The Handbook of Eyewitness Psychology: Volume I: Memory for Events [1 ed.] 0805851518, 9780805851519

Table of contents :
Contents
Preface
About the Editors
PART I Forensic Adult Memory Of Witnesses And Suspects
Chapter 1 Memory for Conversation: The Orphan Child of Witness Memory Researchers
Chapter 2 Interview Protocols to Improve Eyewitness Memory
Chapter 3 The Influence of Emotion on Memory in Forensic Settings
Chapter 4 The Effects of Delay on Long-Term Memory for Witnessed Events
Chapter 5 Alibis in Criminal Investigations and Trials
Chapter 6 Internalized False Confessions
PART II Potential Sources Of Distorted Eyewitness Statements And Postdictors Of Statement Accuracy
Chapter 7 Internal and External Sources of Misinformation in Adult Witness Memory
Chapter 8 False Memory Research: History, Theory, and Applied Implications
Chapter 9 Psychological Impairment, Eyewitness Testimony, and False Memories: Individual Differences
Chapter 10 Recovered Memories
Chapter 11 Using Hypnosis in Eyewitness Memory: Past and Current Issues
Chapter 12 Credibility Assessment in Eyewitness Memory
Chapter 13 Eyewitness Confidence from the Witnessed Event Through Trial
PART III Lifespan Eyewitness Issues: Children
Chapter 14 The Suggestibility of Children's Memory
Chapter 15 Enhancing Performance: Factors Affecting the Informativeness of Young Witnesses
Chapter 16 The Development of Event Memory: Implications for Child Witness Testimony
Chapter 17 False Memory in Children: Data, Theory, and Legal Implications
Chapter 18 A Review of Factors Affecting Jurors' Decisions in Child Sexual Abuse Cases
Chapter 19 Children's Eyewitness Memory: Balancing Children's Needs and Defendants' Rights When Seeking the Truth
PART IV Lifespan Eyewitness Issues: Older Adults
Chapter 20 The Elderly Eyewitness: A Review and Prospectus
Chapter 21 False Memory Susceptibility in Older Adults: Implications for the Elderly Eyewitness
Chapter 22 Eyewitness Memory in Older Adults
PART V Conclusion
Chapter 23 The Relevance of Eyewitness Research: A Trial Lawyer's Perspective
Author Index
Subject Index

Citation preview

HANDBOOK OF EYEWITNESS PSYCHOLOGY

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

HANDBOOK OF EYEWITNESS PSYCHOLOGY Volume 1 MEMORY FOR EVENTS

Edited by

Michael P. Toglia State University of New York, Cortland J. Don Read Simon Fraser University, British Columbia

David F. Ross University of Tennessee, Chattanooga

R. C. L. Lindsay Queen's University, Ontario

Psychology Press Taylor & Francis Group New York London York London

Copyright ©2007 by Lawrence Erlbaum Associates All rights reserved. No part of this book may be reproduced in any form, by photostat, microform, retrieval system, or any other means, without prior written permission of the publisher. First published by Lawrence Erlbaum Associates, Inc., Publishers 10 Industrial Avenue Mahwah, New Jersey 07430 www.erlbaum.com This edition published 2012 by Routledge Routledge Taylor & Francis Group 711 Third A venue New York, NY 10017

Routledge Taylor & Francis Group 2 Park Square, Milton Park Abingdon, Oxon OX14 4RN

CIP information for this volume can be obtained from the Library of Congress

ISBN ISBN

978-0-8058-5151-9-0-8058-5151-8 (case) 978-1-4106-1530-5-1-4106-1530-8 (e book)

Contents Preface

ix

Michael P. Toglia, ]. Don Read, David E Ross, and R. C. L. Lindsay About the Editors

xi

Part I Forensic Adult Memory of Witnesses and suspects Chapter 1 Memory for Conversation: The Orphan Child of Witness Memory Researchers 3 Deborah Davis and Richard D. Friedman Chapter 2 Interview Protocols for Improving Eyewitness Memory

Ronald P. Fisher and Nadja Schreiber

Chapter 3 The Influence of Emotion on Memory in Forensic Settings

Daniel Reisberg and Friderike Heuer

53

81

Chapter 4 The Effects of Delay on Long-Term Memory for Witnessed Events ]. Don Read and Deborah A Connolly

117

Chapter 5 Alibis in Criminal Investigations and Trials 157 Tara M. Burke, John W Turtle, and Elizabeth A Olson Chapter 6

Internalized False Confessions Saul M. Kassin

175

Part II Potential Sources of Distorted Eyewitness Statements and Postdictors of Statement Accuracy Chapter 7 Internal and External Sources of Misinformation in Adult Witness Memory

195

Deborah Davis and Elizabeth E Loftus

v

CONTENTS

vi

Chapter 8 False Memory Research: History, Theory, and Applied Implications JeffreyS. Neuschatz, ]ames M. Lampinen, Michael P. Toglia,

239

David G. Payne, and Elizabeth Preston Cisneros

Chapter 9 Psychological Impairment, Eyewitness Testimony, and False Memories: 261 Individual Differences Sal A Soraci, Michael T. Carlin, ]. Don Read, Terri Krangel Pagoda,

Yvonne Wakeford, Sarah Cavanagh, and Lisa Shin Chapter 10 299 Recovered Memories Steven M. Smith and David H. Gleaves

Chapter 11 Using Hypnosis in Eyewitness Memory: Past and Current Issues

321

Giuliana Mazzoni and Steven ]ay Lynn Chapter 12 Credibility Assessment in Eyewitness Memory

339

Dorothee Griesel and John C. Yuille

Chapter 13 Eyewitness Confidence from the Witnessed Event Through Trial John S. Shaw, III, Kimberley A McClure, and Josie A Dykstra

3 71

Part III Lifespan Eyewitness Issues: Children Chapter 14 The Suggestibility of Children's Memory 401 Laura Melnyk, Angela M. Crossman, and Matthew H. Scullin Chapter 15 Enhancing Performance: Factors Affecting 429 the Informativeness of Young Witnesses Michael E. Lamb, Yael Orbach, Amye R. Warren, Phillip W Esplin, and Irit Hershkowitz Chapter 16 The Development of Event Memory: Implications for Child Witness Testimony 453 Margaret-Ellen Pipe, Karen L. Thierry, and Michael E. Lamb

CONTENTS

vii

Chapter 17 False Memory in Children: Data, Theory, and Legal Implications Valerie E Reyna, Britain Mills, Steven Estrada, and Charles]. Brainerd

4 79

Chapter 18 A Review of Factors Affecting Jurors' Decisions in Child Sexual Abuse Cases 509 Bette L. Bottoms, Jonathan M. Golding, Maggie C. Stevenson, Tisha R. A Wiley, and John A Yozwiak Chapter 19 Children's Eyewitness Memory: Balancing Children's Needs and Defendants' Rights When Seeking the Truth 545 Lindsay C. Malloy, Emilie Mitchell, Stephanie Block, Jodi AQuas,

and Gail S. Goodman

Part IV Lifespan Eyewitness Issues: Older Adults Chapter 20 The Elderly Eyewitness: A Review and Prospectus

Katrin Mueller-Johnson and Stephen]. Ceci

577

Chapter 21 False Memory Susceptibility in Older Adults: 605 Implications for the Elderly Eyewitness

Donna]. LaVoie, Heather K. Mertz, and Tracey L. Richmond Chapter 22 627 Eyewitness Memory in Older Adults Chris]. A Moulin, Rebecca G. Thompson, Dan B. Wright, and Martin A Conway Part v Conclusion Chapter 23 The Relevance of Eyewitness Research: A Trial Lawyer's Perspective

Don Thomson

Author Index

669

Subject Index

699

649

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

Preface In the early days of eyewitness psychology research, investigators such as Stern and Mun~ sterberg focused primarily upon witnesses'recall of actions and events. Later work, begin~ ning about 1970, in contrast, focused heavily upon person identification, an emphasis further stimulated by Wells's (1978) distinction between estimator and system variables; exploration of the latter, in particular, provided the opportunity to examine person iden~ tification in great detail. During this same period the recollection of events was often examined through consideration of more "social" variables like suggestibility and mis~ information. It is striking that virtually all of the research conducted in the areas of both event and person identification at this time used university~age adults as research partie~ ipants. Accordingly, it was reasonable that an edited volume by Ross, Read, and Toglia (1994) focused largely on adults, with considerable emphasis upon person identification issues. However, over the next 12 years, the field of eyewitness psychology expanded dra~ matically in both the sheer amount of active research and the remarkable diversity of the topics studied. The four editors initially considered another one~book project. But our planning was soon guided by both our recognition of the need for an updating of the adult eyewitness literature and by our appreciation of the ways in which the field has expanded, an expansion so great as to preclude the possibility of doing justice to the work in a single volume. As a result, we developed two volumes that reflected an overarching distinction seen in the field over the last century: the recall of events and of persons. The distinction, like most attempts to dichotomize a field, applies only in a general manner because many eyewitness testimony paradigms and much research actually do incorporate investigation of memory for both event and person details. Nonetheless, there is a certain convenience in maintaining this distinction in this Handbook because it has been a practical and mean~ ingful distinction within the realms of research and investigation of criminal events. The companion volume addresses the person component, memory for people, and this volume deals with the first of these components, memory for events. The diversity of the field of ongoing eyewitness research is reflected in this volume in two primary ways. One is the inclusion of several topics that have not previously been reviewed. More specifically, the chapters on memory for conversations or verbal utter~ ances by Deborah Davis and Richard Friedman (Chapter 1), on a taxonomy of alibis and the credibility assigned to them by Burke, Turtle, and Olson (Chapter 5), and research on psychological impairments and individual differences by Soraci et al. (Chapter 9) pre~ sent topics for which the research is very current, the majority of which has been pub~ lished within the last 10 years. The other way in which the current volumes differ from earlier edited volumes is that we include research on participants across the life span, from the astonishing rise in research with children to the numerous investigations with ix

X

PREFACE

university-age participants to a rapidly burgeoning examination of the eyewitness abilities of older adults and the elderly. As special populations, children and older adults are recognized to process, encode, and retrieve event and person information in ways that are different from those exhibited by young university-age adults. More specifically, in regard to children there are six chapters on children's memory and the methods for interviewing them. At the other end of the age spectrum are three chapters dedicated to the special challenges raised by senior or elderly eyewitnesses. In all, this first volume contains 22 chapters and a concluding legal commentary chapter. Volume 2 also concludes with legal discussant analysis. An examination of the contributors in both books reveals a distinct international flavor. In the present volume we recruited researchers from Australia, Canada, England, Germany, Italy, and the United States. The Handbook was conceived as something other than a purely academic researcher's guide to the literature on the some 50 topics gathered together. Instead, chapter authors were provided guidelines that emphasized the needs and interests of those we hoped would make up a significant portion of the target audience: individuals working within the law enforcement and justice systems. In doing this, we actively discouraged the writing of chapters based solely upon an author's last two years in the laboratory and instead encouraged broader reviews of the relevant literature summarized in plain language and followed by recommendations for its application within forensic settings. Obviously, some topics fit into this template particularly well, whereas others met this format somewhat less well. Nevertheless we believe we achieved our overarching intention to provide chapters that would assist those working outside of academe. Although practical applications represent a hallmark of these chapters, the researchers also provide considerable empirical and methodological details of the studies they present. It is very important to point out that such details are not just of interest to the experimentalists who will read this body work, a line of reasoning that is more compelling than when we published our 1994 volume. In fact, two of our authors, Valerie Reyna and Charles Brainerd, argued eloquently for this position, stating that "we both feel that having worked with attorneys and judges that they desperately want and need exactly those empirical and methodological details. Cases turn on those details and attorneys (much to our amazement) wade through them and understand them, and then present those very details in court. Moreover, they much prefer to 'tell their own story' after thoroughly digesting the details rather than having us (the experts) bottom line it for them." We are fortunate to have a particularly legal and especially practical critique of the volume chapters by Don Thomson, a cognitive psychologist and lawyer in Australia. He tells his story in a concluding chapter. We take heart from his chapter because it appears that although not every objective was successfully met in every chapter, there was some considerable overall success in making a useful volume available to our colleagues both in research and within the judicial system in a variety of countries.

References Ross, D. F., Read, J.D., & Toglia, M.P. (Eds.). (1994). Adult eyewitness testimony: Current trends and developments. Cambridge, UK: Cambridge University Press. Wells, G. L. (1978). Applied eyewitness testimony research: System variables and estimator variables. ]oumal of Personality and Social Psychology, 36, 1546-15 57.

About the Editors Dr. J. Don Read is Professor of Psychology and Director of the Law and Forensic Psychology Program at Simon Fraser University. Dr. Read received a B.A. degree from the University of British Columbia followed by M.Sc. and Ph.D. degrees from Kansas State University. His research investigates eyewitness memory, face recognition, reconstructive memory, recollections of childhood abuse, and long-term autobiographical memory and has been funded by the National Sciences and Engineering Research Council of Canada, NATO, and the Alberta Law Foundation. Dr. Read has published some 70 papers and chapters and co-edited four other books including Recollections of Trauma (1997), Eyewitness Memory (1997), and Adult Eyewitness Memory (1994). Dr. Read was recently the Chair of the Society for Applied Research in Memory and Cognition (1998-2002) and the North American Editor of Applied Cognitive Psychology. He is currently Editorial Board member of Applied Cognitive Psychology, Law and Human Behavior, Legal and Criminological Psychology, and The Oxford Press Psychology and Law Book Series.

Rod Lindsay is Professor of Psychology, Queen's University, Canada and Honorary Professor of Psychology, University of Aberdeen, Scotland. He holds a Ph.D. in social psychology from the University of Alberta (1982) and conducts empirical studies on issues in the legal system such as eyewitness reliability, police procedures for obtaining eyewitness evidence, methods of obtaining superior evidence from eyewitnesses, and the effects of eyewitness evidence on jurors. He has published more than 50 articles in social science journals and also has published in law jounals. He has consulted and testified on eyewitness issues in North America, Europe, and Africa (Rwandan War Crimes Tribunal). He has consulted widely on issues of policy concerning eyewitness identification procedures including being a co-author of the National Institute of Justice publication: Eyewitness Evidence: A Guide for Law Enforcement (1999). He has been extensively involved in training of judges, police, and prosecutors in Canada.

Dr. Michael Toglia holds the rank of Professor in the Department of Psychology at State University of New York-Cortland. Since 2003 he has served as the Executive Director of the international organization the Society for Applied Research in Memory and Cognition (SARMAC). He has over 50 scientific publications which in addition to this Handbook include 7 books, most of which are edited volumes devoted to issues on eyewitness memory and testimony. Other editorial experience includes: editor service for 13 journals, a term as Action Editor for the journal Memory, a current appointment on the editorial board for SARMAC's official journal Applied Cognitive Psychology, and reviewer of NSF grant proposals. Similarly, he recently completed a two-year position as a consultant on a

xii

ABOUT THE EDITORS

NIH grant concerning false memory in special populations. He has testified and/or consulted in numerous cases involving the suggestibility of memory, been interviewed by several national newspapers, and appeared on Public Television in the documentary What]ennifer Knew narrated by Susan Saint-James. He is a Fulbright Senior Specialist as well as a Fellow in Division 3 (Experimental), and Division 41 (Psychology and the Law) of the American Psychological Association.

David F. Ross, Ph.D. is a UC Foundation Professor of Psychology at the University of Tennessee at Chattanooga. Dr. Ross received his Ph.D. in Developmental and Social Psychology from Cornell University. Over the last 15 years he has conducted research on factors that influence the accuracy of eyewitness testimony in children and adults, and the psychology of jury behavior. He has published five scientific volumes on psycholegal issues including Children's Eyewitness Memory (Ceci, Ross, & Toglia, Eds.); Perspectives on Children's Testimony (Ceci, Toglia, & Ross, Eds.); Adult Eyewitness Testimony: Current Trends and Developments (Ross, Read, and Toglia, Eds); Handbook of Psychology Volume 1: Memory for Events (Toglia, Read, Ross, and Lindsay, Eds., 2006); Handbook of Psychology Volume 2: Memory for People (Lindsay, Ross, Read, and Toglia, Eds., 2006). Dr. Ross has published numerous articles in top-tier scientific journals; he was invited on two occasions to present his work at NATO conferences on eyewitness memory in children and adults and jury behavior; and he received funding from the National Institute of Justice and from the National Science Foundation for his psycholegal research that subsequently resulted in an award from the American Psychology and Law Society. Dr. Ross is also a member of the editorial board for Law and Human Behavior; and he served as a guest editor for a volume on hearsay testimony and the child witness that appeared in Psychology, Public Policy, and Law, a journal published by the University of Arizona College of Law and the University of Miami School of Law. Most recently Dr. Ross published an invited article on child witnesses that appeared in Law and Contemporary Problems, a journal that is published by the Duke University School of Law. Additionally, Dr. Ross has provided numerous educational seminars to legal and mental health professionals on how to interview the child witness, the accuracy of adult eyewitness testimony, the psychology of jury behavior and jury selection, and techniques for jury persuasion. Dr. Ross has also served as a consultant to lawyers on trial strategy, conducting mock trials, designing opening and closing statements, evaluating eyewitness testimony in children and adults, witness preparation, and jury selection. Dr. Ross also has a specialization in serving as a mitigation expert in death penalty cases. Finally, Dr. Ross has testified as an expert witness on the accuracy of eyewitness testimony in state and federal courts throughout the United States.

HANDBOOK OF EYEWITNESS PSYCHOLOGY

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

I FORENSIC ADULT MEMORY OF WITNESSES AND SUSPECTS

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

1 Memory for Conversation: The Orphan Child of Witness Memory Researchers Deborah Davis University of Nevada, Reno

Richard D. Friedman University of Michigan Law School

The months of late 2003 and early 2004 were filled with accounts of the notorious Martha Stewart insider trading scandal and trial. Hard evidence against Ms. Stewart consisted primarily of the recorded sale of her lmclone stock, and phone records documenting relevant calls to and from Ms. Stewart. Remaining evidence consisted almost exclusively of reports of conversations between Ms. Stewart and her stockbrokers, friends, and aides. Although Ms. Stewart and her brokers described memories of their own conversations, others reported on what they overheard of the conversations, as well as their memories of Ms. Stewart's later comments on them. In the end, Martha Stewart was indicted and convicted of a crime--obstructing justice by lying to investigatorslargely through others' reports of remembered conversations. Martha Stewart's trial may appear to be unusual, if not unique, in its focus on conversation. But nothing could be farther from the truth. Indeed, it is rare that a casewhether civil or criminal-fails to involve some testimony reporting specific individual statements or the tone or contents of specific conversations. Having spent almost 20 years as a jury consultant and expert witness, I (Davis) have been struck by the pervasive presence of testimony regarding conversations and the pervasive absence of expert testimony addressing the varied memory issues inherent in such reports. Although conversational participants and witnesses sometimes choose to lie, their memories are perhaps more commonly subject to the same honest failures and distortions that plague memories for specific persons, locations, objects, and events. It is primarily such honest failures of memory for conversation that we address here, in the hope that our review of relevant

3

4

DAVIS AND FRIEDMAN

research and common areas of testimony will serve as a "call to arms" and road map for future researchers to venture into this underexplored area of witness memory research. We begin by discussing the various ways in which memory for conversation may be relevant to litigation. Since relevance is not sufficient to guarantee admissibility, we then turn to Professor Friedman's review of the rules of evidence governing admissibility of testimony involving statements or conversations. Finally, the bulk of this chapter reviews relevant memory research, focusing on issues relatively unique to memory for conversation and pointing to gaps in existing knowledge.

CONVERSATION AND LITIGATION

Relevant Features of Oral Communications Before describing various purposes for which testimony regarding oral communication may be introduced, it should be noted that many features of verbal interactions can be relevant. Sometimes the only issue is what was said, as is often the case when the conversation involved only two persons, and the time and context are clearly established. But often the issues are far more complex. Witnesses to group interactions may be asked to recall not only what was said, but who said it, and to whom. A particular utterance may have more or less probative value, depending on such contextual issues as the personal histories of the maker of the statement and of those who hear it, what else was occurring concurrently, when the conversation took place, who else might have witnessed it, and which of many conversations in which a particular exchange took place. A suspect might give false testimony concerning a crucial conversation, forgetting that others were there who could impeach his claims. Conversely, that same suspect might falsely remember that others had witnessed a conversation and could verify his testimony, only to be impeached when others denied having witnessed the crucial conversation. A statement might be grounds for termination if made at work and in a particular role, but irrelevant if overheard at the mall. Similarly, accurate memory of which of many similar professional conversations conveyed relevant warnings or advice can be crucial to litigation of malpractice claims. Witnesses may be asked to report subjective impressions of the attitudes, demeanor, emotions, truthfulness, or apparent intent of the participants as context for interpretation of specific statements. Reports of what is not said are sometimes crucialsuch as patient claims that a doctor failed to inform them of crucial facts or risks. Furthermore, the sequence of communications may be significant. Interviewers and police interrogators are increasingly asked to recall who first introduced the possibility of sexually abusive behaviors, or particular references to specific evidence into the conversation; and suspects are asked to recall statements that might have offered incentives or threats to encourage confession. A statement might reflect "guilty knowledge" or "breach of confidentiality," if it occurred prior to widespread awareness of the statement, but not if after.

1.

MEMORY FOR CONVERSATION

5

Finally, memory that a conversation did not actually take place can be crucial. This very form of failure of memory for conversation plays an integral role in a common crime against the elderly known as the "Where's the check?" scam (Schacter, 2001). Essentially, perpetrators convince elderly targets that they had agreed in some way to give the perpetrator a check, whereas in reality the conversation never took place. Those lacking confidence in their memories are susceptible to accepting the perpetrator's claims and often send the check.

Legal Relevance of Oral Communications Specific statements and the many relevant contextual features just described may be offered as evidence pertinent to a wide variety of litigated issues. Perhaps most obviously, a prior statement may be offered as proof of the matter it asserts. That is, the statement may report some event or condition of significance to the litigation and then be presented to the trier of fact for the purpose of proving that the event or condition occurred just as the statement reported. For example, if Wilma testifies that Dottie said to her, "I was at the corner of State and Liberty last week when a blue car came roaring right through the red light," Wilma's testimony is relevant to prove that events occurred just as Wilma says Dottie described them-that is, that the blue car did indeed come roaring through the red light. Similarly, in a criminal case, the prosecution might wish to present testimony by a police officer that witness Whitney told him that "Donald stabbed Victor," for the purpose of showing that Donald did indeed stab Victor. Though the evidence is unquestionably relevant in these cases, it may not be admissible. As discussed in the next section, such use of the evidence raises a problem under the rule against hearsay, and in the criminal case there is also a constitutional problem (posed, in our example, by the inability of the defendant [Donald] to cross-examine his accuser [Whitney], who was not present in trial). Parties to litigation may also offer proof of an oral communication for some purpose other than to prove that a matter asserted by the communication is true. Many contracts are made orally, for example, and a witness may therefore testify that she observed an oral offer, acceptance, or both. Fraud consists of a false representation (which can be oral) that is intended to and does cause another person to rely on it, to that person's detriment. Or a criminal conspiracy may be consummated largely through oral communications. Kingpin's order to Underling, "Deliver two bags to Footman," may help advance their conspiracy to distribute drugs. Sometimes a statement is important for what it reveals about the speaker's state of mind. For example, if Testator says, "Grubber has never been loyal to me," the statement might strengthen a contention that Testator did not intend to make a substantial bequest to Grubber. A statement may also be important to show what information the recipient received. If Landlord received oral notice of a dangerous condition before an accident, such notice might support the conclusion that he was negligent in not correcting the condition. If a phone call gave him timely notice of Tenant's intention to quit the lease, it might end Tenant's obligation. If Donald struck Victor, his contention that he

DAVIS AND FRIEDMAN

6

acted in self-defense might be strengthened by evidence that before the incident someone told him that Victor was a violent man. Often, proof that a witness previously spoke inconsistently with her current testimony is introduced to impeach the witness's credibility. For example, if, some time before the charges were brought against pop singer Michael Jackson, the mother of his accuser had called him an angel, her statement might have strengthened Jackson's defense. In contrast, if her prior statements had been consistent with her current accusations, they may have helped dispel the contention that a laterarising motive or interest had led the witness to distort her testimony.

On the Admissibility of Testimony Regarding Statements and Conversation The preceding examples represent only a few of the many and varied ways in which reports of statements or conversations can be relevant to litigated issues. But relevance does not guarantee admissibility. Normally, the possibility that a witness recounting a prior oral communication may be testifying inaccurately will not cause the evidence to be excluded. But as we will see, the rule against hearsay or, in a criminal case, the accused's right to confront the witnesses against him may nevertheless render the evidence inadmissible.

The Rule against Hearsay. The rule against hearsay might bar admission of, for example, Wilma's testimony that Dottie told her the blue car roared through the red light. The hearsay problem is not that Wilma may be testifying inaccurately. The adjudicative system can address that possibility in the same ways that it would for other events-by allowing cross-examination of Wilma and rebuttal of her testimony. Rather, the hearsay problem arises because Dottie, the out-of-court declarant (the maker of the statement), is not testifying in court, and her statement is offered to prove that it is true. Perhaps Dottie misperceived, misremembered, lied, or spoke inarticulately about what happened. Because the opponent lacks the opportunity to cross-examine Dottie, his ability to test the truth of Dottie's statement would be compromised. Indeed, even if there were no substantial doubt that Dottie made the statement-if, say, the evidence were a videotape of Dottie making the statement-the hearsay issue would remain the same. Many of the illustrations we gave earlier stand in contrast to that of Wilma and Dottie, because in those cases evidence of the out-of-court utterance is offered for some reason other than to prove the truth of a matter that it asserts. In each of these other cases, the fact that the utterance was made is material to the case even if it was not an accurate report of some event or condition. Indeed, in some cases-such as Kingpin's order to Underling to sell drugs-the utterance makes no assertion and hence has no truth value at all. And in the case of an allegedly fraudulent statement, the proponent of the evidence is affirmatively contending that the statement is not true. In other cases, evidence of the statement may have probative value, whether or not the statement was accurate. For example, even if Victor was not violent, the fact that someone told Donald that Victor was violent might support Donald's contention of self-defense. In each of these cases, then, the making of the utterance is an event of significance to the case without reliance on the accuracy of the out-of-court speaker. This oral event

1.

MEMORY FOR CONVERSATION

7

may be proved, as in the case of any other material event, by the testimony of a witness who observed it, and whose testimony is subject to cross-examination. That witness may be the maker of the utterance, its intended audience, or a bystander to the eventwhether known or unknown to the others. Of course, the witness may have failed to perceive or remember the utterance accurately, or her testimony may be intentionally deceptive or unintentionally inarticulate. But as for all of these possibilities, the law takes the attitude that the ordinary means of testing a witness's accuracy, principally cross-examination, are sufficient. The testimony is treated the same as if, instead of reporting utterances, it described an automobile accident, a fight, or a black eye. Now suppose that Wilma does not testify to anything Dottie said, and that she does not remember what happened at the accident-but that she is prepared to testify that she herself said the blue car ran through the light. Significantly, if that testimony is offered to prove the truth of what she asserted, it is still hearsay, notwithstanding that it is the witness's own statement. This principle is mystifying to many, and it has been substantially qualified in recent decades (see, e.g., Federal Rule of Evidence 801 (d) (1) ). But there is more merit to it than is often recognized. Because the prior statement includes information that Wilma does not assert in direct testimony, the driver of the blue car is impaired in cross-examination. Cross-examining a witness about a statement to which the witness does not adhere is like pushing on a string (Friedman, 1996). In short, the law puts great stock in the opportunity to cross-examine a person to test the accuracy of that person's recollection. Without that opportunity, when testimony regarding recalled statements is offered to prove the truth of what they assert, then the testimony is presumptively inadmissible (Federal Rule of Evidence 802), though, in some circumstances, it may be determined admissible in the end. If the witness can be cross-examined, then the possibility of memory failure and inaccuracy will not render the statement inadmissible. To underline this point, consider this hypothetical. A prosecutor is trying to prove that Bully attempted to fix prices at a meeting with competitors. The case does not come to trial until eight years have passed since the meeting. The prosecutor relies principally on two pieces of evidence. One is a memo written the day after the meeting by Competitor One (C1), who died of natural causes three weeks later, saying: "Yesterday's meeting was so extraordinary I want to memorialize it while it is still fresh in my mind. Bully told us that if we raised our prices he would meet them, and that if we didn't he would put us out of business." The other is trial testimony by Competitor Two (C2), to the same effect as the second sentence in C1's statement. Clearly C1's memo poses less of a memory problem than does C2's testimony eight years after the fact. Yet, C2's testimony would clearly be admissible. Although C2 purports to be speaking from personal memory, she is in court and available for cross-examination, whereas the author of the memo presumably would not be. Offered to prove the truth of what it asserts, the memo is hearsay. If not within an exemption to the rule against hearsay, it would be inadmissible. Exemptions. Even if a statement fits within the basic definition of hearsay, it may yet be that the rule does not require exclusion. There are numerous exemptions to the

8

DAVIS AND FRIEDMAN

rule against hearsay. Some consist of rules removing certain categories of evidence (such as certain prior statements made by a person who appears as a witness at trial) from the definition of hearsay. Others simply take the form of exceptions to the presumptive rule excluding hearsay. The most important exemption is the one for party admissions. Statements made by, or attributable to, a party and later offered as evidence against him are not excluded: Anything you say may be used against you. The rule is extended to statements made by an agent of the party, acting within the scope of the agency, and to statements of a conspirator of the party, made during the course of and in furtherance of the conspiracy (Federal Rule of Evidence 801(d)(2)). Numerous exceptions to the rule against hearsay are purportedly based on the twin criteria of trustworthiness and necessity, where trustworthiness, or reliability, is the predominant factor (see generally Strong, 1999, vol. 2, pp. 135-151). We mention just five here. One is an exception for spontaneous declarations, or excited utterances (Federal Rule of Evidence 803 (2)). The received theory, a very contestable one, is that if the declarant was so much affected by the stress of a startling event or condition that her reflective capacities were stilled, the statement is more likely to be reliable. The statement is also more likely to reflect a fresh perception that cannot be replicated by testimony given long after the fact. This exception has sometimes been used to allow evidence of statements made to 911 operators reporting a crime. An exception for statements made for purposes of medical diagnosis or treatment is based on the perception-also a highly contestable one-that a patient is unlikely to speak falsely to a health care provider when her medical care is at stake (Federal Rule of Evidence 803(4)). The exception has been applied in some jurisdictions even when the statement is made to a health care professional who has been consulted in connection with litigation. One of the justifications sometimes offered for this exception is that if a medical expert is offering an opinion on some condition of the patient, then the expert may be allowed to testify anyway to what the patient said, for purposes of demonstrating the grounds for the opinion. It therefore has seemed a small jump to allow the patient's statement to be admitted to prove the truth of what it asserts. The exception for declarations against interest is based on the idea that a statement may be so far against the interests of the declarant-"! chopped down your cherry tree"that it seems unlikely the declarant would have made it unless it were true. Thus, the statement is deemed reliable, and if the declarant cannot testify at trial, this creates a perceived need for the out-of-court statement, and it comes within the exception (Federal Rule ofEvidence 804(b)(3)). These three exceptions are only a sampling of the many that are generally recognized. From time to time, courts and legislators develop new exceptions. For example, in recent years many jurisdictions have adopted rules allowing in some circumstances second-hand evidence of statements by children alleging physical or sexual abuse (see relevant research in later sections). They also raise serious issues of constitutional law. Even if a statement fits none of the specific exceptions, there is a more general one, the residual exception (Federal Rule of Evidence 807). This exception is really an authorization to the courts to admit a hearsay statement, notwithstanding that it falls within none of the specific exceptions, if it appears "more probative on the point for

1.

MEMORY FOR CONVERSATION

9

which it is offered than any other evidence which the proponent can procure through reasonable efforts" and (apart from unusual circumstances) the proponent has given ad~ vance notice of intention to offer the statement and the particulars of it. As this brief summary may convey, the law of hearsay is in an unsatisfactory state. It appears to be a jumble of doctrines based on shaky footings. And yet at the heart of it lies an important principle that until recently has received inadequate attention.

Confrontation. The Confrontation Clause of the Sixth Amendment to the United States Constitution guarantees the accused in a criminal trial the right "to be con~ fronted with the witnesses against him." Until recently, the United States Supreme Court had treated the clause as little more than a constitutionalization of the law of hearsay, aimed at generating reliable evidence. Under Crawford v. Washington (542 U.S. 36 (2004)), if the statement is deemed "testimonial" in nature, then the Confrontation Clause prohibits its use against a criminal defendant, unless the defendant had a prior opportunity to cross-examine the declarant and the declarant is unavailable to testify at trial. In that case, the prior statement is deemed an adequate second-best substitute for in-court testimony. The Crawford Court did not adopt a comprehensive definition of the key term "testimonial." It did indicate, however, that the term applies "at a minimum to prior testimony at a preliminary hearing, before a grand jury, or at a former trial; and to police interrogations." And it recited three definitions as possibilities. The one that seems most suitable would treat as testimonial all "statements that were made under cir~ cumstances which would lead an objective witness reasonably to believe that the statement would be available for use at a later trial" (541 U.S. at 52). Crawford suggests that if a statement is not testimonial the Confrontation Clause will not impose a rigorous bar to admissibility, and the court has made this suggestion definite in Davis v. Washington, decided on June 19, 2006 while this chapter was in press. Davis, the report of which also includes the Court's decision in Hammon v. Indiana, begins, not altogether successfully, the attempt to clarify the meaning of"testimonial." Because Crawford is so new and dramatically different from the previously prevailing doctrine, it is difficult to predict its eventual consequences (Friedman, 2004). If the term "testimonial" is not given an unduly restrictive meaning, however, it offers hope of rationalizing hearsay law. Much of what is worth retaining of the law of hearsay can be explained by the principle that if a testimonial statement is offered against a party, the party ought to have an opportunity to cross~examine the witness.

Exposing Testimonial weaknesses When testimony regarding statements is admitted, the opposition may attempt to expose its weaknesses. As illustrated in Crawford, cross-examination is the most common method and is widely considered the most powerful. Courts often quote the decla~ ration by John Henry Wigmore, the most prominent Evidence scholar of the first half of the twentieth century, that cross-examination is "beyond any doubt the greatest legal engine ever invented for the discovery of truth" (Wigmore, 1974, vol. 5, 1367, at 32; see, e.g., Lilly v. Virginia. 527 U.S. 116, 124 (1999) (plurality opinion); White v. Illinois, 502 U.S. 346, 356 (1992)). With the aid ofleading questions that facilitate control of the

10

DAVIS AND FRIEDMAN

witness, the cross-examiner may fully explore the facts of the case and any conditions that interfere with the witness's ability or inclination to perceive, remember, and narrate the material events accurately. Moreover, the cross-examiner is usually given considerable leeway to go beyond the facts of the particular case, to explore defects in the witness's testimonial capacities, such as impaired perceptual or cognitive abilities or tendencies toward deception. Cross-examination clearly presents a skilled attorney with considerable opportunity to expose weaknesses in the witness's testimony, sometimes with dramatic results. For example, in the Martha Stewart trial, the day after a prosecution witness testified on direct examination to a very damaging statement that Ms. Stewart assertedly made, the witness was compelled to concede that the supposed statement may in fact have been "just a thought in my mind." Although it is perhaps the most common method used to attack or impeach a witness's credibility, cross-examination is only one among many possible tools. Thus, the impeaching party may present (1) rebuttal evidence relevant to any material proposition to which the witness testified, including evidence that the primary witness previously made statements inconsistent with those made at trial; (2) evidence from other witnesses indicating that the primary witness's testimonial capacities were impaired, such as by bias or poor memory; (3) evidence of prior bad acts and of certain criminal convictions that appear to bear on the witness's credibility, (4) a secondary witness's opinion that the primary witness has a bad character or bad reputation for truth-telling; or (5) expert evidence that helps assess the credibility of the primary witness. Not surprisingly, attacks against witness credibility may be met by opposing evidence supporting it. When statements by an outof-court declarant are admitted under exemptions to the hearsay rule, the credibility of the declarant may be impeached and supported by essentially the same methods (apart from cross- and redirect examinations) as if she had been a witness in court. Summary. When evidence of an out-of-court utterance has probative value on grounds other than to prove the truth of a matter it asserts, the law treats the utterance much like other events. A percipient witness may testify to it and thereafter be subjected to cross-examination and other tests of accuracy. In contrast, when evidence of an outof-court statement is offered to prove the truth of a matter it asserts, the evidence will presumptively be excluded as hearsay-not because of the possibility that an in-court witnesses has inaccurately reported the statements, but rather because the statement has not been subjected to cross-examination. Much hearsay is nevertheless admissible, pursuant to hearsay exemptions. The party against whom testimony is offered has a range of tools, including cross-examination, to challenge the witness who gives it; that party has a similar range, but not including cross-examination at trial, when an out-ofcourt statement is offered under one of the hearsay exemptions. Since Crawford, however, it is clear that if an out-of-court statement is deemed to be testimonial and is offered to prove the truth of what it asserts, then evidence of the statement is not admissible against an accused unless he has had an opportunity to cross-examine the person who made the statement and that person is unavailable for trial testimony. This review suggests that the law places great faith in the ability of triers of fact to assess the memory, including the memory for oral communications, of witnesses who are

1.

MEMORY FOR CONVERSATION

l l

in court and subject to cross-examination, and is skeptical of the ability of triers to assess the memory of a declarant who does not testifY in court. Clearly, the importance to litigation of reports of conversations makes salient the empirical question of what factors affect the memory of those making such reports. Answers to this question could affect the legal rules determining the admissibility of reports of statements and conversations, though to what extent is not clear, because those rules are determined in significant part by the judicial system's concept of litigants' rights. But empirical findings certainly could affect the manner in which lawyers present cases to triers of fact, and, to the extent that courts allow such findings to be presented to triers, it could help inform their evaluation of the evidence. With this background, then, we turn to Professor Davis's examination of theoretical and empirical literature on memory for conversation.

STUDYING CONVERSATIONAL MEMORY: THE BEGINNINGS

Arguably, the study of memory for conversation began with the landmark studies of Bartlett (1932) and Allport and Postman (1947, 1958) of the written or oral transmission of stories from one person to another. Participants in these experiments conveyed stories, jokes, descriptive or argumentative prose passages, descriptions of pictures, rumors, or other brief accounts to other participants, who then attempted to reproduce the materials. These were conveyed through a chain of persons to examine the omissions, additions, and distortions that occurred as the information flowed from one person to the next. These studies illustrated many of the same strengths and failures of memory that have since been widely documented across multiple stimulus domains, including written narrative materials and live conversation. First, memory for the gist of the stories was quite good, but with notable failures to preserve the verbatim content. Second, the accounts were substantially degraded both between repetitions by the same person and with successive transmissions through a chain of others. Allport and Postman referred to this degradation in terms of processes of "leveling" and "sharpening," whereby details are dropped so that the account becomes shorter and easier to handle, but incomplete and more sharply focused on the remaining themes and details. Third, in a process Allport and Postman referred to as "assimilation," accounts became distorted by schematic constructive and reconstructive processes to become more consistent with listener expectations. In part, this process of assimilation occurred as a result of failures of understanding. Both Bartlett and Allport and Postman noted that when participants lacked the vocabulary or cultural background to correctly understand what was said, they tended to fill in with guesses consistent with their own experiences and schema-based expectations. These classic studies launched the study of memory for conversation with questions regarding what is remembered. Modern research has since expanded far beyond this to address many of the additional questions introduced earlier-such as who said what to whom, with what demeanor and emotion, in what context, with what response, and so on. We review this literature in the sections to follow. Many readers of this handbook are

DAVIS AND FRIEDMAN

12

familiar with discussions of how memory may fail during encoding, storage, or retrieval. We preserve aspects of this organization, but focus largely upon on aspects particularly important, or even peculiar, to conversational memory.

FAILURES IN ENCODING OF CONVERSATION

Determinants of failures of encoding in conversation are many and often quite different from those leading to failures of witness encoding for other classes of events. Because, like memory in general, memory for spoken utterances is often for the gist of the core of the perceived meaning rather than the literal auditory content, the original interpretation given the utterance is of utmost importance. Perhaps the greatest difference between processes affecting memories for other events versus those for conversation lies in the much larger potential for misunderstanding inherent to statements relative to actions. Clearly, actions and other events can be misunderstood or misperceived. But the order-of-magnitude difference in difficulties of misunderstanding in conversation are attested to by the massive industry devoted to attempts to prevent (e.g., communication skills training programs, workshops, and literature) or correct (e.g., marital therapy) conversation-based misunderstandings. Why should this be the case?

Stimulus Ambiguity and Levels of Meaning in Conversation Perhaps most fundamentally, conversational content offers greater potential for misunderstanding because of the many different levels at which meaning is encoded and understood, and hence at which comprehension can fail. With respect to the verbal aspects of conversation, the following levels of meaning have been addressed (though nonverbal cues are central to accurate comprehension as well; e.g., Planalp & Knie, 2002).

Leuel 1: Speech Perception At the speech perception level, the sensory sound stream must be successfully parsed into phonemes and words. This level of comprehension may be compromised by both environmental interference and chronic or acute personal impairments in hearing. Older persons, for example, have greater difficulty with speech perception, particularly in more difficult environmental contexts (e.g., see Schneider & Pichora-Fuller, 2000, for a review). If the person successfully comprehends the sounds, comprehension then takes place across at least six different levels. Models of discourse comprehension, which focus primarily on text comprehension, include five additional levels of comprehension (see Graesser, Millis, & Zwaan, 1997, for a review).

Leuel 2: Surface Code or Verbatim Wording The second level of comprehension is the surface code, consisting of verbatim wording and syntax. This level is rarely preserved in memory and tends to be replaced by a

1.

MEMORY FOR CONVERSATION

13

general representation of higher level interpretations of the semantic meaning of the utterance (e.g., Bock & Brewer, 1974; Brewer & Hay, 1984; Graesser & Mandler, 1975; see Fletcher, 1994; Graesser, et al., 1997).

Leuel 3: Propositional Content The third level of interpretation is called the textbase or propositional content. This consists of explicit propositions (such as "It is raining.") in stripped-down form that preserves the meaning, but not the exact surface code, along with a few basic inferences needed to make the propositions coherent. Often the exact wording of a statement is crucially important in forensic contexts. For example, the statement "I'm going to kill the son of a bitch!" may constitute a clear threat, whereas "The son of a bitch doesn't deserve to live!" might be taken as a statement of opinion. The two, however, would often be understood as equivalent, and consequently remembered equivalently laterparticularly if the person in question ended up dead. Initial understanding, and therefore memory, is also affected by simple ability to track the propositional content of conversations, which is often difficult. Some speakers tend to be particularly disorganized and difficult to track. Generally, linguistically simple statements are understood more easily. Long sentences that include numerous clauses, subordinate clauses, diversions from the main point, and so on, impose significantly greater processing demands on the listener. To understand such a sentence, a hearer has to hold all information extracted from all clauses in working memory long enough to integrate all parts, extract an overall meaning, and transfer the meaning to long-term memory. Hence, older or distracted listeners and those with otherwise impaired processing capacity are particularly likely to fail to comprehend complex utterances (see Davis & Loftus, 2005).

Leuel 4: The Referential Situation Model The fourth level of meaning, the referential situation model, refers to understanding of the referents of the propositions-or what is being talked about (in other words, understanding of the people, places, and other referents in the microworld the statements are referring to). The model of Bransford and Johnson (1972), for example, provided a well-known demonstration of difficulty in understanding a paragraph describing doing laundry among those who were not told that the paragraph was about "doing laundry." The text was comprehensible only for those who knew what was being talked about. When the paragraph was read without the title "doing laundry," virtually none of the participants understood that it described the process of doing laundry. Even simple pronouns, such as he, as well as more complex abstract concepts, such as work or spirit, likewise can be understood only in the context of the referential situation model, which the listener relies upon to understand who and what is being talked about. It is at this level that understanding often fails-often as a result of failures of "perspective-taking" on the part of the speaker and/or listener.

14

DAVIS AND FRIEDMAN

Failures of Perspective-Taking and Mistaken Assumptions of Common Ground Allport and Postman (1958) noted (as did Bartlett) that "even those with normally acute hearing often mishear or misinterpret words for which they have no supporting mental context" (p. 60) and that these misunderstandings form the basis of distortions in later reports. Lack of supporting mental context (i.e., an inadequate "referential situational model") can stem from features of the listener, such as lack of vocabulary or lack of attention to relevant contextual features. But many result from speaker failure to provide needed clarity and context. Speakers are often disorganized and egocentric, failing to provide sufficient structure, clarity, and context for their listeners. They leave out referents, comment suddenly on something they have been thinking about (for which the listener has no context), and generally engage in sudden topic shifts, rendering it difficult for the listener to track the new context. These are examples of failures of "common ground," or shared knowledge and context, which is necessary to ensure that the listener can understand the words and overall meaning of a speaker's remarks. To provide such clarity, the speaker must understand the knowledge level and perspective of the listener and fill in missing knowledge the listener would need to comprehend the speaker's intent. In turn, to understand, the listener must try to understand the utterance through the perspective of the speaker. Indeed, many have argued that perspective-taking and shared common ground are the fundamental tasks of accurate communication and communicative competence (e.g., Krauss & Fussell, 1996), and hence so is specifically the ability to coordinate the perspectives of speaker and listener so that utterances stay within their "common ground" (e.g., Clark, 1996), most fundamentally with respect to referents-or what is being talked about (see Clark, 1996; Holtgraves, 2002, for detailed reviews of the many forms of common ground necessary for understanding). Listeners do take advantage of their understanding of common ground to interpret speaker statements. However, listeners also tend to manifest an egocentric bias, often failing to sufficiently consider common ground, relying more strongly on linguistic precedents within the conversation and expectations of topical coherence (e.g., Barr & Keysar, 2002). In turn, speakers on the whole successfully adjust their statements to reflect listener perspectives (see Holtgraves, 2002, for a review). This is less likely, however, when the speaker is under pressure (e.g., Horton & Keysar, 1996). Apparently, like many processes requiring controlled processing (conscious efforts to accomplish a particular task), perspective-taking is replaced by more automatic word choices when cognitive resources are compromised. In this respect, unlike memory for other events, witness memory for conversation depends not only upon the witness and the conditions of witnessing, but also upon the condition of the speaker and the clarity with which he or she produces the to-be-remembered statements. Hence, the characteristic communicative abilities of both witness and speaker will affect the accuracy of witness understanding and interpretation, and thereby memory. Notably, as well, third-party witnesses to others' conversations may be particularly lacking in the common ground necessary for accurate understanding-especially if the participants do not intend to communicate to the third party. In consequence, the

1.

MEMORY FOR CONVERSATION

15

eavesdropper may well completely misunderstand the speakers' intended meanings. The relative difficulty in tracking others' conversations for which one lacks common ground was cleverly demonstrated by Kent, Davis, and Shapiro (1978), who disassembled transcriptions of conversations between either friends or strangers and later asked new subjects to try to reconstruct the conversations. Participants were much less accurate at reconstructing conversations between friends-presumably reflecting the unique common ground shared between acquainted speakers, which renders their conversations less comprehensible to others. Unfortunately, we cannot assume that findings from studies of participants' memories of conversation will generalize to observers of others' conversations. Indeed, a strong argument against such generalizations has been made by Benoit, Benoit, and Wilkie (1996; see also Street, 1993), who also provided evidence of differences in amount of memory, what is remembered, and factors affecting memory between participants and observers (for example, observers appear to make more schema-based false reports). Participants and observers have different background information, tasks (e.g., listening, planning contributions, managing face, etc.), and goals during conversation that can affect information processing and hence memory (e.g., Street, 1993). Failures of Common Ground in Professional Conversations. Failures of communication are often at the heart of civil malpractice suits (see case examples in Davis, Kemmelmeier, & Follette, 2005), and, not surprisingly, crucial issues of negligence are evaluated on the basis of reported conversations between the parties. Both original encoding and later memory of these conversations can fail because of mistaken assumptions of common ground by both sides. This problem has been studied most extensively in the context of medical interactions, although other interactions involving complex or unfamiliar material will doubtless suffer the same problems of understanding. Patients sometimes fail to tell their doctors specific problems, assuming that doctors can detect them without discussion. The doctors might later be sued for malpractice over failure to diagnose the problems. Doctors, on the other hand, tend to use sophisticated and highly specialized vocabulary that is inaccessible to the layperson and no doubt is partially responsible for the fact that even immediately after their visits patients are unable to recall about half of what they have been told by their physicians (DiMatteo, 1985; Ley, 1966, 1979, 1986; Ley & Spelman, 1967; see Schneider, 1998, for a review); and, as with memory for other events, memory for conversation decays further with time (e.g., Stafford, Burggraf, & Sharkey, 1987). Domain-specific knowledge, which patients often lack, enhances both comprehension and memory for text or conversation (e.g., L. M. S. Miller, 2003). Unless patients aggressively ask for clarification of unfamiliar concepts, they may fail to understand much of what is said, as a consequence fail to remember it, and hence later deny being told crucial information. In one study, for example, 73% of breast cancer patients misunderstood the word median, as in median survival time (Lobb, Buttow, Kenny, & Tattersall, 1999). Even terms that fall within everyday language (e.g., those relating to frequency, such as frequently, rare, likely, probably, often, etc.) are understood differently by patients and physicians (Girgis & Sanson-Fisher, 1995; Kong, Barnett, Mosteller, & Youtz, 1986), and no reliable way to ensure that patients and physicians share under-

DAVIS AND FRIEDMAN

16

standings of word or probability estimates of risk, treatment options, or side effects has been identified (Mazur & Hickam, 1991; Nakao & Axelrod, 1983; Ohnishi et al., 2002; Robertson, 1983; Woloshin, Ruffin, & Gorenflo, 1994). What is clear is that profession~ als need to engage in extra effort to ensure that their statements fall within shared "common ground" with the patient-as well as to help patients remember vital information and instructions. Conversational Relevance as Central to the Referential Model. Listener-referential situational models refer to assumptions of what is being talked about, as noted earlier. This understanding is based in part on expectations that others will follow conversational rules and conventions (see Holtgraves, 2002; Turnbull, 2003, for reviews). Since listeners generally expect utterances to contribute to overall coherence and to be relevant to the general context as well as preceding utterances (Grice, 1975), statements that do not clearly relate to the general context of the interaction or to preceding statements likewise do not fit the listener's current referential model of what is being talked about. Hence, seemingly irrelevant contributions posed by disorganization and sudden topic shifts often confuse the listener and result in poor memory for the utterances, along with negative impressions of the speaker and/or negative inferences about speaker attitudes toward the listener (e.g., Davis & Holtgraves, 1984; Davis & Perkowitz, 1979). Davis and Holtgraves (1984) argued that, particularly in response to specific questions, listeners scan conversational remarks for a relevant answer and may simply ignore or misunderstand irrelevant information while waiting for the relevant answer. The authors found that memory (both recognition and recall) was clearly impaired for even tangentially relevant compared to fully relevant replies. In fact, when asked to recall each reply verbatim, subjects often simply wrote "didn't answer the question" for tangential replies.

Level 5: Speech Acts and the Communication Level The fifth level of interpretation is called the communication level; this refers to what the speaker actually intends to communicate. "Speech act theory" (Austin, 1962; Grice, 1975; 1989; Searle, 1969) was developed to explain the way in which hearers understand speakers' intended meaning and what "speech acts" speakers intend to accomplish with an utterance (for example, to express feelings, make a request, forgive or insult someone, etc.). Speech act theorists note that speakers often communicate indirectly, in that the propositional content of the utterance might be quite different from the intended meaning. For example, the statement "It sure is hot in here" might be an indirect speech act requesting the hearer to open the window. Likewise, in response to the ques~ tion "What did you think of my presentation?" the indirect response "It's hard to give a good presentation" might be intended or interpreted as meaning the respondent did not like the presentation (e.g., Holtgraves, 1999). Indirect meaning can likewise be conveyed by a wide variety of additional conversational behaviors (see Holtgraves, 2002, for a review). Many of these involve violations of conversational "rules" or other conventions that are interpreted as conveying indirect meaning. For example, a statement that

1.

MEMORY FOR CONVERSATION

17

is obviously false (thereby violating the "truth maxim" that conversational contributions should be truthful; Grice, 1975) will likely be interpreted as irony or sarcasm, which implies the opposite meaning. Regardless of the specific manner in which they are expressed, however, indirect statements are more open to misinterpretation (Holtgraves, 1994, 1997, 1998, 1999). A statement may have both direct and indirect meaning. For example, an employer's complaint that "Good help is hard to find!" directly states the speaker's opinion about the difficulty of finding good help. But when uttered in the presence of a specific employee, it may also be an indirect comment on the speaker's opinion of that employee. Unfortunately, the inherent ambiguity of indirect speech acts can lead to serious social misunderstandings. The employee may feel insulted, for example, even though the boss was not referring to him at all. Many speech acts are performed indirectly by convention. For example, it is considered more polite to make requests indirectly (e.g., "It sure is hot." versus "Open the window."), particularly when one is addressing someone of equal or higher status. Other politeness strategies likewise involve ambiguity in meaning (see Brown & Levinson, 1987; Holtgraves, 2002; Pennebaker, Mehl, & Niederhoffer, 2003, for reviews). Unfortunately, disputes giving rise to litigation can arise over whether a particular speech act (such as informing someone of crucial information, making a request, giving an order) did or did not take place because it was done indirectly, and therefore may never have been understood (and therefore remembered) by the hearer as it was intended. There are strong individual differences in tendencies to state things indirectly and to look for and recognize indirect meanings in the remarks of others. For example, members of collectivistic cultures (e.g., Korea) versus individualistic cultures (e.g., United States), psychology majors as opposed to business majors, and those high in self-consciousness and need for cognition are more prone to look for indirect meaning. Those from collectivist cultures and those high in private self-consciousness are more prone to use indirect statements, and those high in assertiveness and high in status are less prone to do so (e.g., Holtgraves, 1997, 2002). Given such strong individual differences, some people are more likely to fail to perceive indirect meanings the speaker does intend, whereas others are more likely to read indirect meaning into remarks where none exists.

Level 6: Relational Implicatures and Interpretations One of the most pervasive levels of interpretation of speech and conversation is the relational level, which concerns both intended and unintended messages conveyed about the relationship between participants (e.g., Watzlawick, Beavin, &Jackson, 1967). They provide information about how each person sees the other, how they see themselves, how they view the relationship within and outside the immediate situation, and how they would like to see the relationship proceed in the future. Speech is monitored for clues to the intentions of the other person: Does he like me? Does he think well of my work? Will she spend more time with me? Does this mean she would like me to ask her out? Who is in charge here? Is it time to take this to the next level? Does she find me interesting?

DAVIS AND FRIEDMAN

18

Others' statements and behaviors in conversation are constantly monitored for these and other evaluative implications-particularly professional conversations and those between intimates or would-be intimates. And not surprisingly, researchers investigating social interaction have widely addressed how such messages are deliberately conveyed, how they are interpreted, and sources of misunderstanding. Language researchers tend to focus on two relational dimensions, status and solidarity (i.e., closeness, liking between participants), whereas relationship researchers, of course, tend to focus primarily upon implications for solidarity. It is our observation that misunderstandings at the relational level of communication are often at the heart of both criminal and civil cases. Men's misperceptions of the implications of a woman's conversational remarks for questions such as "Does she want me to ask her out?" or "Should I take this to the next level?" are widely implicated in cases involving rape or sexual harassment. They are also frequently responsible for problems with students. For example, the statement "I'm sorry, but I can't understand your argument" might be interpreted and later remembered as "You just don't know how to communicate" or "You're stupid." In such cases, the gist of the core of the perceived meaning is simply "She doesn't like me." or "She thinks I'm stupid." or "She doesn't respect my work." Or, in the rape example, the gist of the conversation might later be remembered and reported as simply "She made it clear that she wanted it." Although listeners sometimes misperceive relational implications of others' statements, speakers do sometimes deliberately encode indirect relational messages. Rather than tell someone he doesn't like him, the speaker may choose non-intimate topics, respond minimally to conversational overtures from the other person, and generally use distancing forms of speech (such as formal rather than informal dialects, I rather than we, less polite forms, and so on). A speaker who wishes to assume a relatively higher status position might use more formal speech or more direct commands than indirect requests (see Holtgraves, 2002, for a review). Again, like all indirect communications, these relational implications are more susceptible to misunderstanding, particularly among the less socially skilled.

Leuel 7: Broader Impressions and Inferences Impressions of Speakers and Listeners. Not surprisingly, conversational participants and observers form strong impressions of the participants. Language researchers have investigated three primary dimensions of these impressions: status, solidarity, and competence. Status, for example, tends to be inferred from the use of formal linguistic forms, less polite forms (or greater assertiveness), use of informal forms of address toward the other person (given versus surname), and generally more distancing language (such as use of I rather than we). Competence tends to be inferred from such variables as higher speech rate, greater lexical diversity, and forms of speech indicating greater certainty. Solidarity tends to be inferred from such variables as intimate disclosures, informal speech forms and terms of address, or lesser politeness. Overall, liking is strongly affected as well by conversational behaviors, particularly those deemed inappropriate (see Holtgraves, 2002; Pennebaker et al., 2003, for reviews).

1.

MEMORY FOR CONVERSATION

19

Clearly, status, solidarity, and competence are only three among a host of impres~ sions routinely formed of others. However, although such impressions are born of the specific content of the interaction, they quickly become independent of it. Even as the specific bases of the impressions are lost, the inferences and conclusions drawn from them are retained (see Wyer, 2004, for a review of the independence of impressions and conclusions from foundation information in memory). Hence, it is important to recog~ nize that although observers to conversations will often report the conclusions they have drawn, the basis of those conclusions has often been lost, and hence their validity cast into doubt. Reports such as "They were fighting," "One was clearly the boss," "The person must have been a professional," or "They were friends" will often be unknowingly based on observations of linguistic form. But the inference may be completely wrong. A person overheard using baby talk to give instructions might be assumed to be talking to a child, whereas baby talk is often used to talk to the elderly, the sick, the mentally retarded, the temporarily incapacitated, and even romantic partners.

Implicit Causality. Some researchers have specifically investigated the impact of wording on attribution of responsibility for described events and states of mind. For ex~ ample, the sentence The dog bit the man is likely to result in greater responsibility being attributed to the dog than the sentence The man was bitten by the dog. Generally, the sub~ ject of the sentence tends to be seen as more responsible for a described action than the object, although the reverse can be true for a described state of mind (e.g., Fiedler & Semin, 1988; Holtgraves, 2002; Rudolph & Forsterling, 1997). For example, if]ane hates Ruth, the tendency is to believe something about Ruth was the cause. Overall Meaning of the Content. In addition to inferences about the speaker, listener, and the relationship between them, listeners draw broad inferences about the overall message the speaker intends to convey. They process between the lines, remembering not precisely what was said, but rather what was pragmatically implied, the inferences they draw about the speaker's meaning or intentions, or what appears to make most sense in a given situation (e.g., Harris & Monaco, 1978; Hilton, 1995). Such inference processes can lead to what might be considered errors in judgment tasks, although some have reinterpreted these "errors" (such as under-use of base rate information) as rational interpretations (e.g., Krosnick, Li, & Lehman, 1990) based on conversational maxims such as "be relevant" (Grice, 1975). That is, the problem solver would assume that any information given to him or her to solve the task would be relevant to the solution (e.g., Hilton, 1995) and hence give it greater weight in some cases than is logically warranted. The importance of conversational inferences was illustrated in a legal context by Fulero and Finkel (1991), who found juror inferences from the actual content of expert testimony to strongly affect verdicts. The authors presented three versions of a trial involving expert testimony regarding insanity. In one version, the expert presented only the diagnosis that the patient suffered from delusional paranoid disorder. In the second version, the expert provided the diagnosis, but also testified that because of the delusions associated with the disorder, he would be unable to consider the consequences of his behavior or to appreciate the wrongfulness of his acts. Finally, for the third version, the

20

DAVIS AND FRIEDMAN

expert testified to the preceding facts, and further stated the conclusion that the defendant was insane at the time of the crime. No differences in verdict were obtained between the three levels of testimony, in part because jurors tended to infer the final determination of insanity in all conditions, and/or to falsely remember that the expert had stated that conclusion even when he had not. The importance of conversation-based inferences has also been extensively discussed in the literature on coerced confessions (e.g., Kassin, 1997; Kassin, this volume). Kassin noted that judges typically exclude confessions elicited through explicit threats and promises, but admit those elicited through implied threats and promises. However, Kassin's research showed that explicit and implied threats and promises are understood equivalently (Kassin & MeHall, 1991), leading Kassin to conclude that common interrogation tactics essentially violate the intent of the law by using pragmatic implication to accomplish what they are forbidden to do explicitly.

Interpretative Distortions Due to Schematic Processing Schemata are organized knowledge structures that include beliefs and expectations concerning the nature, characteristics, and behaviors or functions of objects, people, events, and other cognizable entities. Such schema-related beliefs and expectations tend to bias interpretation of incoming information toward consistency with expectations (see Davis & Loftus, this volume, for more extensive discussion of the effects of schemas on information processing and memory). Dunning and Sherman (1997), for example, showed that stereotype-based expectations affect inferences regarding what a behavior was, what stimulus the person reacted to, and what caused the person to act. In conversation this would translate to "What did the statement mean?'' "What did the other person say or do to trigger it?" and "Why did the speaker react this way?" Multiple schemata affect interpretation in conversation. At the simplest level, word and sentence meaning is understood in light of schematic expectations involving, at a minimum, syntax, semantics, grammar, expectations of coherence, other expectations about the structure of discourse, and schematic understanding of the speaker and context. In addition, people often fill in sections of speech they failed to hear clearly with words consistent with schema-based expectations. This is particularly likely to occur when the person has very strong expectations for what is likely to be said, when they are distracted and not processing carefully, or when hearing is poor. This is particularly characteristic of elderly listeners, who rely on expectations for content as well as knowledge of syntax and semantics to fill in gaps (Nussbaum, Pecchioni, Robinson, & Thompson, 2000). A particularly interesting example of this occurs with "garden path" sentences that lead the hearer to expect a particular ending. Even when the expected ending is not stated, it may later be falsely remembered. For example, the fragment "I'm gonna ... !"might later be remembered as "I'm gonna kill you!" if the speaker was angry when the statement occurred and the target ended up dead. Schemas for other persons or the circumstances of an interaction also affect expectations, which in turn determine how ambiguous statements will be interpreted. For example, John's ambiguous utterance "I might have to do something serious about this" might be interpreted as more hostile if John is a drug dealer rather than a banker (reflect-

1.

MEMORY FOR CONVERSATION

21

ing a stereotype about the speaker), if his target was a known criminal embezzler rather than the employee of the month (reflecting a stereotype of the referent), if it was said in response to a misbehavior rather than laudable behavior (reflecting schematic expectations oflikely responses to target behaviors), or if he was in a bad versus a good mood (reflecting theories about the impact of mood on hostility), to give but a few examples. The listener might even interpret the statement as a plan for murder if John was a known drug dealer and the current context would prohibit open discussion of murder. Generally, crime-related schemata regarding motives, likely perpetrators, manner of execution, and other issues (Davis & Follette, 2002, 2003; Vanous & Davis, 2001) can lead to distorted interpretation and recall of statements by involved parties. Schemas for particular individuals can also distort memory for conversation. Wilson, Simpson, and Smith (2005), for example, showed that adolescents' memory for the hostility of a discussion of a problem with their parent was distorted toward consistency with their general feelings for the parent-a finding of importance for cases involving allegations of abuse.

Motivated Interpretation The recent focus on "motivated social cognition" has underlined the pervasive processing influences exerted by motivation. Similar motivations can be expected to guide and distort understanding and memory in conversation. Erotomaniacs, for example, who are highly motivated to believe their targets love and desire them, might be prone to misinterpret many neutral behaviors as indicating love. Those high in attachment anxiety, who are chronically hypervigilant in monitoring relationship implications of others' behaviors, are prone to see relationship implications and generally personal evaluations in others' communications where none are intended, whereas those high in attachment avoidance tend to engage in self-protective suppression of attention to relationship implications (see Mikulincer & Shaver, 2003, for a review). Narcissists' recall of social interactions tend to undergo self-aggrandizing, ego-protective distortions, particularly in egothreatening circumstances (e.g., Rhodewalt & Eddings, 2002). Although we have not found studies addressing this issue specifically with regard to conversational memory, in light of substantial research documenting motivated processing of health-related and other personally relevant information (e.g, Campbell & Sedikides, 1999; Renner, 2004), it is clear that motivated distortions in the interpretation of conversational behaviors lie at the heart of many litigated issues, such as medical malpractice, sexual harassment, rape, wrongful termination, discrimination, and others, and of memory errors among the witnesses.

Failures of Perception Given that conversational meaning is highly contextually dependent and encoded and interpreted at multiple levels, failures of understanding may result from a relatively wide range of perceptual failures. Having considered this complexity with which conversational meaning must be processed, we can now step back and consider what must be accurately perceived to promote accurate understanding.

DAVIS AND FRIEDMAN

22

Requirements of the Perceptual Field It might seem at first glance that simple perception of verbal utterances would require only an adequate ability to hear. Instead, however, there is ample evidence to document the importance of multichannel perception for accurate understanding of speech (see recent reviews of conversation and language processing by Clark, 1996; Holtgraves, 2002; Krauss & Chiu, 1998; Levinson, 2000; Planalp & Knie, 2002). Since such nonverbal cues as tone of voice, gestures, facial expressions, body movement, and posture may add to or completely alter the meaning of an utterance-such as may be the case with sarcasm, irony, or generally for utterances with nonliteral intended meaning-it is often necessary to see the speaker to fully understand his or her meaning. Multichannel perception, as well as attention to context, is also important for evaluation of the emotional underpinnings of the message (although people are surprisingly poor at judging the exact emotion others are expressing; Carroll & Russell, 1996) and of its veracity (see DePaulo et al., 2003; Planalp & Knie, 2002). Furthermore, it is often necessary to see both participants clearly, as awareness of what the speaker is responding to (for example, the facial expression of the other participant) may inform the interpretation of the speaker's meaning or judgments of veracity. Unfortunately, when describing others' actions perceivers tend to omit contextual information such as what triggered the action, leading hearers to make less accurate judgments of the actor's behavior than the original perceiver (e.g., Gilouich, 1987). Awareness of the general context of an utterance is also often necessary for accurate interpretation of both simple meanings, such as a referent (e.g., nut as "food" vs. nut as "crazy person"; e.g., Cook & Myers, 2004), and those as complex as the general meaning of the utterance (Sperber & Wilson, 1995). To interpret the statement Geez, it's hot in here, one might need to know the gender and age of the speaker, relevant features of other people in the room, room temperature, the status of the windows or thermostat, and the preceding behaviors and statements of others present to know whether the statement was a simple complaint, a postmenopausal hot flash, an indirect request for someone to open the window or alter the thermostat, or a comment on beautiful bodies or the hostility of an ongoing exchange. Hence, "what a speaker means with an utterance (what he intends to accomplish) can only be derived with some reference to a context." (Holtgraves, 2002, p. 5). Therefore, accurate perception of the literal utterance as well as relevant nonverbal cues and contextual features requires adequate opportunity to see, hear, and attend to a relatively wide focal and peripheral perceptual field. The importance of each perceptual channel is enhanced as the impairment of others increases. Visual cues, such as gestures or lip movements, can compensate for an impaired auditory channel and make message transmission more resilient to background noise and other auditory disturbances. This compensatory process is particularly important for the hearing impaired, who chronically read lips to make up for auditory failures (Davis & Loftus, 2005; Nussbaum & Coupland, 2004; Nussbaum et al., 2000; Villaume, Brown, & Darling, 1994). However, such compensation is not fully effective, as interpretations based on the availability of or attention to restricted perceptual channels may be in error. The presence of noise, distraction, personal sensory impairments, or other factors that cause difficulty in simple perception has the additional effect of increasing cognitive

1.

MEMORY FOR CONVERSATION

23

load and thereby compromising comprehension. That is, even though individuals may understand all of the words, they are less likely to understand the overall meaning, draw appropriate inferences from utterances, and integrate new information with their existing knowledge, thereby resulting in diminished comprehension and retention (e.g., Schneider, Daneman, Murphy, & See, 2000). In effect, one must try so hard to understand the words that the meaning of the entire sentence is lost. This problem often plagues older listeners, whose processing resources are already compromised relative to those of younger people, particularly in the presence of background noise (see Davis & Loftus, 2005; Schneider, Daneman, & Pichora-Fuller, 2002). Some evidence also suggests that distraction (specifically multitasking) can selectively impair nonverbal decoding abilities in introverts (e.g., Lieberman & Rosenthal, 2001) and thus impair full comprehension. Unfortunately, even if perceptual abilities are intact and the perceiver has full access to the perceptual field, necessary context sometimes also includes knowledge of the recent and/or historical conversational context of current remarks; of chronic and acute characteristics or states of speakers and others present; of relevant knowledge, emotions, attitudes, and other mental states of all present (and their beliefs about these states in others); of the general purpose of the interaction; of current goals of the parties; and of other relevant internal states of speakers and listener(s). Recall that understanding of these internal states is referred to as perspective-taking, and is necessary for understanding others' statements as well as for encoding one's own statements to others in understandable terms (see Holtgraves, 2002, for a review). To summarize, then, accurate encoding of others' communications requires a wide range of both external perception and internal knowledge. This extreme context dependence of language interpretation sets it apart from other witnessed events. If one person shoots another, the nature of the action itself is crystal clear. The motive may be at issue, but the action itself remains the same. Statements, on the other hand, have multiple possible meanings that can be made unambiguous only by reference to the full context of their utterance. Also, as noted earlier, accurate understanding depends not just upon the listener, but upon the speaker as well. Hence, the communication abilities of all parties will affect listener encoding and memory. A host of individual differences in the ability to successfully encode and decode verbal and nonverbal communications has been examined in the communications literature. A review of such differences is beyond the scope of this chapter (however, see Riggio & Feldman, 2005).

Failures of Attention To understand the conditions under which a particular conversation, part of a conversation, or a particular person participating in the conversation will be best understood and remembered, one must consider what determines the amount, depth, and direction of attention the observer pays to the conversation, its context, and its participants. Attention will determine not only memory for gist, but also for specific wording. Kintsch and Bates (1977), for example, hold that the specific wording of utterances is encoded and remembered to the extent that the wording itself draws attention. Davis and Follette (2001; see also Schacter, 1999, 2001) reviewed in detail the way in which attention is generally determined by both personal characteristics and states, and contextual

24

DAVIS AND FRIEDMAN

features of a target event. Here we focus more exclusively upon issues and investigations specifically regarding the role of attention in memory for conversation. Individual Differences. Conversational memory will be determined partly by the overall extent to which attention is focused on the conversation or one's interaction partners. Conversational involvement, for example, which might be considered an index of degree of attention to the conversation as a whole, enhances overall memory for ideas mentioned in the interaction (Cegala, 1984). Similar relationships have been found for self-schemata for sociability (Dworkin & Goldfinger, 1985), low attachment avoidance (which in part measures the lack of interest in other people; Miller, 2001), and low levels of anxiety (Stafford & Daly, 1984). Interpersonal competence has also been related to memory for others' remarks, but primarily for those in same-sex interactions, and more so for women than for men (e.g., Miller, de Winstanley, & Carey, 1996; Miller & de Winstanley, 2002). Furthermore, participants tend to remember more than observers (e.g., Benoit & Benoit, 1994; Benoit, Benoit, & Wilkie, 1996; MacWhinney, Keenan, & Reinke, 1982; Stafford, Waldron, & Infield, 1989), and observers who know the participants remember more than those who do not (MacWhinney et al., 1982). Observers in one study produced more errors (reports of segments not actually in the conversation) than participants (Stafford et al., 1989). Arguably, these differences are a function of greater interest and attention among participants than among observers, and for acquaintances versus strangers. Finally, there is evidence that females are more attentive and verbally and nonverbally responsive to interaction partners' remarks than males (see Miller & de Winstanley, 2002: Stiles et al., 1997, for reviews) and that younger participants remember conversation topics better than older (although recall for both was relatively poor, 67% versus 55%; Kausler & Hakami, 1983). Attention in conversation is also selective and thereby produces selective memory for specific statements or aspects of the conversation. Attention to Interesting or Goal-Related Information. As is the case for all stimuli, conversational contributions relevant to one's interests or personal processing goals elicit more attention. Such interests and goals may be situation specific or chronic reflections of personality-related concerns such as power, control, impression management, self-esteem maintenance, attachment motivations, and the like-or a combination of some or all of these. This point was illustrated by Hirst and Gluck (1999) in their study comparing the testimony of John Dean in the 1973 Senate Watergate Hearings with transcripts of White House recordings of 1972 conversations between Richard Nixon, H. R. Haldeman, and Dean. They used an unpublished transcription prepared by the Impeachment Inquiry Staff for the House Judiciary Committee at the time of the Watergate hearings (National Archives, 1996, http://nixon.archives.gov/find/subjects/deepthroat.html#tapes), which contained more complete transcriptions than the White House transcripts (Gold, 1974). From these, the authors selected the transcription of the September 15 meeting of Nixon, Haldeman, and Dean, in which they reviewed events related to Watergate. Each person's contributions to the original conversation were coded as narration, facilitating remarks (mentoring), or monitoring (evaluating and correcting). The authors found a

1.

MEMORY FOR CONVERSATION

25

number of interesting relationships between conversational role and the content of Dean's statements in the Watergate hearings. Although there are many potential explanations of the results, the authors suggest that the observed patterns may reflect the influences of both schematic processing and conversational roles and goals (Hirst & Gluck, 1999). Perhaps the largest observed difference was in memory for Dean's own contributions versus those of others. Whereas Dean offered 58% of the narrations in the meeting, 71% of those offered in his statement to the committee were his own. The authors attributed this difference to the greater relevance of Dean's own contributions to his preexisting knowledge and beliefs regarding the Watergate events, along with personal motivations, such as need for self-importance. This generation effect has also been observed in laboratory studies (e.g., Miller et al., 1996). The authors also predicted that because Dean's own schematic organization would dictate both what he chose to tell the others and what he attempted to elicit from them, Dean would be more likely to later recall and report these schema-consistent narratives than those reflected in answers to others' less schema-relevant contributions. Indeed, Dean's memory for conversational contributions due to his own initiatives was superior to those resulting from the initiatives of others. The narratives elicited from Dean by Haldeman and Nixon were more likely to remain unstated than to appear in Dean's statement. On the other hand, the narratives elicited by Dean from Haldeman and Nixon were more likely to make it into the statement than remain unstated. One might have expected that power relationships would have an effect on memory, such that Dean would be more likely to remember something said by a powerful person than something said by a less powerful individual. In contrast, Dean's memory appeared to be affected by a combination of perceived expertise and authority. That is, he correctly believed he possessed the greatest expertise about the events that had happened since the Watergate break-in. However, Dean realized that Nixon had the authority to decide future actions concerning Watergate, and that Nixon's beliefs concerning how the Watergate situation would unfold in the future would affect those decisions. From this perspective, it is not surprising that Dean remembered his own narratives better than Nixon's, but that he instead remembered Nixon's prognostications and plans for the future better than his own. Generally, this is consistent with other findings indicating that conversations deemed "important" are remembered better (e.g., Miller et al., 1996). Dean's memory was also notably affected by "hindsight biases" (i.e., the tendency to remember the past as consistent with present outcome knowledge; see Davis, 1991; Davis, Kemmelmeier, & Follette, 2005; Davis & Loftus, this volume; Hawkins & Hastie, 1990; for in depth discussion), in that he falsely remembered having warned Nixon that the Watergate cover-up would be discovered and that this discovery would cost him the presidency (Neisser, 1981). A recent laboratory conducted by Davis and colleagues (Davis, Lopez, Koyama, et al., 2005) likewise found hindsight biases in memory for a conversation between a couple arguing over an interaction between the girl and a male classmate. Later some participants learned that the couple had broken up, whereas the others learned the female was found dead, and her boyfriend was charged with murder. Those who learned he was charged with murder were both more likely to remember accurately threatening statements he made to his girlfriend during the argument and more

26

DAVIS AND FRIEDMAN

likely to falsely remember that he had hit his girlfriend during the vignette and that he had also threatened her classmate. Robert Wyer and his colleagues (Wyer, Budesheim, & Lambert, 1990) conducted a systematic set of studies examining the impact of processing goals on memory for expectation-consistent versus inconsistent information. Subjects listened to a conversation between a man and woman about a third person after having first received handwritten descriptions of the target by both speakers. When subjects were told to form an impression of the target, memory was superior for information inconsistent with the female's handwritten initial description, whereas when subjects were told to infer each speaker's true impression of the target, memory was superior for information inconsistent with the initial description of the speaker who mentioned them; and when subjects were told to form impressions of the speaker, memory was superior for speakers' comments that were inconsistent with the other speaker's initial description. Presumably, information inconsistent with the listener's task-related working hypothesis drew the greatest attention and was hence most memorable. Acute personal needs can also selectively affect attention in social interaction. For example, recent rejection experiences (which presumably enhance social motivations) selectively enhance memory for explicitly social events recounted in others' diaries (e.g., Gardner, Pickett, & Brewer, 2000).

Attention to the Central Themes of Narrative Sequences. Attention to others' narratives is guided by the story structure and causal flow of the narrative. The listener tries to comprehend aspects of the narrative, such as coherence among actions, events, and outcomes; why episodes in the narrative occur and why the author explicitly mentions specific information; and motives of characters' actions, causes of events, and so on. Hence, attention is greater (and therefore memory as well) for aspects of the narrative related to these central themes and listener concerns (see Fletcher, Hummel, & Marsolek, 1990, for a review of enhanced memory for information relevant to causal themes; and Graesser, Olde, & Klettke, 2002; Wyer, 2004; Wyer, Adaval, & Colcombe, 2002, for general reviews of attentional processes in processing of narratives). Attention to Wording. Memory for specific utterances is affected by the function these utterances play in a conversation (Keenan, MacWhinney, & Mayhew, 1977; MacWhinney, Keenan, & Reinke, 1982). Specific wording is remembered better for information that conveys a great deal about the communicative situation as well as the speaker and his "intentions, beliefs, and his relations to the listener" (Keenan et al., 1977, p. 550). Concretely, this category includes, but is not limited to, statements of personal opinions, evaluations, jokes, profanity, and sexually suggestive language (see also Bates, Kintsch, Fletcher, & Giuliani, 1980; Bates, Masling, & Kintsch, 1978; Pezdek & Prull, 1993). Often such statements involve violations of listener expectancies and hence draw attention. Indeed, wordings that violate expectations tend to be remembered better (e.g., Gibbs, 1981; Kemper & Thissen, 1981), whereas wording conforming to cultural conventions or script-based expectations tends to be processed "mindlessly" or automatically (Langer, Blank, & Chanowitz, 1978) and not to be remembered (e.g., Kitayama & Burnstein, 1988).

1.

MEMORY FOR CONVERSATION

27

Attention to Misbehavior. Generally, violations of social norms or strongly held ex, pectations tend to recruit attentional resources, presumably in efforts to resolve the dis, crepancy between expectation and reality (e.g., Graesser, Gordon, & Sawyer, 1979). For example, one of the implicit expectations in many conversations is that participants will follow norms of politeness and modesty, that is, that they do not engage in unfavorable comments about each other, inappropriate sexual innuendo, or excessive sel£,praise. Statements violating norms of politeness and modesty tend to be more memorable than the remainder of the conversation (Pezdek & Prull, 1993; Wyer, Budesheim, Lambert, & Swan, 1994). Attention to Violations of Social Roles. Typically, lower,status persons are more polite to those of higher status (e.g., Lee, 1993). Consistent with the idea that social sta, tus induces politeness,related expectations, Kemper and Thissen (1981) found that im, polite requests by low,status speakers (not fitting our expectations) are more likely to be remembered, whereas such requests by high,status speakers are more likely to be forgot, ten or distorted. Conversely, polite requests by high,status speakers were more memo, rable than the same requests by low,status speakers (but see Holtgraves, 1997, for more ambiguous results). Likewise, when speakers state a request in a manner that is inappro, priate for the specific context, the wording of the request is more likely to be remem, bered than when the request was made in an appropriate way (Gibbs, 1987). Attention to Violation of Conversational Rules. Violations of conversational rules tend to draw attention, elicit efforts to understand the meaning of the violation, and generally tend to be interpreted in a manner consistent with the preservation of the "cooperative principle" (e.g., Grice, 1975) specifying that conversational contributions should be truthful, appropriately informative, relevant, and clear (see Hilton, 1995; Holt, graves, 2002, for reviews). For example, when the statement It sure is hot in here is re, peated for the fourth time (violating the principle of quantity [appropriate informative, ness]), the listener may infer the indirect meaning that the person is requesting a change in temperature setting. Whereas discrepancy between expectation and reality is generally attention grabbing, in other instances expectation,consistent information may be remembered better. For in, stance, Gamst (1982; Study 2) tested the notion that people have expectations with regard to the structure of a dialogue and have better memory for conversations that do conform to a generic script-perhaps because they are initially easier to process and understand. Attention to People. Likeable, attractive, or high,status persons are typically rela, tively novel, relatively attractive, and relatively important and hence tend to draw at, tention. Lower,status individuals tend to be highly vigilant toward the behavior and utterances of higher status individuals, in particular when their own outcomes are de, pendent on them (Berscheid, Graziano, Monson, & Dermer, 1976; Fiske & Depret, 1996). For similar reasons, attractive, likeable, and higher,status persons elicit more at, tention and more responsiveness from conversational partners who are interested in and eager to please and impress them (e.g., Davis, 1982, 1984), and generally those with whom

28

DAVIS AND FRIEDMAN

we have or expect important relationships elicit more attention (see Gardner, Pickett, & Brewer, 2000). Attention to one's partner may also be affected by the presence of other people, in that some evidence indicates that memory for conversation is reduced when the participants are aware of being observed (Lord, Saenz, & Godfrey, 1987).

Attentional Consequences of Anxiety and Emotion. Just as criminal violence begets intense emotions in victims and witnesses, so can verbal exchanges. Such exchanges may involve strong emotions generated by the exchange itself and may also take place in the context of strong emotions derived from other events. They might begin when participants are already unusually happy, excited, or distressed, or events external to the conversation may trigger intense emotions after it is under way. Moreover, specific elements or words of an utterance or conversations may be emotion provoking, even as the overall emotional tone remains stable. Hence it becomes of interest to understand how both good and bad emotions might affect memory for an exchange. Study of the role of emotions in memory for criminal events has been substantially restricted in comparison with the many questions that can be asked relevant to the role of emotions in memory for conversation. Although witness memory has commonly been studied in the context of negative emotions more relevant to the typical witness to crime, emotions of all sorts may play out in the many and varied conversations that are later reported in the context of criminal or civil suits. Although strong positive emotions may well produce similar "tunnel memory" for central at the expense of peripheral detail (e.g., Brown, 2003; Christianson, 1992a,b; Safer, Christianson, Autry, and Oesterlund, 1993; Christianson & Safer, 1996), memory researchers have yet to examine the impact of such emotions-or the impact of existing emotions on processing of new emotionconsistent or inconsistent information. It may be, for example, that intense existing positive emotions create motivational processes that suppress attention to negative information and vice versa. Attention and reactions to others in need of help, for example, have been shown to vary in response to anticipated effects on existing moods (e.g., Cialdini et al., 1987). It would not be surprising to see similar mood-regulation goals affect attention to conversational material. Unfortunately, there have been few studies of emotion and memory for conversation. One study found reduced memory for unpleasant relative to neutral statements by romantic partners (Sillars, Weisberg, Burggraf, & Zietlow, 1990; although only in relatively satisfied couples). Others have found poorer overall memory among participants reporting high levels of communication anxiety in getting-acquainted conversations (e.g., Stafford & Daly, 1984) or high levels of "psychological discomfort" during a conversation about conflicting attitudes (e.g., Singer, 1969). Some studies have shown that emotional arousal generated outside the conversation is linked to decreased memory for conversation (e.g., Goss, Neuliep, & O'Hair, 1985), just as violent TV programs tend to impair memory for interpolated advertisements (e.g., Bushman, & Bonacci, 2002; see also Mundorf, Drew, Zillmann, & Weaver, 1990). In contrast, one may retain a particularly strong though not necessarily accurate memory for emotional events within the conversation. Indeed, Mac Whinney, Keenan, and Reinke (1979) showed that conversational episodes causing physiological arousal

1.

MEMORY FOR CONVERSATION

29

were more likely to be remembered than those that were less arousing (see also Bolls, Lang, & Potter, 2001). Words involving feared stimuli are remembered better (e.g., spider; Kulas, Conger, & Smolin, 2003), and several studies of therapeutic interactions have shown that "important" therapeutic events are remembered better by clients and therapists (e.g., Cummings, Martin, Hallberg, & Siemon, 1992). Apparently, emotional arousal can sometimes promote long-term retention of the emotion-arousing event or message (Parkin, Lewinsohn, & Folkard, 1982). This may also explain why jokes and other emotion-eliciting utterances tend to be retained better than neutral utterances (Schmidt, 1994; cf. Keenan, MacWhinney, & Mayhew, 1977). However, memories for specific aspects of the event may be in error. As Brown & Kulik (1977) and others have shown in research on "flashbulb memories," although the core features of the events can be well retained, memories for peripheral detail (such as where or when it happened) are often quite mistaken. Notwithstanding the potential of emotion to facilitate memory, it may also be expected to impair memory in some such circumstances. Anxiety and other unpleasant emotions, for example, may be responsible for the well-known failure of patients toremember the content of their conversations with physicians (e.g., DiMatteo, 1985; Ley, 1966, 1979, 1986; Ley & Spelman, 1967). During aspects of physical examinations, for example, patients may invoke distracting strategies to avoid thinking about the unpleasant aspects of the examination, with the unintended consequence of reduced memory for information conveyed during the examination. Both physicians and patients are under special stress during discussions of serious or life-threatening diseases. However, physicians appear to have anticipatory anxiety that peaks early in interactions with patients (though it may persist for some time afterward), whereas patients' anxiety increases later in the consultation and even after an initial consultation about serious news has been concluded (Ptacek & Eberhardt, 1996). Physician anxiety may interfere with clear expression, attention to content, and noticing cues of confusion that may be presented by the patient as the conversation proceeds and the patient becomes more anxious. Attempts to suppress emotional expression (as may occur in a variety of high-stakes interactions) or to lie (also requiring more effort) have been shown to impair communication processes, for example by reducing responsiveness to one's partner, grammatical complexity, and verbal fluency and increasing rates of grammatical errors and speech disturbances (see Richards, 2004). Efforts to control emotions or behavior also disrupt attention in social situations (e.g., Bonanno, Papa, O'Neill, Westphal, & Coifman, 2004; Butler, Egloff, Wilhelm, Smith, Erickson, & Gross, 2003; Gilbert, Krull, & Pelham, 1988) and impair memory for statements encountered in emotion-evoking situations (e.g., Fraley, Gamer, & Shaver, 2000; Richards & Gross, 1999, 2000), although suppression through reconstruing negative events as less aversive does not appear to impair memory (Richards, Butler, & Gross, 2003). Furthermore, people who habitually attempt to suppress emotions in interaction report more memory problems (Richards, 2004). Although the existing studies have provided interesting beginnings for the study of emotion and conversational memory, the area might still be regarded as embryonic. The literature is not yet at a point where general conclusions can be drawn regarding whether

30

DAVIS AND FRIEDMAN

or when emotional content will enjoy an advantage in memory; the conditions under which positive and negative material will be remembered better; or how these effects may depend upon existing emotions, personality, context, or whether one is a participant versus observer. Finally, it is important to note that one's own emotions are distinct from those of interaction partners. There is evidence that expressions of negative affect from others are more accurately perceived and more likely remembered than neutral or positive expressions, perhaps because they draw greater attention; though in some cases memory seems to be superior for affective statements consistent with one's expectations of a partner (see Sillars, Weisberg, Burggraf, & Zietlow, 1990).

Order Effects. Though order has been studied primarily with respect to other domains of memory, persuasion, or person perception, the order in which particular statements occur is likely to affect attention and memory, perhaps more so in conversations of greater length. lgou and Bless (2003) suggested that order effects will depend upon where the listener expects the most important information to fall (and hence where most attention is allocated); and that these expectations will be based upon conversational maxims (e.g., Grice, 1975) specifying that contributions should be relevant and informative and avoid wordiness. Reasoning on these bases, the authors derived and supported predictions regarding conditions under which primacy versus recency effects might be expected (see also Krosnick, Li, & Lehman, 1990, for similar reasoning with respect to order effects on judgment). Attention at Encoding versus Retrieval. It is of particular importance for forensic contexts that there may be a significant discrepancy between witness concerns and focus of attention during a target conversation versus the nature of forensic questions and issues regarding that conversation. As discussed in later sections, for example, an interviewer may be entirely focused upon the information elicited from an interviewee and pay no attention to how it is elicited. But how is crucial to cases involving such issues as disputed child abuse allegations or coerced confessions. Unfortunately, information unattended at encoding tends to be lost and cannot be retrieved when its importance becomes clear in hindsight. In fact, communications researchers have investigated the issue of how questions that require reorganization in memory tend to degrade memory performance (e.g., Ericsson & Simon, 1984). Ericsson and Simon describe four levels of memory reports. Level 1 requires only that the information be reported in the same form in which it was experienced and stored, such that it requires no recoding. Level 2 requires that information experienced in one form (e.g., nonverbal behaviors) be reported in another (e.g., in words). Level3 requires reports of selected information such that respondents must scan and filter their memories to provide the requested report (e.g., How many times did X do Y?). And finally, Level 4 asks the respondent to report on aspects of the conversation that may not ordinarily be attended to, or that require inferences beyond what has been stored in memory-as is the case in many forensic contexts. Indeed, as the level of report escalates, so do inaccuracies in conversational memory performance (e.g., Benoit & Benoit, 1988a,b, 1990; Benoit, Benoit, & Wilkie, 1996).

1. MEMORY FOR CONVERSATION

31

DISTORTIONS DURING STORAGE

Distortions of conversational memory that take place during storage will no doubt be governed by many of the same principles and influences affecting recall of other events. We do not review the full range of these here. Instead, we offer a few examples of distortions specifically in memory for conversation. One of these, the influence of intervening conversations, is covered by Davis and Loftus (this volume). In this section, we cover primarily schematic influences. Several studies have demonstrated schematic effects with respect to verbatim memory (e.g., Bock & Brewer, 1974; Brewer & Hay, 1984; Holtgraves, Srull, & Socall, 1989; Wertsch, 1975). Brewer and Hay (1984), for example, showed that verbatim memory was biased toward the general style of the original remarks. That is, memory for remarks originally presented in the context of a specific style (e.g., academic versus business, etc.) tended to be distorted toward the style of the general context. Another contextual feature-speaker status-also appears to affect memory for verbatim structure, in that statements tend to be remembered as more assertive if the speaker is described as highrather than low-status (e.g., Holtgraves et al., 1989), but only if status is known when the remarks are heard, not iflearned after the fact (suggesting a constructive rather than reconstructive explanation of the effect).

The Special Case of Negation. Of potential importance in a variety of situations leading to litigation, negation appears to be more difficult to process and remember than affirmative statements (e.g., "Bob is not guilty" versus "Bob is innocent"). The schema plus tag interpretation of this effect suggests that the statement Bob is not guilty will evoke schematic information relevant to guilt, along with a tag denoting that the information is false. Unfortunately, the schema-consistent information and the negation tag can become dissociated in memory as time passes, leaving the perceiver with false memories that the original statement affirmed rather than negated the target information (i.e., "Bob is guilty." rather than "Bob is not guilty."), along with false memories of other associations consistent with the originally negated schema (e.g., "Bob has stolen things in the past, too." "Bob appears to be dangerous." See a recent explication and tests of this phenomenon by Mayo, Schul, & Burnstein, 2004). Unfortunately, such errors can prove disastrous. A malpractice case for which I (Davis) served as a consultant may well have been based on this exact process. The patient (for whom it was clearly counterindicated) swore that his doctor had told him to take a relatively large dose of aspirin, whereas the doctor was just as convinced that he had told the patient not to take aspirin. The patient suffered a GI bleed and nearly died, which he attributed to this ill-advised instruction.

FAILURES OF RETRIEVAL

Just as memory for other events may be biased by a host of influences at retrieval, so can memory for conversation. However, in light of the coverage of many of these issues in

32

DAVIS AND FRIEDMAN

other chapters of this handbook, we do not repeat this coverage. Furthermore, for excellent reviews of biasing influences of personal characteristics, goals, and acute and chronic motivations and expectations on both encoding and retrieval, the reader is referred to Pasupathi (2001) and Wyer (2004). We also do not cover accuracy of collaborative recall through conversation, a topic also reviewed in detail by Pasupathi (2001). Here we restrict our review of failures of retrieval to issues of source monitoring, many of which have been investigated specifically for conversation. The much greater variety of source-monitoring errors for conversation relative to other events is yet another significant difference between memory for conversation versus other events.

Who Said What to Whom, When, and Under What Circumstances? Source-Monitoring Errors at Retrieval Source memory refers to memory for the context in which a target object or item of information was encountered, whereas source misattribution (i.e., failure of source monitoring) occurs when a specific occurrence is misattributed to an incorrect time, place, person, or other context (see reviews in Davis & Follette, 2001; Johnson, Hashtroudi, & Lindsay, 1993; Klaver & Meiser, 2000; Koutstaal & Schacter, 2001; Meiser & Broder, 2002; Schacter, 2001). Source monitoring is more likely to be accurate when the original memory is highly elaborated and differentiated (or distinct) from other memories (see also Koutstaal & Schacter, 2001). Hence, source differentiation is encouraged by encoding conditions that encourage sufficiently focused attention and deep elaborative processing and impaired by those involving shallow processing, that makes it more difficult to distinguish information associated with one source from that associated with another (such as when the two sources are similar rather than different). Unfortunately, many if not most conversations take place under conditions that encourage shallow processing and where distinctiveness with other conversations might be low (such as a series of similar conversations between professionals and an endless stream of patients or clients). Hence the opportunities for source monitoring errors regarding conversations and their content are legion. Davis, Kemmelmeier, and Follette (2005) explored ten common categories of errors in conversational source monitoring, nine of which they argued are often at issue in legal contexts: (1) who said what; (2) to whom something is said; (3) whether something is actually said or only has been only considered, imagined, or planned; (4) in which conversation (of a number of possible conversations) a particular exchange took place; (5) when or where a particular exchange took place; (6) the order within a conversation or interaction in which a particular exchange took place, (7) what other participants or witnesses were present, if any; (8) other features of the context or previous utterances that would alter the meaning of the target utterance; (9) whether information was acquired from a particular conversational source or from some other medium; and (10) when a particular conversational contribution was planned, whether one has said these things to this person before. We refer the reader to Davis et al. (2005) for detailed case examples illustrating the involvement of the first nine of these in various civil and criminal cases. Here we review only relevant empirical memory research.

1.

MEMORY FOR CONVERSATION

33

Memory for Who Said What to Whom As indicated in our review of the adjudicative setting, it is often crucial for litigants to remember which of two or more possible speakers made a specific statement as well as the person to whom the statement was made. It may be important, for example, to know which member of a group made a threat to whom, or which person may have said something reflecting guilty knowledge only the perpetrator should have. Just as crucial would be accurate memory for whom the speaker threatened (the deceased or an uninvolved party). Memory researchers have addressed the issue of what determines memory for "who said what" perhaps more extensively than any other issue specific to conversation. Although we know of no experimental studies addressing the issue of the person to whom statements are made, theoretical principles guiding memory for the speaker should affect memory for the target in the same manner. Two general methods have been used to investigate speaker source memory (see the detailed review by Davis et al., 2005). The first, the popular Who said what? paradigm in social psychology (Fiske, Haslam, & Fiske, 1991; Taylor, Fiske, Etcoff, & Ruderman, 1978) was originally developed to study social categorization and prejudice. The general intent of this line of research was to demonstrate the tendency to think in social categories by showing that when asked to remember which of several speakers contributed a particular statement to a discussion, participants are more likely to misattribute the statements of one member of a social category to another person from the same category (i.e., those of the same gender, race, age, physical attractiveness, home town, clothing color, educational status, and many more) than to someone from a different category (see the review by Klauer & Wegener, 1998). For example, Justice Ruth Bader Ginsburg recognized this tendency to confuse members of the same social category in her complaint that attorneys arguing before the Supreme Court commonly confuse her with the other female justice, Sandra Day O'Connor (New York Times, October 5, 1997). Whereas the Who said what? paradigm was largely used to investigate observers' confusions between speakers, the cryptomnesia paradigm was largely used to investigate confusions between one's own contributions and those of others. Participants were asked to develop ideas in groups and later tried to remember which person originally offered which ideas. Hence, the tendency to mistakenly remember others' contributions as one's own was termed "unconscious plagiarism," or "cryptomnesia" (e.g., Brown & Murphy, 1989). Although deliberate instances of plagiarism are common (particularly in academic settings, for example), far more common are instances in which people in everyday life encounter ideas that later occur to them in a different context-with no memory of the original source (e.g., Marsh, Landau, & Hicks, 1997). George H. Daniels made a possibly spurious claim of just such a mishap in his published explanation of plagiarized passages in his book Science in American Society, explaining that he had an unusually good memory and must have unconsciously memorized passages from other sources that he later included in his book with (ironically) no memory of where they had come from (Daniels, 1972). Although the social categorization and unconscious plagiarism paradigms are somewhat different in focus, source confusions in the two areas tend to be governed by similar principles.

34

DAVIS AND FRIEDMAN

Determinants of Category-Based Source Confusions. Source misattributions in the Who said what? paradigm are influenced by personal and situational factors affecting (1) the depth with which characteristics of the speakers and their statements are originally processed, (2) the salience of social categories at encoding, and/or (3) use of heuristic versus controlled decision criteria at retrieval. Reflecting limits on processing capacity and depth of processing, for example, category-based source confusions are more common when someone is faced with either distraction or heavy processing demands at encoding than when that person is encouraged to use elaborative encoding strategies. They are also more common among older people (Bayen & Murnane, 1996; Brown, Jones, & Davis, 1995; Fitzgerald, 2000; Mather, Johnson, & De Leonardis, 1999; see Davis & Loftus, 2005; Davis, Kemmelmeier, & Follette, 2005, for reviews), who are generally more susceptible to schematic processing (see reviews by Davis & Loftus, 2005; Hess, 1999; Hess & Slaughter, 1990), and for whom limitations of cognitive resources render source confusions generally more common (see reviews in Davis & Loftus, 2005; Glisky, 2001; Koutstaal & Schacter, 2001). This age difference has been demonstrated in straight learning paradigms where the person is tested on which of two speakers presented each word or item of information, as well as on memory for who says what in conversation, and for which witness offers which testimony in a videotaped mock trial (see the review by Davis, Kemmelmeier, & Follette, 2005). Category-based source confusions are also more common among persons for whom the social category in question is salient, and who tend to think in stereotypical terms regarding the category, and in situations that encourage category-based thinking (such as when one's own category is in competition with the other). Furthermore, the self seems to function as a unique social category, in that confusions of one's own contributions with those of others are less likely than confusions between those of other persons (Raye & Johnson, 1980). Category-based source confusions are also encouraged by heuristic strategies at retrieval (such as guesses based on which statement would be most likely to fit a member of a specific category, rather than carefully controlled attempts to retrieve associations between source and content-see reviews by Brewer, Weber, & Carini, 1995; Davis, Kemmelmeier, & Follette, 2005; Gawronski, Ehrenberg, Banse, Zukova, & Klauer, 2003; Klauer & Wegener, 1998; Klauer, Wegener, & Ehrenberg, 2002). Finally, schema-based source errors are common. For example, the statement I work out almost every day was disproportionately misattributed to a person described as an athlete when it was actually made by a different speaker (i.e., a speaker-inconsistent statement was reattributed to render it speaker-consistent). Older participants made more such stereotype-consistent source attribution errors, and the extent to which they did so was related to scores on neurological tests assessing both frontal and medial-temporal area functioning of the brain (Mather, Johnson, & De Leonardis, 1999; see also Bayen, Nakamura, Dupvis & Yang 2000; Ferguson, Hashtroudi, &Johnson, 1992; and see Davis & Loftus, 2005, for further discussion of the relationship of brain functioning to source monitoring failures). Such schematic effects also appear to be stronger among those with poorer overall memory (Spaniol & Bayen, 2002) and those with stronger stereotyped associations (as measured by the Implicit Attitudes Test; Greenwald, McGhee, & Schwartz, 1998; Gawronski et al., 2003).

1.

MEMORY FOR CONVERSATION

35

Generally, schematic influences on memory tend to be greater under conditions that encourage shallow or heuristic processing, including when veridical memory is poorer, when processing demands are high in comparison with processing resources, or when processing capacity is restricted in some way-all conditions more likely to occur for older adults, who are more likely to engage in schematic processing of a wide variety of stimuli (see Davis & Loftus, 2005).

Determinants of Cryptomnesia. Many of the factors affecting category- or schemabased source confusions also affect cryptomnesia, which tends to become greater (1) for more difficult original generation tasks, (2) with increasing delay between the original group interaction and the later attempt to generate novel contributions, (3) for responses generated by the person responding immediately prior to the subject (presumably selffocused concern regarding one's own turn would be greatest at this point, and reduce "binding" of the source and content of the contribution), (4) when the original information came from a high- rather than a low-credibility source (and hence originally garnered more attention), (5) for contributions from a member of one's own sex (presumably a more similar and therefore more easily confused source), (6) when participants are distracted during the original generation of ideas, (7) when retrieval occurs in a context different from that of the original task, and (8) for older participants-and less under conditions encouraging deeper processing (e.g., Bink, Marsh, Hicks, & Howard, 1999; Brown & Halliday, 1991; Brown, Jones, & Davis, 1995; Macrae, Bodenhausen, & Calvini, 1999; Marsh & Bower, 1993; Marsh & Landau, 1995). Essentially, unconscious plagiarism tends to be most likely under the same conditions that promote other source memory errors, including category-based errors. Based on her studies of childrens' source-monitoring errors in collaborative activities, Foley and her colleagues (Foley & Ratner, 1998; Foley, Ratner, & House, 2002; Ratner, Foley, & Gimpert, 2002) concluded that despite its potential for generating social conflict, unconscious plagiarism may serve a useful learning function. She offered evidence that the processes of imagining or planning activities that are actually performed by others simultaneously induce false memories of performing the action oneself and facilitate learning of the activity. Memory for Medium. Closely related to the issue of who told us something is that of whether information came from a conversational source at all, versus from another medium-such as television, radio, printed material, or one's own imagination. Although this issue remains unaddressed with respect to conversational versus other sources, it is nevertheless sometimes involved in criminal or civil cases. For example, just as pre-trial publicity can contaminate the neutrality of prospective jurors, it can also contaminate the memories of witnesses (Loftus & Banaji, 1989; see Davis & Follette, 2001; Davis & Loftus, this volume for review), including those to conversations. Media and other sources can also convey information or misinformation attributed to professionals later sued for misinforming their clients (see Davis, Kemmelmeier, & Follette, 2005 for examples). These kinds of source confusions remain unaddressed by conversation memory researchers, but offer fruitful research issues for the future.

DAVIS AND FRIEDMAN

36

It should be noted, as well, that some models of language comprehension assume that comprehension of verbal narratives necessarily includes visual imagery of the event (e.g., Wyer, 2004). Hence, in some cases material may be "remembered" as coming from a conversation when it actually came from visual images generated by the conversation. For example, the language from Loftus and Palmer's (1974) classic misinformation study describing one car as having "smashed into" another may evoke visual imagery of speed, leading to a subsequent false memory of having been told the car was speeding, or of having seen broken glass.

"Target-Memory" Confusion: To Whom Was It Said? As noted earlier, we were also unable to locate research investigating what might be called "target memory confusion." However, this issue is particularly relevant in legal settings, where it most commonly becomes relevant with respect to delivery of instructions, warnings, or infor· mation crucial to an impending decision by a professional, such as a doctor or lawyer. Although such target memory confusion can result from a number of factors, probably the most common cause for legally relevant target memory failures is lack of distinctive· ness between the target conversations and countless similar conversations with similar patients or clients (see Wyer, 2004 for review of effects of similarity between events on event-specific memory). Further, whether or not the conversation took place, the pro· fessional often falls back on script-based knowledge of what he or she typically does in such a situation, which can sometimes be the basis of false belief or false memory that the particular client was given the information in the particular context in question. Although this problem of similarity-based target source errors is commonly discussed (although not in this terminology), we located no studies addressing factors that might enhance or prevent them. Future researchers might profitably adopt variations of the paradigms used to study speaker source confusions to address issues of target source confusion.

Reality Monitoring in Conversation Conversational reality monitoring can become an issue when the person falsely re· members either having said or having been told something that was in fact only planned or imagined. Several experiments have examined the conditions under which one can falsely remember having said something, when the true source of the "memory" lies elsewhere. For example, Buehler and Ross (1993) led participants to falsely remember having made statements by exposing them to pictures of the objects involved in the statements (thereby inducing a sense of familiarity)-an effect that increased with the passage of time. Hence, particularly as they became more unsure with the passage of time, participants appeared to become. more and more likely to use a sense offamiliarity to infer whether they had indeed made the target statement. Parks (1997) conducted a series of very clever studies designed to study false memories of having "said the unsaid". Subjects were led to think of a particular phrase, answer, or debate point-but were prevented from actually saying some (via either instruction or interruption), but allowed to say others. Participants often reported confidently that they had actually said things they only thought of. Older adults appear to have more difficulty remember·

1.

MEMORY FOR CONVERSATION

37

ing whether an item was merely thought about or spoken out loud (e.g., Hashtroudi, Johnson, & Crosniak, 1989)-although this has not yet been demonstrated in a conver~ sational context. Distraction or interruption-based failure to carry out conversational intentions are arguably frequent sources of memory failures giving rise to litigation. For example, doctors interrupt their patients within the first 18-22 seconds of their sessions (Beckman & Frankel, 1984; Marvel, Epstein, Flowers, & Beckman, 1999). As a result patients do not fully discuss their symptoms and complaints, and those they raise emerge late in their visits when time constraints become more acute. If these interruptions and time con~ straints prevent patients from actually raising all of their planned issues, they may later complain of a misdiagnosis, fully believing they had informed their doctors of all relevant symptoms (Liang et al., 2002). Similar problems with interruptions, distractions or time constraints may likewise prevent doctors from carrying out plans to state warnings or to give relevant instructions and treatments, thereby contributing to false memories that they had carried out these tasks, and denials of such failures. It should be noted that reality monitoring can also be an issue for the hearer. For ex~ ample, an erotomaniac may actively imagine sexual conversations between himself or herself and the target, which later are misremembered as real. This problem appeared in one case of sexual harassment, where the alleged victim falsely claimed that the accused had repeatedly professed his love and stated plans to leave his wife and move out of state with his accuser. Given the history of the accuser, it was clear both that she firmly be~ lieved what she was saying, and that it was all false. Finally, it is important to consider that like any form of witness memory (see Davis & Loftus, this volume for review of witness conformity effects), conversational memory in general, and conversational source monitoring judgments in particular may be subject to social influence. Loftus & Davis (2006), for example, have recently presented an analysis of priming and social influence effects on the criteria the person applies to de~ termine whether a "memory" represents a veridical event memory versus images or in~ ternal representations generated in some other way.

Memory for Where, When, and Which Conversation Studies of autobiographical memory have shown that memory for when something happened is far less accurate than memory for what happened (e.g., Larsen, Thompson, & Hansen, 1996; Wagenaar, 1986; 1996), and generally poorer for things in the rela~ tively distant past, and when there is nothing distinctive about the timing that associates the target event with others that can clearly be placed in time. Further, as indicated by research on flashbulb memories (e.g., Brown & Kulick, 1977), memory for where one was can also be quite inaccurate, even when memory for what happened is quite good. Little is yet known concerning source memory distinguishing between multiple conversa~ tions-although it is clear that similarity between conversations would impair distinc~ tiveness, and hence source memory. Nevertheless, as we have indicated near the beginning of this chapter, memory for where, when, or in which of many conversations a target exchange took place is often crucial.

DAVIS AND FRIEDMAN

38

Memory for Sequence As we discussed at the outset, the sequence of communications is frequently important to litigation. And yet memory for sequence is often poor (e.g., SchmitterEdgecombe & Simpson, 2001), whether for the sequence of actions within a single event or for the sequence of discrete events. For example, in a civil suit cited by Davis, Kemmelmeier, and Follette (2005), memory for order was crucial to a dispute regarding whether a client's final instruction to his broker had been to buy or not to buy a stock. This issue involved failures of source monitoring for the sequence of the discussion (in which minds may be repeatedly changed and changed back again), and for the association of a particular alternative with the final decision to choose it. Likewise, sequence can be crucial to many other legally relevant inferences. Imagine, for example, a conversation between an accused murderer and his alleged victim that included an extremely hostile and almost violent confrontation and a long period of jovial conversation and laughter. Arguably, the dispute would be a more likely motive for murder had it occurred at the end rather than the beginning of the conversation. Further, as we shortly consider in more detail, once criminal behavior has been alleged, the sequence in which information is introduced into interviews or interrogations of suspects and alleged victims can be crucial to determinations of whether their reports are veridical versus the result of suggestive or coercive influences. It should be noted that memory for order may be poor in part because information conveyed in conversation does not appear to be organized temporally in memory. Wyer et al. (1990), for example, used a measure of clustering in recall of information conveyed in conversations and found no temporal clustering in any of a series of studies.

Memory for Other Participants Memory for whether other persons witnessed or participated in a conversation can also be crucial, as noted earlier. In source monitoring terms, other people who either witness or participate in a conversation provide context for the focal event. In order for such contextual features to be successfully "bound" to memory of the focal event, they must receive adequate encoding and processing. Highly charged emotional conversations may lead to tunnel memory in which memory for conversational content itself is enhanced, but at the expense of more peripheral aspects of the context such as (particularly uninvolved) others (cf. Safer et al., 1993).

Memory for Contextual Information That Would Clarify Meaning and Implications Finally, we have seen that memory for contextual features of an interaction can be critical in determining its meaning or implications. The full meaning of a statement is revealed not only in the words themselves, but in the complete personal and situational context in which it occurred-including the immediate and historical context of each person and the dyad or group, and the external circumstances of the encounter.

1.

MEMORY FOR CONVERSATION

39

Although this is true in a wide variety of conversational contexts, we focus on two in particular that are of considerable forensic interest-both involving the issue of triggers of particular conversational contributions. Clearly, responses must be interpreted in light of what they are responses to. Gilbert and Jones (1986) coined the term "perceiver induced constraint" to refer to the way individuals constrain the statements and actions of their interlocutors-typically with little to no awareness of their own influence. Without awareness of such influences, it is not surprising that memory for them is often simply lacking or highly inaccurate. We illustrate with two areas of forensic application where the issue is essentially whether an interlocutor/interviewer/interrogator elicited true or false statements from the target: (1) conversations or interviews with children who are alleged victims of abuse; and (2) interrogations of criminal suspects. In each arena, those conversing with, interviewing, or interrogating the target are later asked to report the content of these interactions in court.

Hearsay Evidence in Cases of Child Abuse. We have referred to the hearsay exceptions adopted by many jurisdictions allowing second-hand reports of allegations of abuse made by children. Parents, police, treating physicians, psychologists, social workers or others may testify to the content of conversations in which the child alleged the occurrence of abuse. Apart from the possibility that, at least in some circumstances, such provisions violate the constitutional right of an accused to confront the witnesses against him, there has been great concern in both legal and psychological communities regarding the reliability of this form of evidence (see Ceci & Friedman, 2000; R.D. Miller, 2003; Warren, Nunez, Keeney, Buck, & Smith, 2002; the special issue of Psychology, Public Policy and Law, 1999, Vol. 5, #2 for multiple reviews). Since defendants in child abuse cases often contend that alleged victims were misled or coerced into making false allegations by suggestive conversations and interviews, it is crucial for fact finders to possess accurate knowledge of misleading and coercive influences, if any, in these interactions. Unfortunately-to highlight just one concern among many-almost nothing is known about whether either lay or professional adults can recognize, remember, and accurately report their own contributions to the interactions that may have misled or coerced the children. A literature has begun to emerge addressing the accuracy of this form of hearsay testimony as well as the nature of juror reactions to it. Generally, as is the case for other utterances, memory for the gist of childrens' reports seems to be quite good (e.g., Bruck, Ceci, & Francoeur, 1999; Warren & Woodall, 1999). However, memories of the details of how they were elicited tend to be less accurate. For example, Bruck et al. (1999) explored mothers' memories for interviews with their own preschool children. As expected, mothers' memories for meaning were superior to those for exact wording or structure. However, they had difficulty with two forms of source monitoring-(!) whether the child's statements had been spontaneous or prompted, and (2) whether particular utterances were made by the child or by the mother herself. Further, when shown altered transcripts of their interviews with the children, mothers often endorsed them as completely accurate, even when the experimenters had removed much of the suggestion and coercion that mothers had used when interviewing their children, or when the mothers' statements were altered so as to come from their children. These errors occurred

40

DAVIS AND FRIEDMAN

even though mothers had been forewarned to try to remember exact wording used in interviewing their children, and even though they were tested within 3 days-advantages atypical of real life hearsay witnesses, who tend to suffer much longer delays between interviews and later reports of them, and, in the case of professional interviewers, who experience substantial interference in memory from overwhelming case loads and countless similar intervening interviews and interactions (see Pence & Wilson, 1994). Indeed, Warren and Woodall (1999) obtained similar results with experienced professionals, who likewise remembered the gist of children's reportsfairly well, but failed to recall their own specific leading interview questions (about 20%) and the specifics of children's answers. Although not directly addressing memory for interviews, Brown, Jones, and Davis (1995) attempted to examine the full range of source judgments relevant to a more naturalistic conversation-including both which speaker asked a question, and which answered the question (either oneself or one of three other participants). The exchange was structured so that each person played each role (inquirer, respondent, listener) an equal number of times. Generally, source memory was best for the respondent roleprimarily due to greater success at self-identifications by responders. Strangely, with respect to the other active role-that of inquirer-source memory was poor, and was no better among the inquirers themselves than among the respondents or listeners. Similarly, responders were not better than listeners at identifying who the inquirer had been. After one week, memory for the inquirer role had dropped to chance. Finally, as in other arenas, source memory was poorer among older than younger participants. Apparently, people are not good at remembering who asked questions, but have better than chance memory for who answered them. Although not a direct test, these results support the notion that when interviewer questions suggest information the source of the suggestion may be remembered poorly. Studies of hearsay and question source memory, as well as the general literature on second-hand accounts (e.g., Bartlett, 1932; Allport & Postman, 1947; 1958) and perceiverinduced constraint (Gilbert &Jones, 1986), clearly indicate that interviewers are not sufficiently aware of their coercive influence, nor will their hearsay reports of these interviews reflect significant contextual detail-especially leading and coercive influencesrelevant to evaluation of the truthfulness of a child's story.

Coercion and Police Interrogations. Likewise, this literature calls into question the reliability of police accounts of individual interrogations. Almost immediately after confessing to alleged crimes, many suspects claim their confessions were false and induced by coercive police tactics (see reviews by Davis & O'Donohue, 2004; Davis & Leo, in press; Ofshe & Leo, 1997). Although there is a trend toward requiring that interrogations be recorded, most jurisdictions do not yet have or enforce this requirement. Hence, most interrogations are unrecorded, and fact finders evaluating claims of coercion must rely upon reported recollections of interrogators and suspects. Attempts to recall the coercive influences in police interrogations or other conversations and interviews will necessarily largely fail. The literature on recall of interviews with children clearly illustrates these failures empirically. However, these failures are just as clearly guaranteed by basic processes of attention, labeling, and schematic processing.

1.

MEMORY FOR CONVERSATION

41

It is well known, as reviewed earlier, that memory for verbatim content is poor under most circumstances. Further, one is likely to recall the gist of events in terms of concepts and schemas used to encode them. The legal questions that must be addressed when attempting to determine whether a report of child abuse or a confessed criminal behavior are true or coerced require witnesses to report in hindsight aspects of the conversation which they did not attend to or understand in the same terms as they unfoldeda task known to pose considerable difficulty (e.g., Higgins, McCann, & Fondacaro, 1982). As noted earlier, consistent with literature on "transfer-appropriate-processing" (e.g., Blaxton, 1989) or "encoding specificity" (i.e., the tendency for memory to improve when the same types of processes are used at encoding and retrieval, (e.g., Tulving & Thompson, 1973)), the greater the discrepancy between the way conversational information is encoded and the question posed at retrieval, the poorer memory performance becomes (e.g., Benoit & Benoit, 1988b; 1990; Benoit, Benoit, & Wilkie, 1996; Ericsson & Simon, 1984).

CONVERSATION AS BOTH MECHANISM AND TARGET OF MEMORY

A final distinction between conversation and other targets of memory is that it can be simultaneously a target and mechanism of memory retrieval. As such, conversations themselves are among the many mechanisms that can distort memory (see Davis & Loftus, this volume, for discussion of the processes through which conversations can distort event memory).

CONCLUSIONS

We conclude by reiterating our "call to arms" for memory researchers to join in the investigation of issues of memory specific to statements and conversations. Although basic memory research offers a rich source of hypotheses regarding the determinants of memory for conversation, it remains for future research to examine how principles governing memory for conversation converge (or not) with those governing memory for other events, and to explore the complexities governing processing of, and memory for, conversation that do not apply to other events. While speaking at a recent Continuing Legal Education (CLE) seminar, the first author asked for a show of hands from those who had ever had a case not involving any testimony regarding statements or contents of conversations. Not a single hand went up. But notwithstanding the pervasiveness of such testimony, researchers have largely neglected basic research in this area, and memory experts have rarely been asked to testify regarding determinants of accuracy in memory for conversation. We hope to see memory researchers amass a body of research to more fully inform the issues of conversational memory faced by litigators, and to see more testimony from memory experts to help attorneys provide context for jurors who must evaluate conversation-related testimony.

42

DAVIS AND FRIEDMAN

REFERENCES Allport, G. W., & Postman, L. (1947). The psychology of rumor. New York: Russell & Russell. Allport, G. W., & Postman, J. (1958). The basic psychology of rumor. In E. E. Maccoby, T. M. Newcomb, & E. L. Hartley (Eds.), Readings in social psychology (3rd ed., pp. 54-65). New York: Holt, Rinehart & Winston. Austin, J. L. (1962). How to do things with words. Oxford: Clarendon Press. Barr, D. J., & Keysar, B. (2002). Anchoring comprehension in linguistic precedents. Journal of Memory and Language, 46,391-418. Bartlett, F. C. (1932). Remembering. Cambridge: Cambridge University Press. Bates, E., Kintsch, W., Fletcher, C. R., & Giuliani, V. (1980). The role of pronominalization and ellipsis in texts: Some memory experiments. Journal of Experimental Psychology: Human Learning and Memory, 6, 671-691. Bates, E., Masling, M., & Kintsch, W. (1978). Recognition memory for aspects of dialogue. Journal of Experimental Psychology: Human Learning and Memory, 4, 187-197. Bayen, U. J., & Murnane, K. (1996). Aging and the use of perceptual and temporal information in source memory tasks. Psychology and Aging, 11, 293-303. Bayen, U. J., Nakamura, G. V., Dupuis, S. E., & Yang, C. L. (2000). The use of schematic knowledge about sources in source monitoring. Memory and Cognition, 28, 480-500. Beckman, H. B., & Frankel, R. M. (1984). The effect of physician behaviour on the collection of data. Annals of Internal Medicine, 101, 692-696. Benoit, P. ]., & Benoit, W. L. (1988a). Conversational memory employing cued and free recall. Central States Speech Journal, 39, 18-27. Benoit, W. L., & Benoit, P. J. (1988b). Factors influencing the accuracy of verbal reports of conversational behavior. Central States Speech Journal, 39, 218-232. Benoit, W. L., & Benoit, P. J. (1990). Memory for conversational behavior. Southern Communication Journal, 56, 24-34. Benoit, P. ]., & Benoit, W. L. (1994). Anticipated future interaction and conversation memory using participants and observers. Communication Quarterly, 42, 274-287. Benoit, W. L., Benoit, P. ]., & Wilkie, J. (1996). Participants' and observers' memory for conversational behavior. The Southern Communication Journal, 61, 139-155. Berscheid, E., Graziano, W., Monson, T., & Dermer, M. (1976). Outcome dependency: Attention, attribution, and attraction. Journal of Personality and Social Psychology, 34, 978-989. Bink, M. L., Marsh, R. L., Hicks, J. L., & Howard, J.D. (1999). The credibility of a source influences the rate of unconscious plagiarism. Memory, 7, 293-308. Blaxton, T. A. (1989). Investigating dissociations among memory measures: Support for a transferappropriate processing framework. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 977-989. Bock, J. K., & Brewer, W. F. (1974). Reconstructive recall in sentences with alternative surface structures. Journal of Experimental Psychology, 103, 837-843. Bonanno, G. A., Papa, A., O'Neill, K., Westphal, M., & Coi6nan, K. (2004). The importance of being flexible: The ability to enhance and suppress emotional expressions predicts long-term adjustment. Psychological Science, 15,482-487. Bolls, P. D., Lang, A., & Potter, R. F. (2001). The effects of message valence and listener arousal on attention, memory and facial muscular responses to radio advertisements. Communication Research, 28, 627-651. Bransford, J.D., &Johnson, M. K. (1972). Contextual prerequisites for understanding: Some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behavior, 11. Brewer, M. B., Weber,]. G., & Carini, B. (1995). Person memory in intergroup contexts: Categorization versus individuation. Journal of Personality and Social Psychology, 69, 29-40.

1.

MEMORY FOR CONVERSATION

43

Brewer, W. F., & Hay, A. (1984). Reconstructive recall of linguistic style.Journal of Verbal Learning and Verbal Behavior, 23, 237-249. Brown, A. S., & Halliday, H. E. (1991). Cryptomnesia and source memory difficulties. American Journal of Psychology, 104, 475-490. Brown, A. S., & Murphy, D. R. (1989). Cryptomnesia: Delineating inadvertent plagiarism. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 432-442. Brown, A. S., Jones, E. M., & Davis, T. L. (1995). Age differences in conversational memory. Psychology and Aging, 10, 111-122. Brown, J. M. (2003). Eyewitness memory for arousing events: Putting things into context. Applied Cognitive Psychology, 17, 93-106. Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, UK: Cambridge University Press. Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5, 73-99. Bruck, M., Ceci, S. J., & Francoeur, E. (1999). The accuracy of mothers' memories of conversations with their preschool children. Journal of Experimental Psychology: Applied, 5, 89-106. Buehler, R., & Ross, M. (1993). How do individuals remember their past statements? Journal of Personality and Social Psychology, 64, 538-551. Bushman, B., & Bonacci, A.M. (2002). Violence and sex impair memory for television ads. Journal of Applied Psychology, 87, 557-564. Butler, E. A., Egloff, B., Wilhelm, F. H., Smith, N.C., Erickson, E. A., & Gross, J. J. (2003). The social consequences of expressive suppression. Emotion, 3, 48-67. Campbell, W. K., & Sedikides, C. (1999). Self-threat magnifies the self-serving bias: A metaanalytic integration. Review of General Psychology, 3, 23-42. Carroll, J. M., & Russell, J. A. (1996). Do facial expressions signal specific emotions? Judging emotion from the face in context. Journal of Personality and Social Psychology, 70, 205218. Ceci, S. J., & Friedman, R. D. (2000). The suggestibility of children: scientific research and legal implications. Cornell Law Review, 86, 33. Cegala, D. J. (1984). Affective and cognitive manifestations of interaction involvement during unstructured and competitive interactions. Communication Monographs, 51, 3 20-338. Christianson, S.-A. (1992a). Emotional stress and eyewitness memory: A critical review. Psychological Bulletin 112, 284-309. Christianson, S.-A. (1992b). Remembering emotional events: Potential mechanism. In S.-A. Christianson (Ed.), The handbook of emotion and memory: Research and theory (pp. 307-340). Hillsdale, NJ: Lawrence Erlbaum Associates. Christianson, S.-A., & Safer, M.A. (1996). Emotional events and emotions in autobiographical memories. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 218-243). Cambridge: Cambridge University Press. Cialdini, R. B., Schaller, M., Houlihan, D., Arps, K., Fultz, J., & Beaman, A. L. (1987). Empathybased helping: Is it selflessly or selfishly motivated? Journal of Personality and Social Psychology, 52, 749-758. Clark, H. H. (1996). Using language. Cambridge: Cambridge University Press. Cook, A. E., & Myers, J. L. (2004). Processing discourse roles in scripted narratives: The influences of context and world knowledge. Journal of Memory and Language, 50, 268-288. Cummings, A. L., Martin, J., Hallberg, E., & Siemon, A. (1992). Memory for therapeutic events, session effectiveness, and working alliance in short-term counseling. Journal of Counseling Psychology, 39,306-312. Daniels, G. H. (1972). Acknowledgment. Science, 175, 124-125. Davis, D. (1982). Determinants of responsiveness in dyadic interaction. In W. Ickes & E. S.

44

DAVIS AND FRIEDMAN

Knowles (Eds.), Personality, roles, and social behavior (pp. 85-139). New York: SpringerVerlag. Davis, D. (1984). Antecedents and consequences of responsiveness in dyadic interaction. Paper presented at the International Communications Association, San Francisco. Davis, D. (1991). What would I think ifl didn't know the patient died? From The Mind's Eye, 1, 1-6. Davis, D., & Follette, W. C. (2001). Foibles of witness memory for traumatic/high profile events. Journal of Air Law and Commerce, 66, 1421-1549. Davis, D., & Follette, W. C. (2002). Rethinking probative value of evidence: Base rates, intuitive profiling and the postdiction of behavior. Law and Human Behavior, 26, 133-158. Davis, D., & Follette, W. C. (2003). Toward an empirical approach to evidentiary ruling. Law and Human Behavior, 27, 661-684. Davis, D., & Holtgraves, T. M. (1984). Perceptions of unresponsive others: Attribution, attraction, understandability, and memory for their utterances. Journal of Experimental Social Psychology, 20, 383-408. Davis, D., Kemmelmeier, M., & Follette, W. C. (2005). Conversational memory on trial. In Y. I. Noy & W. Karwowski (Eds.), Handbook of human factors in litigation (pp. 12-11 to 12-29). New York: CRC Press. Davis, D., & Leo, R. (in press). Strategies for prevention of false confessions. In M. Kebbell & G. Davies (Eds.), Practical psychology for forensic investigations and prosecutions. New York: John Wiley. Davis, D., & Loftus, E. F. (2005). Age and functioning in the legal system: Victims, witnesses and jurors. In Y.l. Noy & W. Karwowski (Eds.), Handbook of human factors in litigation (pp. 11-11 to 11-53). New York: CRC Press. Davis, D., Lopez, P., Koyama, M., White, B., Knaack, D., Kusal, T., et al. (2005). Memory for threats in conversation enhanced by later knowledge of violence between participants. Paper presented at the American Psychological Society, Los Angeles, CA. Davis, D., & O'Donohue, W. T. (2004). The road to perdition: "Extreme influence" tactics in the interrogation room. In E. L. W. T. O'Donohue (Ed.), Handbook offorensic psychology (pp. 897-996). New York: Elsevier Academic Press. Davis, D., & Perkowitz, W. T. (1979). Consequences of responsiveness in dyadic interaction: Effects of probability of response and proportion of content-related responses on interpersonal attraction. Journal of Personality and Social Psychology, 37, 534-550. DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129, 74-112. DiMatteo, M. R. (1985). Physician-patient communication: Promoting a positive health care setting. In J. C. Rosen & L. J. Solomon (Eds.), Prevention in health psychology (pp. 328-365). Hanover, NH: University Press of New England. Dunning, D., & Sherman, D. A. (1997). Stereotypes and tacit inference. Journal of Personality and Social Psychology, 73, 459-471. Dworkin, R. H., & Goldfinger, S. H. (1985). Processing bias: Individual differences in the cognition of situations. Journal of Personality, 53, 480-501. Ericsson, K. A., & Simon, H. A. (1984). Protocol analysis: Verbal reports as data. Cambridge, MA: MIT Press. Ferguson, S. A., Hashtroudi, S., & Johnson, M. K. (1992). Age differences in using sourcerelevant cues. Psychology and Aging, 7, 443-452. Fiedler, K., & Semin, G. R. (1988). On the causal information conveyed by different interpersonal verbs: The role of implicit sentence context. Social Cognition, 6, 21-39.

1.

MEMORY FOR CONVERSATION

45

Fiske, A. P., Haslam, N., & Fiske, S. T. (1991). Confusing one person with another: What errors reveal about the elementary forms of social relations. Journal of Personality and Social Psychology, 60, 656-674. Fiske, S. T., & Depret, E. (1996). Control, interdependence, and power: Understanding social cognition in its context. In W. S. M. Hewstone (Ed.), European Review of Social Psychology (Vol. 7, pp. 31-61). Chichester, UK: Wiley. Fitzgerald, J. M. (2000). Younger and older jurors: The influence of environmental supports on memory performance and decision making in complex trials. Journal of Gerontology: Psychological Sciences, 55B, 323-331. Fletcher, C. R. (1994). Levels of representation in memory for discourse. In M.A. Gernsbacher (Ed.), Handbook of psycholinguistics (pp. 589-607). San Diego, CA: Academic Press. Fletcher, C. R., Hummel, J. E., & Marsolek, C. J. (1990). Causality and the allocation of attention during comprehension. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 233-240. Foley, M. A., & Ratner, H. H. (1998). Children's recoding in memory for collaboration: A way of learning from others. Cognitive Development, 13, 91-108. Foley, M. A., Ratner, H. H., & House, A. T. (2002). Anticipation and source-monitoring errors: Children's memory for collaborative activities. Journal of Cognition and Development, 3, 385-414. Fraley, R. C., Garner, J.P., & Shaver, P.R. (2000). Adult attachment and the defensive regulation of attention and memory: Examining the role of preemptive and postemptive defensive processes. Journal of Personality and Social Psychology, 79, 816-826. Friedman, R. D. (1996). Prior statements of a witness: A nettlesome corner of the hearsay thicket. Supreme Court Review, 1995. Friedman, R. D. (2004). Adjusting to Crawford: High court decision restores confrontation clause protection. Criminal Justice, 19, 4-13. Fulero, S. M., & Finkel, N.J. (1991). Barring ultimate issue testimony: An "insane" rule? Law and Human Behavior, *15, 495-507. Fulton, A., & Bartlett, J. C. (1991). Young and old faces in young and old heads: The factor of age in face recognition. Psychology and Aging, 6, 623-630. Gamst, G. (1982). Memory for conversation: Toward a grammar of dyadic conversation. Discourse Processes, 5, 33-51. Gardner, W. L., Pickett, C. L., & Brewer, M. B. (2000). Social exclusion and selective memory: How the need to belong influence memory for social events. Personality and Social Psychology Bulletin, 26, 486-496. Gawronski, B., Ehrenberg, K., Banse, R., Zuhova, J., & Klauer, K. C. (2003). It's in the mind of the beholder: The impact of stereotypic associations on category-based and individuating impression formation. Journal of Experimental Social Psychology, 39, 16-30. Gibbs, R. W. (1981). Your wish is my command: Convention and context in interpreting indirect requests. Journal of Verbal Learning and Verbal Behavior, 20,431-444. Gibbs, R. (1987). Memory for requests in conversation revisited. American Journal of Psychology, 100, 179-191. Gilbert, D. T., & Jones, E. E. (1986). Perceiver-induced constraint: Interpretations of selfgenerated reality. Journal of Personality and Social Psychology, 50, 269-280. Gilbert, D. T., Krull, D. S., & Pelham, B. W. (1988). Of thoughts unspoken: Social inference and the self-regulation of behavior. Journal of Personality and Social Psychology, 55, 685-694. Gilovich, T. (1987). Second hand information and social judgment. Journal of Experimental Social Psychology, 23, 59-74.

46

DAVIS AND FRIEDMAN

Girgis, A., & Sanson-Fisher, R. W. (1995). Breaking bad news: Consensus guidelines for medical practitioners. Journal of Clinical Oncology, 13, 2449-2456. Glisky, E. L. (2001). Source memory, aging, and the frontal lobes. In M. Naveh-Benjamin, Moscovitch, M., & Roediger, H. L. (Eds.), Perspectives on human memory and aging: Essays in honour of Fergus Craik (pp. 265-276). New York: Psychology Press. Gold, G. (Ed.). (1974). The White House transcripts. New York: Bantam. Goss, B., Neuliep, J. W., & O'Hair, D. (1985). Reduced conversational memory as a function of negative arousal: A preliminary analysis. Communication Research Reports, 2, 202-205. Graesser, A. C., Gordon, S. E., & Sawyer, J. D. (1979). Recognition memory for typical and atypical actions in scripted activities: Tests of a script pointer plus tag hypothesis. Journal of Verbal Learning and Verbal Behavior, 18,319-322. Graesser, A. C., & Mandler, G. (1975). Recognition memory for the meaning and surface structure of sentences. Journal for Experimental Psychology: Human Learning and Memory, 104, 238-248. Graesser, A. C., Millis, K. K., & Zwaan, R. A. (1997). Discourse comprehension. Annual Review of Psychology, 48, 163-189. Graesser, A. C., Olde, B., & Klettke, B. (2002). How does the mind construct and represent stories? In M. C. Green, J. J. Strange, & T. C. Brock (Eds.), Narrative impact: Social and cognitive foundations (pp. 229-262). Mahwah, NJ: Lawrence Erlbaum Associates. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464-1480. Grice, H. P. (1975). Logic and conversation. In P. Cole &J. Morgan (Eds.), Sentax and semantics 3: Speech acts (pp. 41-58). New York: Academic Press. Grice, H. P. (1989). Studies in the way of words. Cambridge, MA: Harvard University Press. Harris, R. J., & Monaco, G. E. (1978). Psychology of pragmatic implication: Information processing between the lines. Journal of Experimental Psychology: General, 107, 1-22. Hashtroudi, S., Johnson, M. K., & Crosniak, L. D. (1989). Aging and source monitoring. Psychology and Aging, 4, 106-112. Hawkins, S. A., & Hastie, R. (1990). Hindsight: Biased judgments of past events after the outcomes are known. Psychological Bulletin, 107, 311-327. Hess, T. M. (1999). Cognitive and knowledge-based influences on social representations. In T. M. Hess & F. Blanchard-Fields (Eds.), Social cognition and aging (pp. 239-263). San Diego: Academic Press. Hess, T. M., & Slaughter, S. J. (1990). Schematic knowledge influences on memory for scene information in young and older adults. Developmental Psychology, 26, 855-865. Higgins, E. T., McCann, D., &Fondacaro, R. (1982). The "communication-game"' Goal-directed encoding and cognitive consequences. Social Cognition, 1, 21-37. Hilton, D. J. (1995). The social context of reasoning: Conversational inference and rational judgment. Psychological Bulletin, 118, 248-2 71. Hirst, W., & Gluck, D. (1999). Revisiting John Dean's memory. In E. Winograd & R. Fivush (Eds.), Ecological approaches to cognition: Essays in honor of Ulric Neisser. Emory Symposia in Cognition (pp. 253-281). Mahwah, NJ: Lawrence Erlbaum Associates. Holtgraves, T. M. (1994). Communication in context: The effects of speaker status on the comprehension of indirect requests. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20, 1205-1215. Holtgraves, T. M. (1997). Styles of language use: Individual and cultural variability in conversational indirectness. Journal of Personality and Social Psychology, 73, 624-63 7. Holtgraves, T. M. (1998). Interpreting indirect replies. Cognitive Psychology, 37, 1-27. Holtgraves, T. M. (1999). Comprehending indirect replies: When and how are their conveyed meanings activated? Journal of Memory and Language, 41, 519-540.

1.

MEMORY FOR CONVERSATION

47

Holtgraves, T. M. (2002). Language as social action. Mahwah, NJ: Lawrence Erlbaum Associates. Holtgraves, T. M., Srull, T. K., & Socall, D. (1989). Conversation memory: The effects of speaker status on memory for the assertiveness of conversation remarks. Journal of Personality and Social Psychology, 56, 149-160. Horton, W. S., & Keysar, B. (1996). When do speakers take into account common ground? Cognition, 59, 719-729. Igou, E. R., & Bless, H. (2003). Inferring the importance of arguments: Order effects and conversational rules. Journal of Experimental Social Psychology, 39, 91-99. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28. Kassin, S. (1997). The psychology of confession evidence. American Psychologist, 52, 221-233. Kassin, S., & McNall, K. (1991). Police interrogations and confessions: Communicating promises and threats by pragmatic implication. Law and Human Behavior, 15, 233-251. Kausler, D. H., & Hakami, M. K. (1983). Memory for topics of conversation: Adult age differences and intentionality. Experimental Aging Research, 9, 153-157. Keenan, J., MacWhinney, B., & Mayhew, D. (1977). Pragmatics in memory: A study of natural conversation. Journal of Verbal Learning and Verbal Behavior, 16. Kemper, S., & Thissen, D. (1981). Memory for the dimension of requests. Journal of Verbal Learning and Verbal Behavior, 20,552-563. Kent, G. G., Davis, J.D., & Shapiro, D. A. (1978). Resources required in the construction and reconstruction of conversation. Journal of Personality and Social Psychology, 36, 13-22. Kintsch, W., & Bates, E. (1977). Recognition memory for statements from a classroom lecture. Journal of Experimental Psychology: Learning, Memory and Cognition, 3, 150-159. Kitayama, S., & Burnstein, E. (1988). Automaticity in conversations: A reexamination of the mindlessness hypothesis. Journal of Personality and Social Psychology, 54, 219-224. Klauer, K. C., & Meiser, T. (2000). A source-monitoring analysis of illusory correlations. Personality and Social Psychology Bulletin, 26, 1074-1093. Klauer, K. C., & Wegener, I. (1998). Unraveling social categorization in the "Who said what?" paradigm. Journal of Personality and Social Psychology, 73, 1155-1178. Klauer, K. C., Wegener, I., & Ehrenberg, K. (2002). Perceiving minority members as individuals: The effects of relative group size in social categorization. European Journal of Social Psychology, 32, 223-245. Kong, A., Barnett, G. 0., Mosteller, F., & Youtz, C. (1986). How medical professionals evaluate expressions of probability. New England Journal of Medicine, 315, 740-744. Koutstaal, W., & Schacter, D. L. (2001). Memory distortion and aging. In M. Naveh-Benjamin, M. Moscovitch, & H. L. Roediger (Eds.), Perspectives on human memory and cognitive aging: Essays in honour of Fergus Craik (pp. 362-383). New York: Psychology Press. Krauss, R. M., & Chiu, C. Y. (1998). Language and social behavior. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 2, pp. 41-88). New York: McGraw-Hill. Krauss, R. M., & Fussell, S. R. (1996). Social psychological models of interpersonal communication. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 655-701). New York: Guilford. Krosnick, J. A., Li, G., & Lehman, D. R. (1990). Conversational conventions, order of information acquisition, and the effect of base rates and the individuating information on social judgment. Journal of Personality and Social Psychology, 59. Kulas, J. F., Conger, J. C., & Smolin, J. M. (2003). The effects of emotion on memory: An investigation of attentional bias. Journal of Anxiety Disorders, 17, 103-113. Langer, E. J., Blank, A., & Chanowitz, B. (1978). The mindlessness of ostensibly thoughtful action: The role of "placebic" information in interpersonal interaction. Journal of Personality and Social Psychology, 36, 635-642.

48

DAVIS AND FRIEDMAN

Larsen, S. F., Thompson, C. P., & Hansen, T. (1996). Time in autobiographical memory. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 129-156). Cambridge: Cambridge University Press. Lee, F. (1993). Being polite and keeping MUM: How bad news is communicated in organizational hierarchies. Journal of Applied Social Psychology, 23, 1124-1149. Lieberman, M.D., & Rosenthal, R. (2001). Why introverts can't always tell who likes them: Multitasking and nonverbal decoding. Journal of Personality and Social Psychology, 80, 294-310. Levinson, S.C. (2000). Presumptive meanings. Cambridge, MA: MIT Press. Ley, P. (1966). What the patient forgets. Medical Opinion Review, 1, 71-73. Ley, P. (1979). Memory for medical information. British Journal of Social and Clinical Psychology, 18, 245-256. Ley, P. (1986). Cognitive variables and noncompliance. Journal of Compliance in Health Care, 1, 171-188. Ley, P., & Spelman, M.S. (1967). Communicating with the patient. London: Staples. Liang, W., Burnett, C. B., Rowland, J. H., Merapol, N. ]., Eggert, L., & Yi-Ting, H. e. a. (2002). Communication between physicians and older women with localized breast cancer: Implications for treatment and patient satisfaction. Journal of Clinical Oncology, 20, 1008-1016. Lieberman, M.D., & Rosenthal, R. (2001). Why introverts can't always tell who likes them: Multitasking and nonverbal decoding. Journal of Personality and Social Psychology, 80, 294-310. Lobb, E. A., Buttow, P. N., Kenny, D. T., & Tattersall, M. H. (1999). Communicating prognosis in early breast cancer: Do women understand the language used? Medical Journal of Australia, 171, 290-294. Loftus, E. F., & Banaji, M. (1989). Memory modification and the role of the media. In V. A. Gheorghiu, P. Netter, H. J. Eysenck, & R. Rosenthal (Eds.), Suggestibility: Theory and research (pp. 279-294). Berlin: Springer-Verlag. Loftus, E. F., & Davis, D. (2006). Recovered memories. Annual Review of Clinical Psychology, 2, 469-498. Loftus, E. F., & Palmer,]. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589. Lord, C. G., Saenz, D. S., & Godfrey, D. K. (1987). Effects of perceived scrutiny on participant memory for social interactions. Journal of Experimental Social Psychology, 23, 498-517. Macrae, C. N., Bodenhausen, G. V., & Calvini, G. (1999). Contexts of cryptomnesia: May the source be with you. Social Cognition, 17, 273-297. MacWhinney, B., Keenan, J. M., & Reinke, P. (1982). The role of arousal in memory for conversation. Memory & Cognition, 10, 308-317. MacWhinney, B., Keenan, J. M., & Reinke, P. (1982). The role of arousal in memory for conversation. Memory & Cognition, 10,308-317. Marsh, R. L., & Bower, G. H. (1993). Eliciting cryptomnesia: Unconscious plagiarism in a puzzle task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 673-688. Marsh, R. L., & Landau, J.D. (1995). Item availability in cryptomnesia: Assessing its role in two paradigms of unconscious plagiarism. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 1568-1582. Marsh, R. L., Landau, J.D., & Hicks, J. L. (1997). Contributions of inadequate source monitoring to unconscious plagiarism during idea generation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23, 886-897. Marvel, M. K., Epstein, R. M., Flowers, K., & Beckman, H. B. (1999). Soliciting patient's agenda: Have we improved? Journal of the American Medical Association, 281, 283-287. Mather, M., Johnson, M. K., & De Leonardis, D. M. (1999). Stereotype reliance in source monitoring: Age differences and neuropsychological test correlates. Cognitive Neuropsychology, 16,437-458.

1.

MEMORY FOR CONVERSATION

49

Mayo, R., Schul, Y., & Burnstein, E. (2004). 'I am not guilty' vs 'I am innocent': Successful negation may depend on the schema used for its encoding. Journal of Experimental Social Psychology, 40, 433-449. Mazur, D. J., & Hickam, D. H. (1991). Patients' interpretations of probability terms. Journal of General Internal Medicine, 6, 237-240. Meiser, T., & Broder, A. (2002). Memory for multidimensional source information. Journal of Experimental Psychology: Learning, Memory, and Cognition, 28, 116-137. Mikulincer, M., & Shaver, P.R. (2003). The attachment behavioral system in adulthood: Activation, psychodynamics, and interpersonal processes. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 35, pp. 53-152). San Diego: Academic Press. Miller, J. B. (2001). Attachment models and memory for conversation. Journal of Social and Personal Relationships, I8, 404-422. Miller, J. B., & de Winstanley, P. A. (2002). The role of interpersonal competence in memory for conversation. Personality and Social Psychology Bulletin, 28, 78-89. Miller, J. B., deWinstanley, P. A., & Carey, P. (1996). Memory for conversation. Memory, 4, 615-631. Miller, L. M. S. (2003). The effects of age and domain knowledge on text processing. Journal of Gerontology: Psychological Sciences, 58B, 217-223. Miller, R. D. (2003). Testimony by proxy: The use of expert testimony to provide defendant testimony without cross-examination. The Journal of Psychiatry and the Law, 3I, 21-41. Mundorf, N., Drew, D., Zillmann, D., & Weaver, J. (1990). Effects of disturbing news on recall of subsequently presented news. Communication Research, I7, 601-615. Nakao, M.A., & Axelrod, S. (1983). Numbers are better than words. Verbal specifications of frequency have no place in medicine. American Journal of Medicine, 74, 1061-1065. Neisser, U. (1981). John Dean's memory: A case study. Cognition, 9, 1-22. Nussbaum, J. F., & Coupland, J. (Eds.). (2004). Handbook of communication and aging research. Mahwah, NJ: Erlbaum. Nussbaum, J. F., Pecchioni, L. L., Robinson, J.D., & Thompson, T. L. (2000). Communication and aging. Mahwah, NJ: Lawrence Erlbaum Associates. Ofshe, R. J., & Leo, R. A. (1997). The decision to confess falsely: Rational choice and irrational action. Denver University Law Review, 74, 979-1122. Ohnishi, M., Fukui, T., Matsui, K., Hira, K., Shinozuka, M., & Ezaki, H., et al. (2002). Interpretation of and preference for probability expressions among Japanese patients and physicians. Family Practice, I9, 7-11. Parkin, A. J., Lewinsohn, J., & Folkard, S. (1982). The influence of emotion on immediate and delayed retention: Levenger & Clark reconsidered. British Journal of Psychology, 73, 389-393. Parks, T. E. (1997). False memories of having said the unsaid: Some new demonstrations. Applied Cognitive Psychology, II, 485-494. Pasupathi, M. (200 1). The social construction of the personal past and its implications for adult development. Psychological Bulletin, I2 7, 651-6 72. Pence, D., & Wilson, C. (1994). Team investigation of child sexual abuse: The uneasy alliance. Thousand Oaks, CA: Sage. Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. (2003). Psychological aspects of natural language use: Our words, our selves. Annual Review of Psychology, 54, 547-577. Pezdek, K., & Prull, M. (1993). Fallacies in memory for conversations: Reflections on Clarence Thomas, Anita Hill, and the like. Applied Cognitive Psychology, 7, 299-310. Planalp, S., & Knie, K. (2002). Integrating verbal and nonverbal emotion(al) messages. InS. R. Fussell (Ed.), The verbal communication of emotions (pp. 55-78). Mahwah, NJ: Lawrence Erlbaum Associates. Ptacek, J. T., & Eberhardt, T. L. (1996). Breaking bad news: A review of the literature. Journal of the American Medical Association, 276, 496-502.

50

DAVIS AND FRIEDMAN

Ratner, H. H., Foley, M.A., & Gimpert, N. (2002). The role of collaborative planning in children's source-monitoring errors and learning. Journal of Experimental Child Psychology, 81, 44-73. Raye, C. L., & Johnson, M. K. (1980). Reality monitoring vs. discriminating between external sources of memories. Bulletin of the Psychonomic Society, 15, 405-408. Renner, B. (2004). Biased reasoning: Adaptive responses to health risk feedback. Personality and Social Psychology Bulletin, 30, 384-396. Rhodewalt, F., & Eddings, S. K. (2002). Narcissus reflects: Memory distortion in response to egorelevant feedback among high-and low-narcissistic men. Journal of Research in Personality, 36, 97-116. Richards, J. M. (2004). The cognitive consequences of concealing feelings. Current Directions in Psychological Science, 13, 131-134. Richards, J. M., Butler, E. A., & Gross, J. J. (2003). Emotion regulation in romantic couples: The cognitive consequences of concealing feelings. Journal of Social and Personal Relationships, 20, 599-620. Richards, J. M., & Gross, J. J. (1999). Composure at any cost? The cognitive consequences of emotion suppression. Personality and Social Psychology Bulletin, 25, 1033-1044. Richards, J. M., & Gross, J. J. (2000). Emotion regulation and memory: The cognitive costs of keeping one's cool. Journal of Personality and Social Psychology, 79, 410-424. Riggio, R. E., & Feldman, R. S. (Eds.). (2005). Applications of nonverbal communication. Mahwah, NJ: Lawrence Erlbaum Associates. Robertson, W. 0. (1983). Quantifying the meanings of words. Journal of the American Medical Association, 249, 1061-1065. Rudolph, W., & Forsterling, F. (1997). The psychological causality implicit in verbs: A review. Psychological Bulletin, 121, 192-218. Safer, M.A., Christianson, S.-A., Autry, M. W., & Oesterlund, K. (1993). Tunnel memory for traumatic events. Applied Cognitive Psychology, 12, 99-117. Schacter, D. L. (1999). The seven sins of memory: Insights from psychology and cognitive neuroscience. American Psychologist, 54, 182-203. Schacter, D. L. (2001). The seven sins of memory. New York: Houghton Mifflin. Schmidt, S. R. (1994). Effects of humor on sentence memory. Journal of Experimental Psychology: Learning, Memory and Cognition, 20, 953-967. Schmitter-Edgecombe, M., & Simpson, A. L. (2001). Effects of age and intentionality on content memory and temporal memory for performed activities. Aging, Neuropsychology, and Cognition, 8, 81-97. Schneider, B. A., Daneman, M., Murphy, D. R., & See, S. K. (2000). Listening to discourse in distracting settings: The effects of aging. Psychology and Aging, 15, 110-125. Schneider, B. A., Daneman, M., & Pichora-Fuller, M. K. (2002). Listening in aging adults: From discourse comprehension to psychoacoustics. Canadian Journal of Experimental Psychology, 56, 139-152. Schneider, B. A., & Pichora-Fuller, M. K. (2000). Implications of perceptual deterioration for cognitive aging research. In F. I. M. Craik & T. A. Salthouse (Eds.), The handbook of aging and cognition (pp. 155-219). Mahwah, NJ: Lawrence Erlbaum Associates. Schneider, C. E. (1998). The practice of autonomy: Patients, doctors, and medical decisions. Oxford: Oxford University Press. Searle, J. R. (1969). Speech acts. Cambridge, UK: Cambridge University Press. Sillars, A. L., Weisberg, J., Burggraf, C. S., & Zietlow, P. H. (1990). Communication and understanding revisited: Married couples' understanding and recall of conversations. Communication Research, 17, 500-522. Singer, S. (1969). Factors related to participants' memory of a conversation. Journal of Personality, 37, 93-110. Spaniol, J., & Bayen, U. J. (2002). When is schematic knowledge used in source monitoring? Journal of Experimental Psychology: Learning, Memory, and Cognition, 28, 631-651.

1.

MEMORY FOR CONVERSATION

51

Sperber, D., & Wilson, D. (1995). Relevance: Communication and cognition (2 ed.). Malden, MA: Blackwell Publishing. Stafford, L., Burggraf, C., & Sharkey, W. (1987). Conversational memory: The effects of time, recall mode, and memory expectancies on remembrances of natural conversations. Human Communications Research, 14. Stafford, L., & Daly, J. A. (1984). Conversational memory: The effects of recall mode and memory expectancies on remembrances of natural conversations. Human Communications Research, 10, 379-402. Stafford, L., Waldron, V. R., & Infield, L. L. (1989). Actor-observer differences in conversational memory. Human Communications Research, 15, 590-611. Stiles, W. B., Lyall, L. M., Knight, D.P., Ickes, W., Waung, M., Hall, C. L., et al. (1997). Gender differences in verbal presumptuousness and attentiveness. Personality and Social Psychology Bulletin, 23, 759-772. Street, R. L. (1993). Analyzing messages and their outcomes: Questionable assumptions, possible solutions. Southern Communication Journal, 58, 85-90. Strong, J. W. (Ed.). (1999). McCormick on evidence (5th ed.). St. Paul, MN: West Group. Taylor, S. E., Fiske, S. T., Etcoff, N. L., & Ruderman, A. J. (1978). Categorical and contextual bases of person memory and stereotyping. Journal of Personality and Social Psychology, 36, 778-793. Tulving, E., & Thompson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80, 352-373. Turnbull, W. (2003). Language in action: Psychological models of conversation. New York: Psychology Press. Vanous, S., & Davis, D. (2001). Motive evidence: Probative or just prejudicial? Paper presented at the Rocky Mountain Psychological Association, Reno, Nevada. Villaume, W. A., Brown, M. H., & Darling, R. (1994). Presbycusis, communication and older adults. In M. L. Hummert, J. M. Wiemann, &J. F. Nussbaum (Eds.), Interpersonal communication in older adulthood: Interdisciplinary theory and research (pp. 83-106). Thousand Oaks, CA: Sage. Wagenaar, W. A. (1986). My memory: A study of autobiographical memory over six years. Cognitive Psychology, 18, 225-252. Wagenaar, W. (1996). Autobiographical memory in court. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 180-196). Cambridge: Cambridge University Press. Warren, A. R., Nunez, N., Keeney, J. M., Buck, J. A., & Smith, B. (2002). The believability of children and their interviewers' hearsay testimony: When less is more. Journal of Applied Psychology, 87, 846-857. Warren, A. R., & Woodall, C. E. (1999). The reliability of hearsay testimony: How well do interviewers recall their interview with children? Psychology, Public Policy and Law, 5, 355-371. Watzlawick, P., Beavin, J. H., &Jackson, D. D. (1967). Pragmatics of human communication. New York: Norton. Wertsch, J. (1975). The influence of listener perception of the speaker on recognition memory. Journal of Psycholinguistic Research, 4, 89-98. Wigmore, J. H. (1974, revised by Chadbourn, J. H.). Evidence in trials at common law. Boston: Little, Brown. Wilson, C. L., Simpson, J. A., &Smith, S.M. (2005, January). Attachment and false memory: Linking working models and memory mistakes. Paper presented at the Society for Personality and Social Psychology, New Orleans, LA. Woloshin, K. K., Ruffin, M. T., IV, & Gorenflo, D. W. (1994). Patients' interpretation of qualitative probability statements. Archives of Family Medicine, 3, 961-966. Wyer, R. S., Jr. (2004). Social comprehension and judgment: The role of situation models, narratives, and implicit theories. Mahwah, NJ: Lawrence Erlbaum Associates.

52

DAVIS AND FRIEDMAN

Wyer, R. S., Jr., Adaval, R., & Colcombe, S. J. (2002). Narrative-based representations of social knowledge: Their construction and use in comprehension, memory and judgment. In M.P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 133-197). San Diego: Academic Press. Wyer, R. S., Jr., Budesheim, T. L., & Lambert, A. J. (1990). Cognitive representation of conversations about people. Journal of Personality and Social Psychology, 58, 218-238. Wyer, R. S., Budesheim, T. L., & Lambert, A. J. (1994). Person memory and judgment: Pragmatic influences on impressions formed in a social context. Journal of Personality and Social Psychology, 66, 254-267.

2 Interview Protocols to Improve Eyewitness Memory Ronald P. Fisher and Nadja Schreiber Florida International University

Despite the obvious appeal of television police shows, where attractive investigators and their white-smocked laboratory colleagues solve crimes by conducting high-tech computer-assisted analyses of paint samples, cloth fibers, burn patterns, and the like, in fact most crimes are solved by ordinary-looking police who ask victims and witnesses: "What happened?" One of the best predictors of whether police will close a criminal investigation and how the court case will be resolved is the quality of an eyewitness's report. Furthermore, eyewitness evidence is likely to play a key role for the foreseeable future (Bartol & Bartol, 2004). Indeed, several recent studies have found that police investigators agree that eyewitnesses usually provide central leads in their investigations (Berresheim & Weber, 2003; George & Clifford, 1992; Kebbell & Milne, 1998; Kebbell & Wagstaff, 1997). Unfortunately, however, police also note that witnesses rarely provide sufficient information (Kebbell & Milne, 1998). Ideally, witnesses would observe crimes under optimal viewing conditions, would have detailed and accurate recollections of the crimes, and would voluntarily and accurately communicate these recollections on request. Police cannot control these factors (estimator variables; Wells, 1978), and, as such, they represent more wishful thinking than effective police investigative technique. A more productive approach is to concentrate on those factors that police can control (system variables). Of those, the most important, and the focus of this chapter, is how they interview cooperative witnesses. We start by illustrating a typical police interview of a cooperative witness. As this reflects the training received in the police academy, we then describe the training provided to police in various western countries. We then examine some of the scientific research on interviewing witnesses, specifically with an eye toward developing better techniques. There is consensus among researchers about several interviewing techniques, and these have formed the core elements that appear in many interviewing protocols. We focus on one specific interviewing protocol, the Cognitive Interview (CI), 53

54

FISHER AND SCHREIBER

which incorporates these core elements along with several unique techniques to facilitate witness recollection. Much of the scientific research that tests specific interviewing protocols examines the Cl, and we review the most recent round of research. To complement this empirical review, we examine some important conceptual, methodological, and practical issues associated with the Cl and witness interviewing. Finally, we summarize some of the findings in the child witness interviewing literature and resulting interview guidelines.

CURRENT INTERVIEWING PRACTICES

Although police interviewing procedures are critical for eliciting extensive, accurate witness reports, we know relatively little about how police actually conduct interviews with cooperative witnesses. Approximately two decades ago, Fisher, Geiselman, and Raymond (1987) and George and Clifford (1992) described typical interviewing protocols of American and British police, respectively. The results were surprisingly uniform and discouraging. Following a perfunctory effort to establish rapport, police interviewers generally began interviews by making an open-ended request to the witness: Tell me what happened. After listening to an initial outburst of crime-related facts, often no more than just a few seconds, the interviewer interrupted the witness's narrative response and asked a barrage of direct, short-answer questions, on the order of: How tall was he? How much did he weigh? Did he have a weapon? These questions appear to have been directed by a list of generically salient crime facts and were often asked in the same order across witnesses. In addition to these neutral questions, police often asked leading or suggestive questions, for example, Was he wearing a red shirt? When witnesses did not provide the requested details, police interviewers did little or nothing to assist the witnesses' memory efforts. We note, with a sigh of renewed discouragement, that a very recent analysis of German police found a similar pattern of poor interviewing procedures (Berresheim & Weber, 2003). Here, following an initial open-ended request, police reverted to the conventional style of asking many more closed-ended and multiple-choice questions (mean = 28.9 per interview) than open-ended questions (5.7 per interview). As a result, only a small portion (7%) of the witness's testimony was given as a free narrative. Similar to the American (Fisher et al., 1987) and British (George & Clifford, 1992) samples, German investigators interrupted the witness frequently (on average five times per interview) and allowed for very few pauses. In preparing for the present chapter, we conducted yet one more analysis of American police interviews, almost 20 years after the earlier study (Fisher et al., 1987). This time, we enlisted the assistance of a local police department in south Florida. Experienced detectives (N = 23) from a variety of divisions (robbery, sexual battery, homicide, traffic homicide, and internal affairs) tape-recorded their interviews with victims and witnesses and submitted the tapes for analysis. These interviews represented typical investigative interviews as conducted in each division. Despite the two decades since our earlier study, and following a spate of scientific progress in the area, the current police interviewing behaviors were disappointingly similar to those we had observed earlier.

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

55

The interviewers did little to establish effective social dynamics: they made few attempts to build rapport, and only rarely did they explicitly instruct witnesses about their ex, pected role during the interview, such as telling witnesses to report everything in a de, tailed fashion. Committing many of the same errors as their predecessors, the current group of police officers frequently interrupted witnesses in the middle of the their narra, tive responses. Furthermore, police questioning tactics were poor. They asked mainly di, rect, short,answer questions, and only rarely did they ask open,ended questions. They often asked compound questions, combining several questions into one, and, to com, plete the package, they asked many suggestive or leading questions. We are disappointed to see that not much has changed about police interviewing procedures in the past 20 years, despite the extensive scientific research that has been conducted during that time. Obviously, there are limitations to the data we examined, as they were all collected from one police department, within one section of the United States. Based on our contacts with a limited number of other police departments around the United States, however, we believe that the interviewing procedures described here are likely to characterize many American police interviews. By comparison, we have noticed that some other, non,American police used more sophisticated interviewing techniques. We suspect that the relatively poor quality of American police interviews may reflect the training they receive-or do not receive. We therefore conducted an in, formal survey of various training academies for police and other investigative agencies in the United States and elsewhere. Within the United States and Canada, police training practices vary across states and provinces, so that investigators are subject to the individual training plan of their police department or of their state's or provincial police training commission. Training also varies across experience level, with new trainees receiving a broader array of police, related knowledge, but at the expense of depth of training. Many training programs in, elude sections on interviewing, however, much of that time is allocated to interrogating suspects, leaving insufficient time for interviewing cooperative witnesses. Within the segment on cooperative interviews, considerable time is spent on providing background information on terminology and legal issues, formal explanations of the purpose of in, terviewing witnesses, and the requirements for establishing witness competency and credibility. Based on the descriptions we received from various agencies, we believe that relatively little time is spent on explaining the scientific principles underlying witness memory and communication, or on conducting practice exercises of interviewing coop, erative witnesses. By and large, local police investigators in North America receive comparably ineffective training on interviewing cooperative witnesses. At the federal level, training facilities for interviewing cooperative witnesses are somewhat better. The FBI, for example, provides a 51,hour block of interview and in, terrogation training during their 17 ,week standardized training program for special agents. Training requirements distinguish between the challenges of interviewing coop, erative witnesses and interrogating suspects, with approximately half of the training block spent on each topic. The witness interview training follows the CI approach (Fisher & Geiselman, 1992) and includes further training on agency,specific requirements, such as generating reports. We have heard from instructors affiliated with the recently

56

FISHER AND SCHREIBER

formed Office of Homeland Security that they will provide concentrated training on interviewing cooperative witnesses, again following the CI approach, but this is not yet formally in place. In comparison with the varied, poorly documented, and relatively modest training provided to police in the United States and Canada, current police training practices in Britain are more uniform, better documented, more grounded in scientific research, and considerably more thorough. George (1991) noted that, until the early 1990s, any formal police training in interviewing was minimal, and informal on-the-job learning was the accepted way of gathering interviewing skills. Following several changes in public policy, there began to appear standardized training objectives and empirically based training models. Today there are national standards for interview training leading to a variety of formal training options. Police officers and civilian staff in Britain are given interview training, depending on their training needs. According to a national training plan, investigators can be trained on up to five levels ("tiers"). Tier 1 is the beginner level for interviewing witnesses and suspects for low-level crimes and is taught to all police officers. The training follows a simplified version of the CI for interviewing cooperative witnesses and a course on Conversation Management (George, 1991) for interviewing suspects. At Tier 2, a more elaborate version of the CI is presented for investigators, and in some constabularies for all police officers. At this level, specially trained instructors explain and conduct practice exercises in a six-day course. Tier 3 is intended for specialist interviewers, who may interview child witnesses, witnesses with special communication needs, serious sexual offense victims, and key witnesses. Here, interviewer training can take up to five weeks and follows either the Cl or the British Home Office's "Achieving Best Evidence" when there are specific challenges concerning the witness's cognitive ability (Home Office Communication Directorate, 2002). The suspect interrogation course at this tier typically lasts 12 days. Trainees must complete Tiers 3 and 4 successfully (passing a formal test) before being accepted to Tier 5, which represents an interview advisor role. The Tier 5 course is in the process of being written and will run approximately 8-10 days. A Tier 5 interviewer would have the highest level of interviewing responsibility, advising on murder or rape cases and/or interviewing police officers involved in these cases. This formalized tier structure seems to enhance training motivation, so that police officers appreciate the movement between tiers as a skills progression through their careers. Apparently, the British system has stimulated motivation among investigative interviewers, as the demands for training are so high that trainers do not have adequate time to conduct all of the requested programs (Clark, personal communication, February 3, 2004). Sweden and Norway do not provide as extensive or sophisticated training as Britain; however, they do provide relatively high-level national training for police. In Sweden, many investigators have been trained recently in the CL This training includes six hours of lectures on memory and perception and practical training exercises using videotaped materials. In addition, investigators may voluntarily participate in another police interviewing course that includes two hours of theory and one complete day filled with practical materiaL In Norway, investigators are trained in a 20-hour basic witness interview course. The course consists of modules on memory and witness psychology, principles of

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

57

the CI, and the dynamics of police interviews. It also includes practice exercises with feedback on conducting interviews with cooperative witnesses. In Germany, the structure of training parallels that in the United States, where training differs across states and between federal and state training academies. Federal investigators receive 18 hours of training on interviewing (only 1% of their total train, ing), but this time is divided between interviewing cooperative witnesses and interrogat, ing suspects and includes large segments on nonpsychological topics, such as legal and criminological aspects of interviewing, and testifying in court. Local investigators receive training from institutions within their respective states, and some of these training pro, grams are intense and sophisticated. For instance, the state of Nordrhein,Westfalen of, fers a 3 Yz,day training program ("structured witness interviewing") modeled after the CI. To summarize, training practices differ widely across countries. The most advanced and rigorous training is found in Britain, where police often collaborate with academic researchers to develop and evaluate training programs. Other European countries have also recently implemented research,based training programs, as have Australia, New Zealand, and Israel. We find it interesting that the United States and Canada, which have produced many of the leading theoretical advances in the psychology of memory, have not kept pace with their western counterparts by incorporating this knowledge into police training programs.

RESEARCH ON INTERVIEWING COOPERATIVE WITNESSES

Despite the obvious importance of interviewing within a police investigation, there has been relatively little research on the topic. There is considerably more scientific research on the related issues of conducting identification tests (lineups) and on detecting decep, tion in suspect interrogations (see in particular the chapters in Section 2 of volume 2 of this handbook). We believe there are several reasons for the limited research on inter, viewing. First, research on interviewing is resource,demanding, as it requires individual testing of participants, tape,recording of the interview, transcription of the taped inter, view, and then scoring of the transcripts, each of which is a painstaking operation. Sec, ond, police and other investigative interviewers often do not appreciate the subtleties of conducting effective interviews with cooperative witnesses, and so they do not perceive the need to develop better interviewing procedures. Third, police generally do not tape, record or videotape their interviews with cooperative witnesses. As a result, police have little hard evidence that they conduct interviews poorly. Fourth, the consequences of conducting interviews poorly-usually, collecting less information than might be col, lected, although in some instances, incorrect statements are also elicited-are typically less dramatic than the consequences of conducting identification tests poorly or failing to detect deception among guilty suspects, both of which may lead directly to guilty peo, ple being set free or innocent people being arrested. There are currently several different protocols on interviewing that have been developed throughout the world. Of these, the best known are the CI, Conversation

FISHER AND SCHREIBER

58

Management, the Memorandum of Good Practice, and the Step-wise method (Yuille, Hunter, Joffe, & Zaparniuk, 1993). Each of these protocols is composed of many specific techniques that have generally been found to (a) increase the amount of information gathered and/or (b) decrease the likelihood of a commission error. Common to all of these protocols are several core elements, including (a) developing rapport with the witness, (b) asking open-ended questions primarily, (c) asking neutral questions and avoiding leading or suggestive questions, and (d) funneling the interview, beginning with broader questions and narrowing the process down to more specific questions. These core elements are found not only in police interviews of eyewitnesses, but also in other domains of investigative interviewing, such as eliciting medical histories, journalistic interviews, oral histories, accident investigations, etc. (Fisher & Geiselman, 1992). That there are common elements across different areas of investigative interviewing should not come as a surprise, as the process of eliciting information from a respondent is similar, regardless of whether the content of the interview is a crime, a set of medical symptoms, a political event, or a set of autobiographical events.

The Cognitive Interview Although there are many interviewing protocols, almost all of the validity studies have examined only one protocol, the Cl. Certainly, other interviewing protocols have been used in experimental research. Their use, however, has been mainly in the service of examining other phenomena. For instance, the Step-wise method has been used in the service of testing prostitutes' recollections of sexual assaults (Cooper, Kennedy, Herve, & Yuille, 2002). It is not obvious to us why so many validation studies have targeted the CI and left other protocols untested-except as the comparison/control group to test the Cl. Nevertheless, in keeping with this bias, our review of the literature will focus on those studies that test the validity of the Cl. Does the CI elicit more or better information from respondents than a comparable or control type of interview? Before reviewing the empirical literature that tests the Cl, we first briefly describe its major components. For a complete description, see Fisher and Geiselman (1992). The core elements of the CI are organized in terms of the three basic psychological processes: memory/general cognition, social dynamics, and communication.

Memory/General Cognition Witnesses can do little to prepare themselves to encode crimes effectively. For cognitive theory to be useful in an eyewitness memory task, it must therefore focus on the retrieval phase of memory. Second, both the witness and the interviewer are engaged in demanding cognitive tasks. The witness is attempting to recall and describe in detail a complex, and perhaps threatening, event; the interviewer is listening for details and subtleties within the witness's response, generating and testing hypotheses about the crime, formulating questions, and noting the witness's answers. Because these tasks are demanding, the witness's and interviewer's cognitive resources must be used efficiently.

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

59

Context Reinstatement. Following from the encoding specificity principle, memory retrieval is most efficient when the context of the original event is reinstated at the time of recall (Tulving & Thomson, 1973). Therefore, at the beginning of the interview, interviewers should instruct witnesses to mentally recreate their affective, physiological, cognitive, and emotional states that existed at the time of the original event. Context reinstatement might also be implemented if the interviewer conducts the interview at the scene of the crime or accident. Limited Mental Resources. People have only limited mental resources to process information (Baddeley, 1986; Kahneman, 1973). Thus, performance suffers when we engage in several difficult tasks at once. Witnesses may conduct superficial searches through memory because they are concurrently listening to the interviewer's questions. Interviewers may fail to process subtle nuances within a witness's response because they are concurrently generating hypotheses about the crime and formulating questions. Interviewers can minimize overloading witnesses by refraining from asking questions while witnesses are searching through memory, and, in general, by asking fewer but more openended questions. Asking fewer questions should also increase the interviewer's efficiency by not deflecting his or her mental resources from the task of listening to the witness's description. Interviewers can also promote more efficient use of witnesses' limited mental resources by encouraging them to close their eyes when recalling (Bekerian & Dennett, 1997) and by minimizing physical distractions, such as phone calls during the interview.

Witness-Compatible Questioning. Each witness's mental representation of an event is unique. Some witnesses may have focused on the perpetrator's face, whereas others may have focused on the weapon. Similarly, some witnesses have more personal interest and/or perceptual skills about some aspects of the crime event (e.g., vehicles) than about others (clothing). Interviewers should therefore tailor their questions to the mental representation and perceptual skills of each particular witness instead of asking all witnesses the same set of questions. Interviewers often violate this rule by using a standardized checklist to guide their questioning (Fisher, Geiselman, & Raymond, 198 7). Within each witness, the accessibility of event details also varies systematically throughout the course of the interview. Memory for the perpetrator's face, for instance, should be more accessible when the witness is thinking about the perpetrator than when the witness is thinking about the getaway car. In general, event details will be most accessible when they are perceptually related to the witness's current mental image (Pecher, Zeelenberg, & Barsalou, 2003). Interviewers therefore should be sensitive to the witness's currently active mental image, so as to time their questions most efficiently. This may require interviewers to defer asking questions about investigatively relevant details until later in the interview, when the questions are compatible with the witness's current mental image. Varied Retrieval. Different retrieval cues may access different aspects of a complex event (Anderson & Pichert, 1978). Therefore, witnesses should be encouraged to think

FISHER AND SCHREIBER

60

about events in many different ways, for instance, in terms of their temporal or spatial properties, e.g., thinking about the event in different temporal orders or from different spatial perspectives (Gilbert & Fisher, in press). Similarly, events are represented conceptually and in terms of their sensory properties (Paivio, 1971). Witnesses should therefore be encouraged to describe an event in terms of its semantic properties and its various sensory properties. In general, when witnesses think about an event in more and different ways, they should be able to recall more details.

Metacognitive Monitoring. People can monitor the accuracy of their knowledge well and can maintain high accuracy if they are not actively encouraged to guess (e.g., Koriat & Goldsmith, 1996). To promote high accuracy in recall, interviewers should explicitly instruct witnesses not to guess but, preferably, to indicate that they "don't know." Furthermore, interviewers should refrain from applying social pressure on witnesses or otherwise encouraging them to answer questions they are uncertain of. This is particularly important when children are being interviewed. Minimizing Constructive Recall. At times, memory is a constructive process, whereby the rememberer incorporates information from other (nontarget) sources to construct the target episode (Bartlett, 1932; Bransford & Franks, 1971; Loftus & Palmer, 1974; see also Roediger & McDermott, 1995, for the proposition that all remembering is constructive). A witness, for instance, might incorporate knowledge gathered from other witnesses or the media to help construct his or her memory of a crime. Interviewers should therefore warn witnesses not to expose themselves to these sources. More important, witnesses may also acquire and use information from the interviewer, based on his or her verbal or nonverbal behavior (Ceci & Bruck, 1995). Interviewers should therefore be careful about not leaking information to witnesses either via their nonverbal behavior or by asking suggestive questions.

Social Dynamics Witnesses and victims often are asked to describe unpleasant experiences to people whom they do not know. Two critical features mark this social interaction. First, witnesses must be psychologically comfortable with the interviewer as a person to go through the mental effort and emotional distress of describing crime-related details. Second, there are social and knowledge-based inequalities between the witness and the interviewer. The interviewer (e.g., police officer) generally has higher social or expert status, which normally dictates that that the interviewer should control the interview; however, the witness has more first-hand knowledge of the crime event, which normally dictates that the witness should control the interview. Resolving this apparent conflict is critical for a successful interview.

Developing Rapport. Witnesses, and especially victims, are often asked to give detailed descriptions of intimate, personal experiences to police officers, who are complete strangers. If anything, the police investigator's official appearance (badge, uniform, gun)

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

61

may create a psychological barrier between the police officer and the witness. To overcome this natural barrier, police interviewers should invest time at the outset of the interview to develop meaningful, personal rapport with the witness (Collins, Lincoln, & Frank, 2002), a feature often absent in police interviews (Fisher, Geiselman, & Raymond, 1987).

Active Witness Participation.

The witness has extensive first-hand information about the target event. Therefore the witness, and not the interviewer, should be doing most of the mental activity during the interview. In practice, however, police interviewers often dominate the social interaction with the witness by asking many questions and by asking questions that elicit only brief answers. This relegates the witness to sitting passively, waiting for interviewers to ask questions (Fisher, Geiselman, & Raymond, 1987). To compound the problem, police interviewers often discourage witnesses from taking an active role by interrupting them in the middle of a narrative response. Interviewers can induce witnesses to take more active roles and to volunteer information by (a) explicitly requesting them to do so, (b) asking open-ended questions, (c) not interrupting witnesses during their narrative responses, and (d) constructing the social dynamic so that witnesses perceive themselves to be the experts and therefore the dominant person in the conversation. This last point is especially important when children are being interviewed.

Communication For police interviews to be effective, investigators must communicate their professional, investigative needs to the witness. Witnesses must also communicate their unique knowledge of the target event to the investigator. Ineffective communication will lead witnesses to withhold valuable information or to provide irrelevant, imprecise, or incorrect answers.

Promoting Extensive, Detailed Responses. Police interviews require witnesses to describe people, objects, and events in more detail than civilians normally do in casual conversation. Inducing such an extraordinary level of description mandates that police convey this goal explicitly, which they rarely do. To compound the problem, witnesses often withhold information because they do not know what is relevant for a police investigation. To minimize witnesses' withholding of information, interviewers should instruct witnesses to report everything they think about, whether it is trivial or out of chronological order, or even if it contradicts a statement made earlier. (This instruction does not imply that witnesses should lower their output threshold, and guess uncertain answers, as is sometimes misinterpreted in the literature. It merely directs witnesses to give expression to events when they think of them.) If contradictions arise within a witness's testimony, interviewers should defer resolving these contradictions until later in the interview. Police investigators generally direct witnesses to describe relevant information by asking many specific, short-answer questions about normatively relevant topics, such as the perpetrator's age, height, weight, clothing, weapon, etc. (Fisher, Geiselman, & Raymond, 1987). This questioning style minimizes irrelevant information, which is obviously desirable-especially when there are time constraints on the interview-but it es-

FISHER AND SCHREIBER

62

tablishes this efficiency at the cost of (a) minimizing unsolicited, but relevant, witness information; (b) reducing the depth and precision of witnesses' responses; and (c) sometimes inducing incorrect responses. Code-Compatible Output. Interviewers and respondents often exchange ideas only through the verbal medium. Some people, however, are more expressive nonverbally, and some events are better described nonverbally (Leibowitz & Guzy, 1990). Ideally the response format should be compatible with the witness's mental representation of the event, thereby minimizing the need to transform the mental representation into an overt response (cf. ideo-motor theory, Greenwald, 1970). For example, if an event is inherently spatial (e.g., location of objects within a room), then witnesses should respond spatially, for example, by drawing a sketch of the room or by placing model objects within a (model) room.

Sequence of the Cognitive Interview The CI follows a designated order intended to maximize the effectiveness of the individual techniques. The general strategy is to guide the witness to those memory codes that are richest in relevant information and to facilitate communication when these codes have been activated. The recommended questioning sequence, which is common to the Step-wise method (Yuille et al., 1993) and the Memorandum of Good Practice (Home Office, 1992), is to progress from asking open-ended questions to more specific follow-up probes. The CI is divided into five sections: Introduction. The introduction establishes the appropriate psychological states and interpersonal dynamics required to promote efficient memory and communication during the remainder of the interview. The interviewer attempts to (a) develop rapport with the witness; (b) convey his or her investigative needs for extensive, detailed information; and (c) encourage the witness to play an active role by volunteering information. Open-Ended Narration. The initial open-ended narration permits the interviewer to infer the witness's overall representation of the event and to develop an efficient strategy for probing the various memory codes. The interviewer notes the witness's "mental images" of the perpetrator and significant objects (e.g., weapons) and develops a preliminary plan about which images to probe, in what order, and which questions to ask when each image is evoked. If the initial narrative is not interrupted, witnesses will convey a strikingly large portion of their total knowledge of the critical event (e.g., Roberts & Higham, 2002). Probing. In the probing stage, the interviewer guides the witness to the richest sources of knowledge ("mental images") and thoroughly exhausts these sources of their contents. The interviewer might direct the witness to close his or her eyes and to think

2. INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

63

about the best view he or she has of the perpetrator, and then to describe that image in extensive detail. Follow-up questions related to this image are asked to elicit additional information. Other information that is not associated with a specific image is also gathered during the probing stage.

Review. Reviewing the information already recorded allows the interviewer to check on its accuracy. It also provides the witness with an additional opportunity to recall the event. During the review stage, the interviewer clarifies any uncertainties or discrepancies (e.g., contradictory claims) that occurred earlier in the interview. The interviewer should read his or her notes back to the witness at the end of the interview and ask the witness to (a) correct any errors or omissions in the interviewer's notes and (b) inform the interviewer of any new recollections that occurred during the review. The interviewer should also point out in a nonchallenging way any ambiguities or contradictory statements within the witness's statement and ask the witness to clarify these matters, even if that means indicating that the witness is not certain about the matter. Close. When closing the interview, the interviewer fulfills any official police requirements associated with the investigation, such as collecting background information about the witness. The interviewer should also offer suggestions to extend the functional life of the interview by encouraging the witness to contact him or her several days later when the witness thinks of new information. Although this is the normative sequence, the interview invariably deviates from this plan, as new, unexpected information and events arise during the interview. Interviewers must be flexible enough to adapt to unexpected conditions and not impose the normative plan when it is inappropriate. In that regard, the CI is more of a general guideline or a collection of tools than it is a recipe for conducting an interview. The successful interviewer, much like a skilled worker, needs to determine which tools to use for a particular job and then to use the tools properly.

Empirical Testing of the Cognitive Interview Much of the early research on the CI has been reviewed elsewhere (e.g., Bekerian & Dennett, 1993; Fisher, Brennan, & McCauley, 2002; Memon & Koehnken, 1992). Therefore, we will just summarize this early research and focus on the more recent literature. The earliest research studies examined a precursor of the current version, the so-called original version of the Cl, in which witnesses were instructed to (a) reinstate the context of the original event, (b) not edit their recollections before reporting, and (c) use varied retrieval, that is, change the order of reporting (chronological and reverse order) and the personal perspective (the witness's and another person's). The typical experimental procedure was for witnesses to observe either a videotape of a simulated crime or a live, innocuous event. Shortly thereafter, the witnesses were interviewed either with the CI or with the comparison interview. In almost all of these studies, the CI elicited more correct information than did the comparison interview, and the number of commission errors was approximately the same for the two interview types. The same

64

FISHER AND SCHREIBER

pattern held across events (live or videotaped; crime or innocuous), witnesses (college students, noncollege adults, and young children), and location of testing (United States, Britain, Germany). Research on the Cl underwent its first major shift with the development of the socalled revised or enhanced (current) version, as briefly described earlier in this chapter. The only experimental study that directly compared the revised and original versions of the Cl found the revised version to be far superior, eliciting almost 50% more correct information, and with approximately the same proportion of commission errors (Fisher, Geiselman, Raymond, Jurkevich, & Warhaftig, 1987). The revised and original versions of the Cl were also compared indirectly in a meta-analysis, as some of the component studies used the original version and other studies used the revised version (Koehnken, Milne, Memon, & Bull, 1999). The results of the meta-analysis suggest that the two versions of the Cl are not differentially effective. However, that conclusion is muddled by the confounding between versions of the Cl tested and other important methodological factors. Specifically, the original version of the Cl was usually compared with a typical police interview, which is infused with many interviewing errors, whereas the revised version of the Cl was usually compared with a superior control condition (the Structured Interview, which incorporates generally accepted principles of interviewing, minus those thought to be unique to the Cl). In our view, the direct comparison between the original and revised versions of the Cl (Fisher, Geiselman, Raymond, Jurkevich, & Warhaftig, 1987) is the more illuminating study, as it controls for other potentially confounding factors. We conclude, therefore, that the revised version is demonstrably superior to the original version. Apparently, the research and practitioner communities share this view as (a) most researchers who examine the Cl do, in fact, test the revised version, and (b) most agencies that provide training on the Cl generally present some variant of the revised version. As a final note, it is difficult to imagine that the revised version of the Cl would not be superior to the original version, given the evolution of the technique. That is, the revised version includes all of the components of the original version; in addition, it incorporates other generally accepted principles of cognitive and social psychology (e.g., limited mental resources) and makes more systematic use of known interviewing techniques (e.g., developing rapport, asking primarily open-ended questions). Given the demonstrated superiority of the revised version of the Cl (Fisher, Geiselman, Raymond, Jurkevich, & Warhaftig, 1987) one would expect all of the recent research to examine the revised version. To our dismay, many researchers continued to examine the original version of the Cl long after 1987, and inexplicably some researchers are still testing the original version. We strongly recommend to the research community that those who wish to examine the Cl concentrate their efforts on the revised version. Similarly, we strongly encourage police and other practitioners who actually conduct investigative interviews and those who provide training on investigative interviewing to use the revised version of the Cl. Recent studies of the revised Cl use basically the same experimental paradigm as those testing the original version: A controlled event is presented to young adults-typically college students-and then either a Cl or a control interview is conducted. Not surprisingly, the general trend of the current findings mimics those found with the original version, namely increased amount recalled, accompanied by similar accuracy rates.

2. INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

65

The more interesting studies are those that vary the standard procedure, either by using "atypical" witnesses, or by examining other kinds of eyewitness tasks, such as identifying the perpetrator in a lineup type of task. We turn now to these studies.

Nonstandard Witnesses.

With the increasing awareness of the importance of interviewing children and other "nonstandard" witnesses (e.g., older people, people with intellectual deficits) many studies in the late 1990s and at the turn of the millennium examined whether the CI could be used effectively with these populations. Several such studies have been conducted, mainly in England, and almost all conform to the basic experimental paradigm, comparing the (revised) CI with the Structured Interview. In some of these studies, the experimental witnesses are young children (from 7 to 12 years of age; e.g., Geiselman & Padilla, 1988; Larsson, Granhag, & Spjut, 2003; McCauley & Fisher, 1995; Milne & Bull, 1996; Saywitz, Geiselman, & Bornstein, 1992); in some studies, the witnesses are elderly (McMahon, 2000; Mello & Fisher, 1996); and in some, the witnesses are young adults with learning disabilities or some other cognitive deficit (Brown & Geiselman, 1990; Milne, Clare, & Bull, 1999). Rather than describe each study individually, we note the common pattern across studies, namely, the CI elicits considerably more correct information than does the comparison interview, and at approximately equivalent accuracy rates. There are a few exceptions to this pattern, where no differences were found between the CI and the comparison interview (e.g., McMahon, 2000; Memon, Holley, Milne, Koehnken, & Bull, 1994). Some of these nondifferences can be attributed to a lack of statistical power (e.g., McMahon, 2000); others may be explained by interviewers not changing their interviewing styles after CI training (Memon et al., 1994). Other than a few exceptions, however, the literature shows aremarkably uniform pattern of CI superiority. For instance, of the 55 experiments incorporated into the meta-analysis ofKoehnken et al. (1999), 53 of the experiments found that the CI elicited more information than did the comparison interview. Furthermore, the effect was sizable: the median increase in amount of information elicited was 34%. The second common finding relates to the accuracy of responses. Typically, accuracy is measured in two ways: accuracy rate (the proportion of responses that are correct) and absolute number of commission errors. The relation between the CI and comparison interviews depends on which measure of accuracy is used. (Later in this chapter we discuss the preferred measure of accuracy.) The CI and comparison interviews reliably yield very similar accuracy rates, with a slight though nonsignificant advantage to the Cl. When the absolute number of errors is used, the pattern is not as regular. Some studies show that the CI and comparison interviews elicit the same number of commission errors, and other studies show that the CI elicits more errors. The third common finding is that the effect of the CI is similar for all groups of witnesses: "normal" adults, young children, older adults, and adults with cognitive deficits. This holds for measures of amount recalled and for accuracy. That is, for all of these witness groups, the CI elicits more correct information than does the comparison interview, and the accuracy rates are approximately the same for both interview conditions. Differences across witnessing groups are mainly in terms of the amount of correctly recalled information, where, not surprisingly, young "normal" adults recall more information than

66

FISHER AND SCHREIBER

do the other groups. Of perhaps greater interest is that the accuracy rates do not vary much across groups. In fact, accuracy rates hardly vary across experiments, generally hovering between .75 and .85. That is, across a wide variety of experimenters, types of witness, methods of scoring, events to be recalled, and recall conditions, approximately 75% to 85% of all witness responses are accurate. We believe that accuracy rate is high because witnesses naturally monitor the quality of their recollections before volunteering a response, and that this editing mechanism eliminates low-accuracy responsesalthough sometimes at the cost of providing incomplete reports (Koriat & Goldsmith, 1996). Note that these universally high and consistent accuracy rates occur only for description tasks, as in a conventional police interview. Commission errors may be much higher and more variable in recognition tests (e.g., lineups).

Field Studies with Real Victims and Witnesses. Supplementing these controlled, laboratory studies of the Cl, there have also been two field studies with victims and witnesses of real crime. One of these studies was conducted in the United States (Fisher, Geiselman, & Amador, 1989), and one was conducted in England (George & Clifford, 1996). Both of these studies have been reviewed elsewhere, so we note here only that the two studies showed the same pattern as typically found in the laboratory: In Fisher et al., for instance, the CI elicited 50-60% more witness statements than did conventional police interviews. Also replicating the pattern found in laboratory studies, the accuracy rate, as estimated by corroboration of the primary witness's statements with those of other witnesses to the same crime, was slightly (nonsignificantly) higher for the CI-trained interviewers (.95) than for nontrained interviewers (.93). One of the advantages of collecting witness reports from real-world criminal cases is that we can analyze the witness statements for their utility in police investigations. Newlands, George, Towell, Kemp, and Clifford (1999) did just that, conducting a secondary analysis of the witness statements reported in the study by George and Clifford (1996) of real-world criminal cases. In that study, 28 British police officers were trained in the CI, Conversation Management (CM), or both CI and CM, or were untrained (although they were experienced in general interviewing techniques). They tape-recorded witness interviews before and after training. Other police officers then assessed the investigative value of these witness statements on three dimensions: practical use to an investigator, quality of the mental image of the perpetrator that could be formed from the interview, and an overall ranking of the quality of the interview. Newlands et al. reported that the post-training scores in the CI group were the (nonsignificantly) poorest of the four groups, in direct opposition to the pattern found in other studies. We suspect that this occurred because the four groups of interviewers were not equivalent prior to training: the pretraining scores were also the poorest for the CI group. With such unequal groups, the more appropriate measure would have been to score for improvement (from pre-training to post-training). When the results are analyzed in terms of improvement, they mirror those found in traditional laboratory studies, namely, the CI-trained interviewers improved the most from pre- to post-training-although note that the improvement score for the CI group was not significantly different from those of the other groups.

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

67

We suspect that the nonsignificant findings ofNewlands et al. (1999) reflect low statistical power: there was only one pre-training and one post-training tape-recorded interview per interviewer; there were only seven interviewers per training condition; and only a small subset of the data was scored (descriptions of the perpetrator). We are not critical ofNewlands et al. for the lack of power, as we recognize the immense amount of time and resources required to conduct such a study. If anything, we find their study supportive of the value of CI training, and we encourage other researchers (a) to examine real-world police interviews and (b) to score witness statements for measures that are relevant to police investigation. Ideally, if one had the resources, it would be valuable to examine whether the type of police interview affects the likelihood of solving crimes (closure rate). Naturally, conducting a study of that magnitude would take an extended period of time (for cases to be closed) and manpower (training enough police officers to overwhelm the variability in closure rate due to factors other than interviewing effectiveness, e.g., viewing conditions or retention interval). We suspect that such a study, although costly, would go a long way toward developing better police interviewing skills.

Person Identification Tests. As effective as the CI has been for eliciting better witness descriptions of events, it has been decidedly unsuccessful for improving performance in identification tests (e.g., photo lineups). Fisher, Quigley, Brock, Chin, and Cutler (1990) conducted four separate experiments and found no advantage of the CI over a no-instruction control group, even when the CI was modified specifically to facilitate facial processing (promoting interfeatural processing, minimizing verbal descriptions). The failure to find an advantage of the CI was not due to the CI being conducted improperly, inasmuch as it did elicit better descriptions of the target person than did the no-instruction control. Franzen and Sporer (1994, as cited in Sporer, 1996) reported a similar failure of the CI to facilitate face recognition. Gwyer and Clifford (1997) compared the CI with the Structured interview across several possible combinations of lineup type (target-present and target-absent; 48-hour and 96-hour retention intervals) and found that, although the CI elicited considerably better descriptions of the target person, it had only small and irregular effects in the recognition test. Specifically, it had no influence in any of the target-present conditions. In the target-absent condition, it did decrease the number of false identifications in the 48-hour delay test, but it had no effect in the 96-hour delay. Finger and Pezdek (1999) claimed to have compared the CI with a standard police interview; however, their study is not illuminating, as their version of the CI differed considerably from that described by Fisher and Geiselman (1992), and their standard interview did not resemble a conventional police interview. On balance, it appears that the CI has little effect on person identification tests. The failure of the CI to improve identification accuracy surprised us, especially given that Malpass and Devine (1981) observed an increase in identification accuracy when they implemented only the context reinstatement component of the CI (see also Krafka & Penrod, 1985). Upon further reflection, however, we note that there are many instances of experimental manipulations that have large effects in recall (description) tasks but little or no effect in recognition (identification) tasks (e.g., Bower, Clark, Lesgold,

68

FISHER AND SCHREIBER

& Winzenz, 1969). We suspect that this may be an instance of the outshining hypothe~ sis (Smith, 1988). That is, when a potent retrieval cue is provided, as in a recognition item is presented, the retrieval process is test where a copy of the to~be~remembered driven by the retrieval cue, which renders other factors less potent. When a nonspecific retrieval cue is provided, as in a description test where the interviewer asks, in essence, "What happened?" the minimal guidance of the retrieval cue opens the possibility that other factors will have greater influence. One implication for practitioners is that inter~ viewing skills should be more important when there are few other leads to prompt a wit~ ness's memory (e.g., photographs or other artifacts) than when there are many potential prompts. In the complete absence of other memory prompts, when there are no leads other than the witness's memory, the interviewer's skill to probe memory efficiently is of paramount importance. One interesting theoretical prediction for face recognition studies, based on the out~ shining hypothesis, is that experimental manipulations should generally have smaller conditions than in target~absent conditions. That is, when the effects in target~present test stimulus (person or face) is a good replica of the to-be-remembered stimulus (target~ present), the test stimulus will drive the retrieval process and render other manipula~ tions less potent. In contrast, when the test stimulus is not a good replica of the stimu~ lus (target-absent), the retrieval process is more easily influenced by other factors. In fact, a quick review of the person/face recognition literature shows this to be a com~ mon finding. For instance, instruction bias typically has a greater effect in target-absent conditions (see Steblay's meta~analysis, 1997), as does conditions than in target~present lineup format (sequential vs. simultaneous; see Steblay, Dysart, Fulero, & Lindsay's meta~ analysis, 2001) and location of the lineup administrator (potential to bias the witness's response; Haw & Fisher, 2004). Even the one study showing the CI to influence recog~ nition performance found that its effects were limited to reducing False Alarms in Target~ absent lineups (Gwyer & Clifford, 1997). In general, experimental researchers have not been successful at increasing Hits in Target~Present lineups by manipulating instruc~ tional variables. We suggest that a more fruitful line of research is to improve the nature driven process, for example, by selection of the photographs that are used in this data~ of photographs to match the witness's description rather than matching the suspect's photograph (Wells, Rydell, & Seelau, 1993), or perhaps to use videotapes rather than photographs (Cutler & Fisher, 1990).

Overcoming Misleading Suggestion Effects. We know that exposure to post~ event misinformation can distort a witness's memory (e.g., Loftus, Miller, & Burns, 1978). Police and other investigators within the legal system should therefore be trained to conduct interviews without introducing misleading information. The legal system, how~ ever, cannot control all of the potential contacts that may influence a witness's memory. For instance, although we can instruct witnesses to avoid contact with the media and not to discuss the incident with friends or other witnesses, we cannot practically prevent witnesses from interacting with others. In fact, it seems naive to assume that witnesses will not speak to one another, or that they will avoid exposure to the media. It seems even more naive to assume that children will not speak to their parents about a crime

2. INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

69

they have experienced, or that concerned parents will not speak to their children, sim~ ply because police investigators have warned them about the possibility of contaminat~ ing the child's memory. We should assume, therefore, that in the normal course of an in~ vestigation witnesses will be exposed to these other sources of misleading information. How effectively can interviewing procedures overcome the potentially tainting effects of these sources? Several experiments were conducted to examine whether the CI can overcome, or at least reduce, the deleterious effects of misleading information. Specifically, can the CI (a) overcome retroactively the injurious effects of misleading information that was presented at an earlier point in time and (b) inoculate a witness proactively against the injurious effects of misleading information that will be presented after the interview is conducted? Several such studies have been conducted, in which experimental witnesses (often children) were supplied with misleading information (e.g., Geiselman, Fisher, Cohen, Holland, & Surtes, 1986; Hayes & Delamothe, 1997; Milne, Bull, Koehnken, & Memon, 1995). Either a CI or control interview preceded or followed the misleading information. The common pattern across these studies is that the CI may be able to inoculate wit~ nesses against the effects of later misleading information. Once the witness is exposed to misleading information, however, its distorting effects cannot later be undone by the Cl. In practical terms, this suggests to us that it is unlikely that we can retroactively over~ come the distorting effects of a witness having been exposed earlier to misleading infor~ mation. The more likely prophylactic approach to minimizing the influence of sugges~ tive questioning is to conduct an effective witness interview shortly after the critical event, before the witness has been exposed to the influence of misleading suggestions. We recognize the futility of this recommendation in many instances, and especially in child~abuse cases, where police or other investigative agencies do not become aware of the victim's status until long after the victim's memory has been tainted by suggestive in~ fluences. If the witness has been exposed to suggestive information, police and other decision~makers should question the validity of the witness's recollection of the crime. In such instances, the value of corroborating evidence becomes increasingly important.

Practical Application of the Cognitive Interview: Police Interviewers' Feedback Laboratory and field studies have almost universally demonstrated that considerably more witness information can be elicited by the use of a CI than a conventional police interview protocol. What are the reactions of police officers and other investigators who investigations? Geiselman have actually attempted to implement the CI in real~world and Fisher (1997) reported several instances in which the Cl was used successfully to solve real~ world cases, ranging from a witness to a kidnapping, to a child molestation vic~ tim, to a witness of a politically motivated bombing. Just before this chapter was written, an investigator from the Bureau of Alcohol, Tobacco, and Firearms reported conducting CI with a 38~year~old woman who had witnessed a homicide as a an extended (3~day) 4~year~old child. The interview elicited innumerable recollections, many of which were

FISHER AND SCHREIBER

70

corroborated by police records established at the time of the crime. We have even received some encouraging feedback from the National Transportation Safety Board, whose investigators received training on the CI and shortly thereafter conducted extremely effective interviews of crew members of the USS Greenville, the American submarine that collided at sea with a Japanese fishing boat in February 2001. Of course, we recognize that these are merely anecdotes, which are subject to many biases. Offsetting these successes, British police reported that the complete CI was sometimes difficult to implement (Kebbell, Milne, & Wagstaff, 1999). Specifically, British police reported that they were uncomfortable using-and only rarely used-the Varied Retrieval and Imagery techniques. Furthermore, they noted that using the complete CI often required more time than was available. In response to this practical concern, Davis, McMahon, and Greenwood (2006) set out to create a shorter version of the CI that would still capture most of the information gathered by the complete Cl. Their laboratory-based research demonstrated that a pared-down version of the CI, deleting the Varied Retrieval technique, saved considerable time and yet was almost as effective as the complete Cl. We find this a healthy development in the CI research, trying to make it more efficient and sensitive to real-world conditions.

Methodological Issues Although the empirical results on the effectiveness of the CI have been remarkably consistent across studies, there have been some disagreements about methodological issues. Three such issues are (a) how to measure accuracy, (b) determining the appropriate control group, and (c) how to isolate specific components of the Cl. Measuring Accuracy. Accuracy has been reported either as the number of incorrect responses or the proportion of all responses that are incorrect (i.e., accuracy rate). Presumably, the argument for reporting the number of incorrect responses is that each individual error increases the chances of guiding the investigation astray. Accuracy rate, which takes into account not only the number of errors but also the total number of responses, is the more conventional measure (Koriat & Goldsmith, 1994). Note, for instance, other normalizing measures, such as automobile efficiency as miles per gallon, or baseball players' performance as hits per time at bat. We agree that every error increases the chances of reaching an incorrect conclusion, but we think that everyone-other than the perpetrator's attorney-would prefer police to interview a witness who provides 100 details, 2 of which are incorrect (accuracy rate = .98), than a witness who provides 2 details, of which only 1 is incorrect (accuracy rate = .50), even though the second witness makes fewer errors. We recommend that researchers follow the same intuitive logic and report the accuracy rates of their witnesses, and not only the number of incorrect responses. For a more thorough discussion of this argument, see the contrasting papers by Memon and Stevenage (1996) and Fisher (1996). There are interesting practical implications related to the distinction between the absolute number and the relative number of errors. On the one hand, a technique that generates many more correct recollections, even though at the cost of a few additional

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

71

incorrect recollections, creates many more investigative opportunities by opening up new leads, which may be corroborated by other sources. On the other hand, the few ad~ ditional errors of recalling events did not occur may lead to witnesses being challenged in court. We leave it to police investigators and legal strategists to debate the relative value of acquiring considerably more correct information at the cost of a few errors.

Control Group. Two experimental strategies have emerged, differing in terms of the control interview against which the CI is compared. Some researchers have used as the control group either a typical police interview or an interview fashioned after the Memorandum of Good Practice. Other researchers have used the Structured Interview, which is a reduced version of the Cl, that is, the complete CI minus those components thought to enhance memory retrieval. Each experimental strategy has some advantages and disadvantages. The first strategy simply indicates whether the CI is better than a specific comparison interview that is currently in use, but the results may not be gener~ alizable to other interview protocols. The results may also be misleading: If the compar~ ison interview is poor, then the CI is simply better than a poor technique. The advantage of comparing the CI with a known interview protocol is that the results are unambigu~ ous and have direct policy implications. Specifically, if the experimental results show the CI to be superior to an interview protocol that is currently in use, then the CI should replace the currently used interview protocol. The second strategy, eliminating the memory~enhancing components of the Cl, is used for more theoretical purposes, to determine whether the unique features of the CI actually promote better memory retrieval. This strategy suffers from the validity of the underlying assumptions: (a) that we know which components of the CI affect memory retrieval (i.e., which components to delete to create the comparison interview) and (b) that the deleted components affect memory retrieval only and not some other psy~ chological process (e.g., social interaction). How do we know which components of the CI affect memory retrieval? Perhaps instructing witnesses not to guess, which we listed in the category of memory retrieval, actually influences communication (converting the consciously recalled event into a communication act; Tulving, 1983). Second, how do we know that a component technique affects only memory retrieval and not also an~ other psychological process, such as social interaction? Perhaps requesting the witness to draw a sketch of the scene facilitates memory retrieval, but it also permits the inter~ viewer to ask better questions, because he or she has a better understanding of what the witness had observed. As far as we know, these questions have not been answered con~ vincingly enough to support the assumptions behind using the Structured Interview control. Furthermore, the policy implications of the Structured Interview control strat~ egy are limited, given that no investigative agency currently uses the Structured Inter~ view. In overview, although we are sympathetic to the theoretical goals of the Structured Interview approach, we find it to be standing on theoretical grounds that are too shaky to be of much value-at least at the current time. As such, we encourage researchers to compare the CI with one or more interview protocols that are currently in use. The other obvious recommendation is to shore up the theoretical grounds of the Structured Interview.

72

FISHER AND SCHREIBER

Isolating Specific Components of the Cognitive Interview. Most of the research that examines the Cl implements all of the component techniques as a package, and hence we cannot assess which of the various components are the active ingredients. Perhaps almost all of the benefit of the Cl occurs because of only a few components, with the remaining components merely serving as window dressing. Can we determine the effectiveness of specific components? If so, perhaps we can streamline the technique by using only those components that have been demonstrated to increase witness recall. This is particularly important if there is inadequate time to conduct the full Cl (Kebbell, Milne, & Wagstaff, 1999). Efforts have been made to examine the utility of specific components of the Cl (e.g., Boon &Noon, 1994; Geiselman, Fisher, MacKinnon, &Holland, 1986). Unfortunately, even as late as 2002, almost all of these efforts have been directed at the three components of the original version: context reinstatement, recall everything, and varied retrieval (Milne & Bull, 2002). The general strategy in these experiments was to instruct experimental witnesses to use one technique (e.g., Recall everything) or a combination of these techniques (e.g., Recall everything plus Context reinstatement) when recalling the target event. The scores from these conditions were compared with either a Control interview (Boon & Noon, 1994; Geiselman et al., 1986) or with a request for witnesses to "Try again" (Milne & Bull, 2002). Collectively, these studies found only minimal support for using any one technique in isolation, with greater effects being observed when all of the techniques were used in combination. It is difficult to interpret many of these results, however, because, in some of the studies, the techniques were implemented only after the witnesses had already provided an initial free narrative (Boon & Noon, 1994; Milne & Bull, 2002). Furthermore, showing that a technique does not generate more information than an instruction to "Try again" does not necessarily imply that the technique does not generate additional information. It simply is no more effective than instructing witnesses to try again. Finally, the effort is limited in that only the three mnemonics of the original Cl were examined. By our conservative estimate, these three components account for less than 10% of the revised (current) version of the Cl. It is not apparent to us why, in 2005, 18 years after the revised Cl was first demonstrated to be superior to the original version, researchers are still limiting their research to dissection of the original version. Although we are sympathetic to the goal of isolating the components of the Cl, we suspect that the task is considerably more difficult than appears on the surface. The major difficulty is that some components are designed to be implemented in concert with others, not in isolation. For instance, instructing witnesses to give complete answers and to actively generate information should be effective, but only when combined with open-ended questions. Similarly, instructing witnesses to close their eyes (to minimize distractions) can be implemented properly only after rapport is established. Moreover, the contribution made by any one component probably varies across testing conditions. Reinstating context, for example, is likely to be more potent at long retention intervals, after the witness's thoughts have changed over time, than at short retention intervals (Fisher, Falkner, Trevisan, & McCauley, 2000). Certainly, we encourage others to develop effective methods to test specific components of the Cl, but we doubt that many hard-and-fast rules will emerge from this endeavor. The more likely finding, we expect,

2. INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

73

is that all of the components will be somewhat effective under specific conditions. In that regard, the skill of conducting an effective interview is to decide which of the vari~ ous components is most likely to be helpful under the conditions of the present investi~ gation. Making such a decision on~line, under time pressure, and in the context of an important investigation with many possible leads, is likely to require as much training as implementing the component techniques themselves.

INTERVIEWING CHILD WITNESSES

We conclude this chapter by focusing specifically on interviewing children, as this has abuse cases, become an area of intense research since the highly publicized day~care starting in the early 1980s (Nathan & Snedecker, 1995). In these cases, preschoolers ac~ cused their day~care teachers of having sexually abused several teachers and children, with some accusations involving ritualistic elements. Not surprisingly, the validity of these allegations was examined in the ensuing criminal investigations, where expert witnesses identified and criticized several questionable interviewing techniques. We highlight here some of the targeted interviewing techniques that may have encouraged the children to make false allegations (Ceci & Bruck, 1995; see also the chapters in Part III of this volume.) One of the problematic techniques, selective reinforcement, can take several forms: (a) praising or otherwise rewarding a child (You are such a smart boy!) for reporting the desired answer, (b) giving a child negative feedback (Maybe you don't have a very good memory.) for failing to provide the desired answer, or (c) indicating that praise, rewards, or negative consequences will follow, depending on what the child says (I am sure you are smart enough to remember.). Most variations of conventional learning theory would predict that positive reinforcement should increase, and punishment (negative feed~ back) should decrease, the probability of eliciting a desired response. In line with these predictions, Garven, Wood, and Malpass (2000) showed that children who were rein~ forced were more likely to agree with misleading questions than children who were not reinforced. abuse cases is peer A second problematic technique that was identified in day~care information. This refers to interviewers informing child witnesses pressure or co~witness about other witnesses' alleged statements. The concern is that this technique can po~ tentially create conformity pressure on the child to agree with other witnesses' state~ ments (Garven et al., 1998, 2000; Leichtman & Ceci, 1995). A third technique that may promote incorrect testimony, inviting speculation, may occur by asking a child to speculate about the events in question or to pretend that an event has happened. Introducing this new "reality level" into the interview is known to have a detrimental effect on the accuracy of a child's later statement (Ceci, Huff~ man, Smith, & Loftus, 1994; Ceci & Loftus, 1994; Schreiber et al., 2001, Schreiber & Parker, 2004). and other abuse cases Several other interviewing techniques identified in day~care have also been shown to encourage children to make false statements. Three such tech~ niques are interviewer bias (the investigator has a preconceived notion about the event

74

FISHER AND SCHREIBER

or the suspect in question), repeated questioning (asking a question that the witness has already answered), and introducing a negative stereotype of the accused (Ceci, Leichtman, & White, 1995; Garven et al., 2000; Leichtman & Ceci, 1995; Lepore & Sesco, 1994; Poole & White, 1991). Given that each of these techniques, in isolation, has been shown to influence children's statements, it is not surprising that combining these suggestive techniques, the more likely scenario in a real-world investigation, can have a particularly detrimental effect on children's testimony (Garven, Wood, Malpass, & Shaw, 1998; Garven et al., 2000). Based on the empirical research findings, several guidelines have emerged to promote effective child interviewing procedures (see Ceci & Bruck, 1995, and Poole & Lamb, 1998, for comprehensive reviews). Many of the suggested techniques are common to those recommended for interviewing adults, including establishing the desired social dynamics (that the interviewer has little prior knowledge about the critical incident), developing rapport, asking primarily open-ended questions, avoiding complex and suggestive/leading questions, and encouraging the witness to say "I don't know" rather than to guess. The sequential structure of the interview also mirrors that recommended for adults, progressing from broader, open-ended questions and funneling toward more specific follow-up questions. This design is best exemplified in the Step-Wise procedure (Yuille et al., 1993) and the "Memorandum of Good Practice" (Home Office, 1992). Special considerations have been recommended for interviewing children to accommodate their reduced cognitive skills and background knowledge. For instance, children are often found to provide relatively impoverished narratives in response to open-ended questions (Poole & Lamb, 1998). To promote more elaborate narratives, and in general to assess children's verbal skills, interviewers are encouraged to make more effective use of the rapport-building phase. Here interviewers may ask children to describe innocuous, specific events in their lives (e.g., a birthday party) that are unrelated to the focus of the investigation. The resulting narratives provide the interviewer with baseline data on the child's verbal skills and should reinforce and train the child to give informationrich narratives (Yuille et al., 1993). Another method, Narrative Elaboration, provides children with cues to describe the various elements of a narrative, such as participants, settings, actions, and conversations (e.g., Camparo, Wagner, &Saywitz, 2001; Saywitz & Geiselman, 1998). These cues can be presented by showing pictures or providing verbal prompts (Brown & Pipe, 2003). Both of these approaches can be implemented early in the interview, before the critical event is addressed, to train children to provide more extensive narratives. One tempting method of overcoming children's reduced skills and knowledge has been to introduce aids such as drawings or anatomically correct dolls. Many researchers, however, caution against using these aids, because of their potential to distort the child's recollection (Ceci & Bruck, 1995). The safer and generally recommended approach is to allow for such interviewing aids only after the child has already provided a verbal description of the event (Poole & Lamb, 1998). Some unique concerns arise when children are interviewed, and especially when the alleged event is physical or sexual abuse. One of the most challenging problems is to introduce the topic of abuse without the use of suggestive questions. Furthermore, and

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

75

especially with young children, interviewers must assess the child's ability to distinguish between truthful and untruthful responses. Because of space limitations, we cannot develop the relevant issues properly. The interested reader can find comprehensive reviews of these and related topics and several recommended procedures in Poole and Lamb (1998) the "Memorandum of Good Practice" (Home Office, 1992) and within publications by various professional organizations (e.g., National Institute of Health and Human Development; American Professional Society on the Abuse of Children). CONCLUSION AND RECOMMENDATIONS FOR FUTURE RESEARCH AND PRACTICE

A myriad of studies over the past two decades have demonstrated clearly that the CI reliably enhances witness descriptions under a variety of event and test conditions. In fact, it seems to us that we have reached a point of diminishing returns as to the value of testing the CI in yet another minor variant of the basic experimental procedure. We think it is more valuable to pursue other lines of research, such as whether we can improve the CI's effectiveness by (a) training investigators to learn the procedure more efficiently; (b) streamlining he procedure so that police can use it more efficiently in field situations, where resources are often limited; (c) developing new component techniques to expand the CI; (d) developing methods that can be applied after an interview is conducted to discriminate between likely-correct and likely-incorrect statements; and (e) exploring the CI's utility in other (noncriminal) investigations. To end this chapter on a positive note, we are happy to report that considerable progress has been made along these lines. For instance, Powell, Fisher, and Wright (2005) explored various methods to improve training; Davis et al. (2006) have examined ways to streamline the interview for time-constrained situations; Seemungal and Stevenage (2002) demonstrated how post-interview probes can be used to assess the accuracy of witness statements; and Chapman and Perry (1995) have examined how the CI might be used to investigate a car accident. We encourage others to advance the interviewing research in similar directions. Research on interviewing children has run a parallel course, so that we now have a generally good idea of which techniques are effective and which techniques are ineffective. Here, too, the next major advance is likely to come in developing better training procedures so that we can convert our theoretical knowledge into realworld application (see Powell et al., 2005). We recognize that the actual conduct of an investigative interview in a real-world setting is difficult, and it is not simply a matter of following a recipe of pristine scientific principles. Nevertheless, practitioners can improve the quality of their interviews by following scientifically guided protocols, as has been shown repeatedly in laboratory and field studies. Practically, improving investigators' interviewing skills requires that police (and other) administrators properly appreciate the value of interviewing cooperative witnesses and, concomitantly, improve the training provided, following the lead of Britain. Second, practitioners should monitor the quality of their interviews and receive feedback on their performance.

76

FISHER AND SCHREIBER

For the future, both the scientific research and the actual conduct of investigative interviews in real-world settings would benefit by the establishment of closer working relationships between practitioners and scientists. Scientists could conduct better research studies if they were guided by the real-world insights of practitioners, and practitioners could implement the recommended procedures more effectively if they had a better understanding of the underlying scientific concepts. We encourage researchers and practitioners to work more closely with one another, so as to be of mutual benefit.

REFERENCES Anderson, R. C., & Pichert, ]. W. (1978). Recall of previously unrecallable information following a shift in perspective. Journal of Verbal Learning and Verbal Behavior, 17(1), 1-12. Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. London: Cambridge. Bartol, C. R., & Bartol, A. M. (2004). Psychology and law: Theory, research and application. Belmont: Wadsworth. Bekerian, D. A., & Dennett, J. L. (1993). The cognitive interview technique: Reviving the issues. Applied Cognitive Psychology, 7, 275-298. Bekerian, D. A., & Dennett, J. L. (1997). Imagery effects in spoken and written recall. In D. Payne & F. Conrad (Eds.), Intersections in basic and applied memory research (pp. 279-289). Mahwah, NJ: Lawrence Erlbaum Associates. Berresheim, A., & Weber, A. (2003). Die Strukturierte Zeugenvernehmung und Ihre Wirksamkeit [Structured witness interviewing and its effectiveness]. Kriminalistik, 57, 757-771. Boon,]. C. W., & Noon, E. (1994). Changing perspectives in cognitive interviewing. Psychology, Crime and Law, I, 59-69. Bower, G. H., Clark, M. C., Lesgold, A. M., & Winzenz, D. (1969). Hierarchical retrieval schemes in recall of categorized words lists. Journal of Verbal Learning and Verbal Behavior, 8, 323-343. Bransford, J.D., & Franks, J. ]. (1971). The abstraction of linguistic ideas. Cognitive Psychology, 2, 331-350. Brown, C. L., & Geiselman, R. E. (1990). Eyewitness testimony of mentally retarded: Effect of the cognitive interview. Journal of Police and Criminal Psychology, 6, 14-22. Brown, D., & Pipe, M.-E. (2003). Individual differences in children's event memory reports and the narrative elaboration technique. Journal of Applied Psychology, 88(2), 195-206. Campara, L.-B., Wagner, J.-T., & Saywitz, K.-J. (2001). Interviewing children about real and fictitious events: Revisiting the narrative elaboration procedure. Law and Human Behavior, 25(1), 63-80. Ceci, S. ]., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific analysis of children's testimony. Washington, DC: American Psychological Association. Ceci, S. J., Huffman, M. L. C., Smith, E., & Loftus, E. F. (1994). Repeatedly thinking about a non-event: Source misattributions among preschoolers. InK. Pezdek & W. P. Banks (Eds.), The recovered memory-false memory debate (pp. 225-244). San Diego, CA: Academic Press. Ceci, S. ]., & Loftus, E. F. (1994). "Memory work": A royal road to false memories? Applied Cognitive Psychology, 8, 351-364. Chapman, A.]., & Perry, D.]., (1995). Applying the cognitive interview procedure to child and adult eyewitnesses of road accidents. Applied Psychology: An International Review, 44, 283-294. Collins, R., Lincoln, R., & Frank, M. G. (2002). The effects of rapport in forensic interviewing. Psychiatry, Psychology & Law, 91, 69-78.

2. INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

77

Cooper, B.S., Kennedy, M.A., Herve, H. F., & Yuille, J. C. (2002). Weapon focus in sexual assault memories of prostitutes. International]oumal of Law and Psychiatry, 25, 181-191. Cutler, B. L., & Fisher, R. P. (1990). Live lineups, videotaped lineups, and photoarrays. Forensic Reports, 3, 439-448. Davis, M. R., McMahon, M., & Greenwood, K. M. (2005). The efficacy of mnemonic components of the cognitive interview: Towards a shortened variant for time-critical investigations. Applied Cognitive Psychology, 19, 75-94. Finger, K., & Pezdek, K. (1999). The effect of the cognitive interview on face identification accuracy: Release from verbal overshadowing. Journal of Applied Psychology, 84, 340-348. Fisher, R. P. (1996). Misconceptions in design and analysis of research with the cognitive interview. PSYCOLOQUY, 7(6), witness-memory. 12. fisher. (electronic journal). Fisher, R. P., Brennan, K. H., & McCauley, M. R. (2002). The cognitive interview method to enhance eyewitness recall. In M. Eisen, G. Goodman, & J. Quas (Eds.), Memory and suggestibility in the forensic interview. Mahwah, NJ: Lawrence Erlbaum Associates (pp. 265-286). Fisher, R. P., Falkner, K. L., Trevisan, M., &McCauley, M. R. (2000). Adapting the cognitive interview to enhance long-term (35 years) recall of physical activities. Journal of Applied Psychology, 85, 180-189. Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques in investigative interviewing: The cognitive interview. Springfield, IL: C.C. Thomas. Fisher, R. P., Geiselman, R. E., & Amador, M. (1989). Field test of the Cognitive Interview: Enhancing the recollection of the actual victims and witnesses of crime. Journal of Applied Psychology, 74(5), 722-727. Fisher, R. P., Geiselman, R. E., & Raymond, D. S. (1987). Critical analysis of police interviewing techniques. Journal of Police Science & Administration, 15, 177-185. Fisher, R. P., Geiselman, R. E., Raymond, D. S.,Jurkevich, L. M., & Warhaftig, M. L. (1987). Enhancing enhanced eyewitness memory: Refining the cognitive interview. Journal of Police Science and Administration, 15, 291-297. Fisher, R. P., Quigley, K. L., Brock, P., Chin, D., & Cutler, B. L. (1990). The effectiveness of the cognitive interview in description and identification tasks. Paper presented at the American Psychology-Law Society, Williamsburg, VA. Garven, S., Wood, J. M., & Malpass, R. S. (2000). Allegations of wrongdoing: The effects of reinforcement on children's mundane and fantastic claims.Joumal of Applied Psychology, 85, 38-49. Garven, S., Wood, J. M., Malpass, R. S., & Shaw, J. S. (1998). More than suggestion: The effect of interviewing techniques from the McMartin Preschool case. Journal of Applied Psychology, 83, 347-359. Geiselman, R. E., & Fisher, R. P. (1997). Ten years of cognitive interviewing. In D. G. Payne & R. G. Conrad (Eds.), A synthesis of basic and applied approaches to human memory (pp. 291310). Mahwah, NJ: Lawrence Erlbaum Associates. Geiselman, R. E., Fisher, R. P., Cohen, G., Holland, H., & Surtes, L. (1986). Eyewitness responses to leading and misleading questions under the cognitive interview. Journal of Police Science and Administration, 14, 31-3 9. Geiselman, R. E., Fisher, R. P., MacKinnon, D. P., & Holland, H. L. (1986). Enhancement of eyewitness memory with the cognitive interview. American Journal of Psychology, 99, 385-401. Geiselman, R. E., & Padilla, J. (1988). Cognitive interviewing with child witnesses. Journal of Police Science and Administration, 16, 236-24 2. George, R. C. (1991). A Field and Experimental Evaluation of Three Methods of Interviewing WitnessesNictims of Crime. Unpublished thesis, Polytechnic of East London. George, R. C., & Clifford, B. (1992). Making the most of witnesses. Policing, 8, 185-198. George, R.-C., & Clifford, B.-R. (1996). The Cognitive Interview-Does it work? InS. LloydBostock & G. Davies (Eds.), Psychology, law and criminal justice: International developments in research and practice (pp. 146-154). Oxford: Walter de Gruyter

78

FISHER AND SCHREIBER

Gilbert, J. A. E., & Fisher, R. P. (in press). The effects of varied retrieval cues on reminiscence in eyewitness memory. Applied Cognitive Psychology. Greenwald, A. G. (1970). Sensory feedback mechanisms in performance control: With special reference to the ideo-motor mechanism. Psychological Review, 77, 73-99. Gwyer, P., & Clifford, B. R. (1997). The effects of the cognitive interview on recall, identification, confidence and the confidence/accuracy relationship. Applied Cognitive Psychology, 11, 121-145. Haw, R. M., & Fisher, R. P. (2004). Effects of administrator-witness contact on eyewitness identification accuracy. Journal of Applied Psychology, 89, 1106-1112. Hayes, B. D., & Delamothe, K. (1997). Cognitive interviewing procedures and suggestibility in children's recall. Journal of Applied Psychology, 82, 562-577. Home Office (2002). Achieving best evidence in criminal proceedings: Guidance for vulnerable or intimidated witnesses, including children. London: Her Majesty's Stationery Office (also available on the World Wide Web at http://www.cps.gov.uk/publications/prosecution/bestevidencevoll .html; accessed January 31, 2005). Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall. Kebbell, M.-R., & Milne, R. (1998). Police officers' perception of eyewitness performance in forensic investigations. Journal of Social Psychology, 138(3), 323-333. Kebbell, M.-R., Milne, R., & Wagstaff, G.-F. (1999). The Cognitive Interview: A survey of its forensic effectiveness. Psychology, Crime and Law, 5(1-2), 101-115. Kebbell, M. R., & Wagstaff, G.F. (1997). Why do the police interview eyewitnesses? Interview objectives and the evaluation of eyewitness performance. The Journal of Psychology, 131 (6), 595-601. Koehnken, G., Milne, R., Memon, A., & Bull, R. (1999). The cognitive interview: A meta-analysis. Psychology, Crime and Law, 5, 3-27. Koriat, A., & Goldsmith, M. (1994). Memory in naturalistic and laboratory contexts: Distinguishing the accuracy-oriented and quantity-oriented approaches to memory assessment. Journal of Experimental Psychology: General, 123, 297-316. Koriat, A., & Goldsmith, M. (1996). Monitoring and control processes in the strategic regulation of memory accuracy. Psychological Review, 103, 490-517. Krafka, C., & Penrod, S. (1985). Reinstatement of context in a field experiment on eyewitness identification. Journal of Personality and Social Psychology, 49, 58-69. Larsson, A. S., Granhag, P. A., & Spjut, E. (2003). Children's recall and the Cognitive Interview: Do the positive effects hold over time? Applied Cognitive Psychology, 17, 203-214. Leibowitz, H. W., & Guzy, L. (1990). Can the accuracy of eyewitness testimony be improved by the use of non-verbal techniques? Paper presented at the American Psychology-Law Society, Williamsburg, VA. Leichtman, M. D., & Ceci, S. J. (1995). The effects of stereotype and suggestions on preschoolers reports. Developmental Psychology, 31, 568-578. Lepore, S. J., & Sesco, B. (1994). Distorting children's reports and interpretations of events through suggestion. Journal of Applied Psychology, 79, 108-120. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4, 19-31. Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589. Malpass, R. S., & Devine, P. G. (1981). Guided memory in eyewitness identification: Journal of Applied Psychology, 66, 343-350. McCauley, M. R., & Fisher, R. P. (1995). Facilitating children's recall with the revised cognitive interview. Journal of Applied Psychology, 80, 510-516. McMahon, M. (2000). The effect of the enhanced cognitive interview on recall and confidence in elderly adults. Psychiatry, Psychology and Law, 7, 9-32.

2.

INTERVIEWING PROTOCOLS TO IMPROVE EYEWITNESS MEMORY

79

Mello, E. W., & Fisher, R. P. (1996). Enhancing older adult eyewitness memory with the cognitive interview. Applied Cognitive Psychology, 10,403-417. Memon, A., Holley, A., Milne, R., Koehnken, G., & Bull, R. (1994). Towards understanding the effects of interviewer training in evaluating the cognitive interview. Applied Cognitive Psy-

chology, 8, 641-659.

Memon, A., & Koehnken, G. (1992). Helping witnesses to remember more: The cognitive interview. Expert Evidence, 1, 39-48. Memon, A., & Stevenage, V. S. (1996). Interviewing witnesses: What works and what doesn't? Psycoloquy, 7(6) witness-memory.l.memon. Milne, R., & Bull, R. (1996). Interviewing children with mild learning disability with the cognitive interview. In N. Clark & G. Stephenson (Eds.), Investigative and forensic decision making (pp. 44-51). Leicester: British Psychological Society. Milne, R., & Bull, R. (2002). Back to basics: A componential analysis of the original Cognitive Interview mnemonics with three age groups. Applied Cognitive Psychology, 16(7), 743-753. Milne, R., Bull, R., Koehnken, G., & Memon, A. (1995). The cognitive interview and suggestibility. In G. Stephenson & N. Clark (Eds.), Criminal Behaviour: Perceptions, Attributions and Rationality. Division of Criminological & Legal Psychology Occasional Papers No. 22. Leicester: British Psychological Society. Milne, R., Clare, C. I. W., & Bull, R. (1999). Using the cognitive interview with adults with mild learning disabilities. Psychology, Crime and Law, 5, 81-99. Nathan, D., & Snedeker, M. (1995). Satan's silence. New York: Basic Books. Newlands, P. J., George, R. C., Towell, N. A., Kemp, R.I., & Clifford, B. R. (1999). An investigation of description quality from real-life interviews. Psychology, Crime & Law, 5(1-2),

145-166.

Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Reinhart & Winston. Pecher, D., Zeelenberg, R., & Barsalou, L. W. (2003). Verifying different-modality properties for concepts produces switching costs. Psychological Science, 14, 119-124. Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. Poole, D. A., & White, L. (1991). Effects of question repetition on the eyewitness testimony of children and adults. Developmental Psychology, 27, 975-986. Powell, M. B., Fisher, R. P., & Wright, R. (2005). Investigative interviewing. InN. Brewer & K. Williams (Eds.), Psychology and law: An empirical perspective (pp. 11-42). New York: Guilford. Roberts, W. T. & Higham, P. A. (2002). Selecting accurate statements from the cognitive interview using confidence ratings. Journal of Experimental Psychology: Applied, 8, 33-43. Roediger, H. L., & McDermott, K. B (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory & Cognition, 21, 803-814 Saywitz, K. J., & Geiselman, R. E. (1998). Interviewing the child witness. Maximizing completeness and minimizing error. InS. J. Lynn & K. M. McConkey (Eds.), Truth in memory. New York: Guilford Press. Saywitz, K. J., Geiselman, R. E., & Bomstein, G. K. (1992). Effects of cognitive interviewing and practice on children's recall performance. Journal of Applied Psychology, 77(5), 744-756. Schreiber, N., & Parker, J. F. (2004). Inviting witnesses to speculate: Effects of age and interaction on children's recall. Journal of Experimental Child Psychology, 89, 31-5 2. Seemungal, F. V., & Stevenage, S. V. (2002). Using state of awareness judgments to improve eyewitness confidence-accuracy judgments. In M. Izaute & P. Chambres (Eds.), Metacognition: Process, function and use (pp. 219-231). Dordrecht, the Netherlands: Kluwer. Smith, S. M. (1988). Environmental context-dependent memory. In G. M. Davies & D. M. Thomson (Eds.), Memory in context: Context in memory (pp. 13-34). New York: Wiley. Sporer, S. L. (1996). Experimentally induced person mix-ups through media exposure and ways to avoid them. InS. Lloyd-Bostock & G. Davies (Eds.), Psychology, law, and criminal justice: International developments in research and practice (pp. 64-73). Oxford: Walter de Gruyter.

80

FISHER AND SCHREIBER

Steblay, N. M. (1997). Social influence in eyewitness recall: A meta-analytic review of lineup instruction effects. Law and Human Behavior, 21, 283-298. Steblay, N., Dysart, J., Fulero, S., & Lindsay, R. C. L. (2001). Eyewitness accuracy rates in sequential and simultaneous lineup presentations: A meta-analytic comparison. Law and Human Behavior, 25(5), 459-473 T ulving, E. (1983). Elements of episodic memory. Oxford: Oxford University Press. Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80, 352-373. Wells, G. L. (1978). Applied eyewitness testimony research: System variables and estimator variables. Journal of Personality and Social Psychology, 36, 1546-1557. Wells, G. L., Rydell, S. M., & Seelau, E. P. (1993). The selection of distractors for eyewitness lineups. Journal of Applied Psychology, 78(5), 835-844. Yuille, J. C., Hunter, R., Joffe, R., & Zaparniuk, J. (1993). Interviewing children in sexual abuse cases. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses. Understanding and improving testimony (pp. 95-115). New York: Guilford Press.

3 The Influence of Emotion on Memory in Forensic Settings Daniel Reisberg and Friderike Heuer Reed College

Being the witness or victim of a crime is in most cases a highly emotional experienceduring the crime itself and in the minutes (or months) afterward. If the crime involves or threatens violence, the witnesses and victims are likely to feel fear. Other crimes will leave the witnesses and victims feeling angry, and still other crimes might trigger an episode of emotional disturbance (e.g., anxiety or depression) that continues long after the crime itself. And in some cases, these emotional reactions will be quite strongperhaps rising to the level we call trauma. How will these various aspects of emotion influence the witness's or victim's memory? How will they influence the completeness, accuracy, or longevity of the memory, or the ease with which the memory can be retrieved later on? These issues set the agenda for this chapter. In the chapter's first section, we review evidence that indicates that moderate levels of emotion promote memory-at least for an event's central aspects or gist. In the next section, we turn to evidence that suggests that the same levels of emotion may have the opposite effect, undermining memory, for more peripheral aspects of an event. In the following section, we consider the memory effects of stronger emotion and trauma. We next consider at length the memory effect of stress. We then turn to a different issue, asking how emotion, experienced during an event, can influence the ease with which the event can be recalled; we move from this to a consideration of the highly contentious claim that emotion during an event can cause that event to be "repressed" later on, and thus not available for recall. We then shift our focus and examine emotional memories in a group often of special interest for the legal system: children. Finally, we turn to a different issue-namely, the potential use of emotionality, associated with a memory, as a means of distinguishing which memories are correct and which are not. Before we turn to these materials, several preliminary points should be addressed. First is the nontrivial task of defining terms like emotion, stress, arousal, and trauma. Each 81

82

REISBERG AND HEUER

of these definitions has been the topic of some debate, in part because each term can be defined within multiple frameworks. Thus, for example, one might seek to define emotion in terms of its subjective qualities (how does one feel when one feels emotional?), in terms of its psychological consequences (what behavioral or cognitive effects can we detect and measure when someone is emotional?), in terms of its bodily concomitants (how does the pattern of blood chemistry or brain activation change when someone is emotional?), in terms of the situation itself (what is it that defines a fearful event or a happy story?), or, finally, in terms of a mix of these criteria. (For a glimpse of the technical debate over how exactly "emotion" should be defined, see, among others, Gross, 2002; Keltner & Gross, 1999; LeDoux, 1996; Levenson, 1994; Oatley &Jenkins, 1996; Roseman & Smith, 2001.) In addition, people can of course feel a wide variety of different emotions, and these differ in their subjective presentation, their consequences for thought and action, and their effects on the body. We will return to this point later in this chapter, but, for now, simply note that this point adds complexity to the task of defining "emotion," or in saying what psychological or biological states accompany emotion. By the same logic, there are also large differences from one person to the next in their emotional responses. Thus, for example, some people will find an event frightening while others will not; some people will be made slightly angry by an event, while others might become infuriated. These differences are surely shaped by the beliefs or expectations that each individual brings to an event, and also by the personality of each individual, with some people particularly responsive to emotional events and others less so. These considerations are likely to have considerable importance in governing how these events are remembered (cf. Deffenbacher et al. 2004; Levine & Pizarro, 2004; Yuille & Daylen, 1998), and we will return to this issue as well. But, again, these differences add difficulty to the problem of defining emotion, or more broadly, in deciding which events should be counted as "emotional." Similar concerns are attached to the word stress (cf. Kim & Diamond, 2002; Payne, Nadel, Britton, &Jacobs, 2004; Yuille & Daylen, 1998) or trauma. Technical definitions for these terms have been offered-for example, as part of the definitions of posHraumatic stress disorder (PTSD) and acute stress disorder, in the Diagnostic and Statistical Manual for Mental Disorders or DSM (American Psychiatric Association, 1994). However, these definitions are themselves controversial-in part because some individuals seem to develop PTSD as a result of events not included within these definitions (and so, peculiarly, these individuals have a post-traumatic pattern even though, technically, they have experienced no trauma). This suggests that what counts as "traumatic" may vary to some extent from person to person, obviously complicating our efforts to create definitions (c£ Heim, Owens, Polotsky, & Nemeroff, 1997; Schacter, 1996; Yehuda, 1997). In what follows, therefore, we cannot rely on explicit and consensual definitions. Instead, we rely on a variety of markers for emotionality. Thus, for example, we speak of an emotional event if (a) the event includes the sort of content that most people would regard as emotional, (b) the event causes most people to report that it makes them feel emotional, and (c) the event also causes a change in bodily state of the sort usually associated with emotion (e.g., changes in heart rate or an increase in the blood level of nor-

3.

MEMORY AND EMOTION

83

epinephrine). Likewise, we use the term trauma to refer to the response people have to events that are (or are perceived to be) extremely negative, highly threatening (e.g., to life or health), accompanied by a very high level of bodily arousal, and typically associ~ a ted with a sense of complete absence of control. As we will see, these broad definitions will serve us reasonably well in reviewing the available evidence. However, we will need to return to these definitional issues later in this chapter, when we consider the fact that there are different types of emotion, and, with that, differences among people in how they respond to emotional events. As a second preliminary, we should also note the sorts of evidence that are pertinent throughout this chapter. Each of the topics we discuss is well researched, and, for most of the points we consider, the evidence draws on a mix of paradigms. These include laboratory studies that-for purposes of experimental control-use artificial materials (word lists or sentences to be memorized) and also laboratory studies that rely on more realistic materials (e.g., participants are shown a video, a narrated slide sequence, or, in some cases, a live event that conveys an emotional story, and they are unexpectedly asked later to remember this material). The evidence, where possible, also draws on studies of emotional events, and, in some cases, memory in actual witnesses memory for real~life seeking to recall real crimes. Finally, because some of the effects we consider are biolog~ ical in nature, data can also be drawn from animal studies. The procedures in these lat~ ter studies are obviously at some distance from the circumstances of a crime eyewitness, but these studies often allow measurements and manipulations that are not possible with human participants and in this fashion allow important insights into the mechanisms underlying effects observed with actual witnesses. Most of the evidence we describe here is published in carefully documented journal articles, and this is obviously important if this evidence is to survive a Daubert type of review. 1 Moreover, the claims we describe are each based on a number of studies, and preferably a diversity of studies, and it is the convergence of these various data sources that provides our assurance that the claims are correct. However, we suspect that many readers of this chapter would not be well served by our cataloguing of the relevant data sources in the manner customary for scholarly reviews, because many of the sources are in specialized journals that may be inaccessible to attorneys or professionals in law en~ forcement. (Moreover, the chore of synthesizing the primary data sources is probably one best left to the relevant specialists.) For this reason, we have tried in this chapter to cite the relevant secondary sources-research summaries and overviews-rather than the primary data sources themselves. Readers who wish to locate the primary data should easily be able to do so via these secondary summaries.

1 The admissibility of expert evidence in a trial is governed by several standards, but, in U.S. courts, the key criteria for admissibility derive from the Supreme Court's ruling in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993). Among other points, this ruling instructs the courts to pay attention to whether the expert's testimony will involve claims that are scientifically reliable. Reliability, in turn, depends on whether the claims are testable and have been tested; whether the claims have been subjected to peer review and publication; whether the claims have a known and acceptably low rate of error; and whether the claims are generally accepted among the relevant experts.

84

REISBERG AND HEUER

EMOTION PROMOTES MEMORY

A number of studies have documented a strong positive relationship between how emotional an event was at the time that it occurred and how vividly the event is remembered later on. To put the result in simple terms, it seems that the stronger the emotion, the greater the vividness of the subsequent recollection. In one study, for example, the correlation between vividness and emotionality was .89 for sad events, .68 for angry events, and .90 for fearful events (Reisberg, Heuer, McLean, & O'Shaughnessy, 1988; see Reisberg & Heuer, 2004, for related data). 2 The problem with this finding, however, is that vivid memories are not always accurate memories, and, in fact, a close scrutiny of emotional memories reveals that some are highly accurate and others are not. For illustration, consider a study of people's memories for an event, in 1992, in which an El AI cargo plane lost power in two of its engines just after taking off from Amsterdam's Schiphol Airport. The pilot was unable to return the plane to the airport, and the plane crashed into an apartment building, destroying the building and killing 43 people. Ten months later, investigators interviewed Dutch participants about this highly emotional event and found that two-thirds of the people questioned reported seeing the television film of the moment the plane hit the building and were willing to answer a number of detailed questions about what they had seen in the film. However, there was no such film; no camera had recorded the crash, and no film (or any reenactment) was shown on television (Crombag, Wagenaar, & van Koppen, 1996). The participants in the study by Crombag et al. had vivid and detailed recollections of the El AI crash; their memories were reported with considerable confidence. Even so, substantial parts of these memories were mistaken-they were plainly at odds with how the event actually unfolded. (For related examples, also showing substantial error in the memory for emotional events, see Greenberg, 2004; Neisser & Harsch, 1992; Ost, Vrij, Costall, & Bull, 2002; Schmolck, Buffalo, & Squire, 2000.) Clearly, therefore, the accuracy of emotional memories-no matter how compelling the memory-based reportscannot be assumed, but instead must be assessed. In addition, we obviously need to ask whether the data of Crombag et al. are in any fashion representative of how, in general, emotional events are remembered. To be sure, we can also document cases of considerable accuracy in the recall of emotional events (e.g., Conway et al., 1994; Yuille & Cutshall, 1986), and so we know at the outset that the data pattern found by Crombag et al. is not by any means inevitable. But this simply leaves us asking: How well, in general, are emotional events recalled? To address questions like these, many researchers have turned to laboratory studies of emotional memory. In the laboratory, we know how the to-be-remembered event actually unfolded, and so memory accuracy can easily be assessed. Equally important, in the laboratory we can control the nature of the event being recalled and, in that fashion, isolate the effects of emotion, with other factors held constant. What has the laboratory taught us? A wide range of studies indicate that emotion, experienced during an event, promotes memory for the event's gist or central aspects. 2

For readers not familiar with correlation values, the maximum possible correlation is 1.00. A correlation of .89 or .90, therefore, indicates a very strong relationship. We might also mention that the three correlation values mentioned here are not statistically different from each other.

3.

MEMORY AND EMOTION

85

Thus, the gist of emotional events seems to be remembered more accurately and more completely than the gist of emotionally neutral (but otherwise comparable) events. At longer delays, this advantage for emotional events grows even more pronounced, suggesting that the emotionality also slows the process of forgetting. This broad pattern of results has been documented with diverse experimental procedures and with different types of to-be-remembered materials. For example, emotional pictures seem to be better remembered than neutral pictures (e.g., Canli, Zhao, Brewer, Gabrieli, & Cahill, 2000), emotion-laden word lists are better remembered than neutral lists (e.g., Dietrich et al., 2001; Jones, O'Gorman, & Byrne, 1987), and emotional stories (including verbal humor) are better remembered than neutral stories (e.g., Schmidt & Williams, 2001). Most relevant for forensic settings, many studies have shown that the gist and central aspects of emotional events are better remembered than the comparable aspects of emotionally neutral events (Reisberg & Heuer, 2004). Several factors contribute to this well-documented data pattern. Some factors are biological in nature, and, indeed, investigators have been able to document in detail the physiological effects that emotion (and the bodily arousal that goes with it) have on memory. (For broad summaries of these mechanisms, see, for example, Buchanan & Adolphs, 2004; Cahill & van Stegeren, 2003; Hamann, 2001; McGaugh, 2003.) In addition, emotional events are usually important to us, virtually guaranteeing that we will pay close attention as the event unfolds, and close attention contributes to more accurate and more complete remembering. Emotional events are also emotional precisely because they are related to issues we care about and may have thought about in other contexts; this fosters the sort of memory connections that we know promote retention and recall. Finally, we tend to mull over emotional events in the minutes (or hours) following the event, and this is tantamount to memory rehearsal, which, again, has a positive effect on memory. On all of these grounds, the memory advantage for emotional events seems to be nearly inevitable, making it unsurprising that the advantage has been documented in many studies. Surprising or not, however, the evidence is clear in showing that emotion increases the accuracy, completeness, and longevity of memory for an event's center or gist. However, there are three crucial limits on this claim. First, as the result of Crombag et al. reminds us, emotion promotes memory accuracy but does not guarantee memory accuracy-memory errors are possible even for emotional events. Second, we have emphasized that emotion promotes memory for the gist and central aspects of an event; what about materials that are not central to the event? This is our concern in the next section. Third, are these claims independent of the extremity of the emotion? We turn to this point in a later section.

EMOTION AND MEMORY FOR AN EVENT'S "PERIPHERY"

Commonsense proposals sometimes liken memory to a video recorder, but, on many grounds, this analogy is misleading. Among other concerns, if a video camera is turned on and in focus, then it records everything that is in view. Memory, in contrast, is selec-

86

REISBERG AND HEUER

tive. In any particular event, many objects are in view, and each of these objects has many attributes. Different occurrences take place, and these can be described at a coarse grain ("He robbed us") or a fine grain ("He first approached to within 8 feet of us, then said hello, then moved closer... "). However, memory (unlike a camera) does not record all of these, and this requires that we ask: Which of the event's attributes, and which of the occurrences, will be remembered, and at what level of grain? The answer to this question depends, to a large extent, on what the witness is paying attention to, on the straightforward principle that those things that are attended to and thought about are likely to be remembered, whereas those things not attended to are likely never to be recorded into memory. This selectivity can be documented in a number of ways, both formally and informally. As an information demonstration, consider the simple fact that many people seem unable to describe the appearance of their own wristwatches. (Are numerals shown at all 12 positions? Or just at the "compass points"? What is the font in which the numerals are shown?) Of course, people have looked at their watches on countless occasions but, on each occasion, have attended to the time, and not the watch's appearance. As a result, the appearance is not encoded into memory and cannot be recalled. (For a more formal demonstration of strikingly poor memory despite countless exposures, see the study by Nickerson & Adams [1979], showing remarkably poor memory for the layout of an ordinary U.S. penny, despite the fact that people have seen pennies thousands of times, underscoring the fact that it is attention that matters for memory, and not the number of exposures.) A related demonstration is provided by Powers, Andriks and Loftus (1979): They found similar levels of overall memory performance in males and females, but found that each gender tended to remember different aspects of event: Males had better memory for details about a purse snatcher they had seen, but females had better memory for the victim's clothing. For both genders, then, we see the selectivity of memory, and, specifically, better memory for those aspects that were more closely attended. Thus, memory does not record an entire scene (as a camera would), but instead records just those bits, within the scene, to which the witness pays attention. On this basis, we need to ask whether emotion, during an event, changes what people pay attention to. If it does, then it will influence what people remember. One often-quoted claim with regard to this issue was offered many years back by Easterbrook (1959), who argued that bodily arousal (which is obviously an accompaniment to emotion) causes a decrease in the "range of stimulus cues" that an organism takes in. As a result, an aroused organism is likely to focus its attention on just those elements of the scene that are of particular interest and will be correspondingly insensitive to all else within the scene. This focusing of attention would then contribute to the well-documented memory advantage for those elements that are interesting because those elements will presumably receive more of the organism's time and effort. At the same time, this focusing of attention would also produce a memory disadvantage for less interesting, less important materials, because these will be excluded from the narrowed "range" of attention and hence not remembered later on. To put this same idea into more common terms, the proposal is that emotionality will make people less observant. According to the Encarta dictionary (for example) observant

3.

MEMORY AND EMOTION

87

means "paying such careful attention that little or nothing is unnoticed." This is the state disrupted by emotion, because the narrowed range of attention makes it quite likely that a great deal, within the scene, will be unnoticed, although those bits that are noticed will be better remembered than they would have been in the absence of emotion. Many studies have tested these claims, and (with a few exceptions) have confirmed what is often called the Easterbrook Hypothesis: In these studies, research participants who experience an emotional event have more complete, more accurate memories for the event's gist and central materials, compared with participants who experience an emotionally neutral (but otherwise comparable) event. But, in addition, the participants who experience the emotional event tend to remember less of the event's peripheral details, compared with participants who experienced a neutral event (for overviews, see Reisberg & Heuer, 2004; Edelstein, Alexander, Goodman, & Newton, 2004). This pattern has been replicated many times in the laboratory (as one example of the lab evidence, see Burke, Heuer, & Reisberg, 1992). Importantly, however, the pattern has also been observed in real-life situations, in which the research participants are "in" the event, and not just passively viewing it, and with the source of the emotionality literally on the scene, and not merely shown in a photograph. Thus, this pattern of improved memory for central materials and worse memory for peripheral materials has been documented with real-life crime victims (e.g., Christianson & Hubinette, 1993) and with spider phobics, invited into a room in which they observed a large, live spider, contained within a glass jar (Wessel & Merckelbach, 1997, 1998). Consistent with other studies, the spider phobics (presumably quite aroused by the sight of the spider) showed reasonably accurate memory for the event's center (the spider) but worse memory for the event's periphery (other aspects of the room), in comparison with control participants. Moreover, the pattern of these data (again, relatively accurate memory for central materials within the event, but a memory disadvantage for peripheral materials) mirrors another well-documented finding known as "weapon focus." This term refers to a tendency for crime witnesses to lock their attention on the gun or knife, if one is present, to the exclusion of all else. (Even if a weapon is merely threatened, weapon focus can still be observed, with the witness focusing on, say, the hand in the pocket that might be holding a weapon.) Because of this focus, the witness is likely to have detailed memory for the weapon later on (and detailed memory for the hand holding the weapon-such as whether the person was wearing gloves or a ring). At the same time, however, a witness to a crime involving a weapon is likely to have relatively poor memory for other aspects of the scene-including such forensically crucial information as what the perpetrator looked like. Indeed, a broad summary of the studies on this topic indicates that eyewitness identifications of the perpetrator are systematically less accurate in crimes involving weapons-presumably because the witness's attention was focused on the weapon, not on the perpetrator's face (Steblay, 1992). The weapon-focus pattern has been demonstrated in many studies, including studies that probe participants' memory after the event and studies that literally track where participants are pointing their eyes during the event. (Not surprisingly, participants spend a considerable amount of time directing their eyes toward the weapon; Loftus, Loftus, & Messo, 1987; Stanny &Johnson, 2000.) Overall summaries of these data, using a powerful

88

REISBERG AND HEUER

statistical technique called meta-analysis, confirm the reliability of this pattern and, importantly, show that the weapon-focus effect is actually stronger and more reliable in those studies that are more similar to actual forensic settings, in comparison with more artificial lab settings (Steblay, 1992). Thus, the weapon-focus effect seems not to be apeculiar by-product of the specific situations created in the lab; instead, the effect is probably underestimated in the laboratory studies. Can the weapon-focus pattern be detected in actual police investigations? In one study, Tollestrup, Turtle, and Yuille (1994) surveyed the police records of a suburban community, examining the files for all cases of robbery committed between 1987 and 1989. They found that, when a weapon was on the scene, only 44% of the identifications made by the witnesses were of the person the police had identified as their suspect; in contrast, 71% were of the police suspect in crimes without weapons. If one assumes that the police suspect was, in fact, the actual perpetrator, then this suggests markedly diminished accuracy in the presence of a weapon (in accord with the laboratory data). Other findings, however, complicate this picture. For example, Tollestrup et al. found that the presence of a weapon did not influence the accuracy of witnesses' or victims' estimations of the perpetrator's age, height, or weight. They also found that victims and witnesses of crimes with weapons reported more details (about the clothing or physical appearance of the perpetrator) than did the victims or witnesses of weaponless crimes. Taken at face value, this seems to be the opposite of what one might expect in light of the standard weapon-focus pattern, inasmuch as it implies that witnesses to crimes with weapons took in more information about the scene, rather than less. (For a similar finding, see Cooper, Kennedy, Herve, & Yuille, 2002.) A further challenge to the weapon-focus claim comes from a study by Valentine, Pickering, and Darling (2003). They surveyed 640 attempts by eyewitnesses to identify the culprit in actual lineups organized by the Metropolitan Police in London; the cases involved a wide range of crimes (including murders, rapes, armed robbery, burglary, etc.). They found no effect of weapon presence on the likelihood that a witness would select the police suspect from a lineup. Similarly, Behrman and Davey (2001) examined 271 cases drawn from the police files in Sacramento, California. They also found no effect of a weapon's presence, either in the number of overall identifications or in the percentage of witnesses selecting the police suspect. What should we make of this mixed pattern? On one hand, these various studies of actual police records (and the study by Cooper et al. [2002], which interviewed actual crime victims) have an enormous advantage: they examine memories from real witnesses (and victims) to real crimes, and, in that regard, are vastly more realistic than the laboratory studies. On the other hand, these real-world studies also have a number of crucial weaknesses. For example, an assessment of how many details someone reports (as in the studies by Tollestrup et al. or Cooper et al.) may tell us nothing about how much the person actually remembers. One reason for this is that we have no way of assessing the accuracy of these reported details, and, as already discussed, accuracy certainly cannot be assumed. Thus, at least some of these findings (e.g., the sheer number of details reported) are irrelevant to claims about weapon focus; these claims are concerned with what the person actually remembers, and not with what the person says that he or she remembers.

3.

MEMORY AND EMOTION

89

Even if we set the issue of accuracy to the side, we need to ask: did the witnesses in the study by Tollestrup et al. report more details about crimes with weapons because they remembered more? Or did they report more details about these crimes for an entirely different reason, quite independent of what they remembered? As one option, it seems plausible that witnesses to crimes with weapons may simply choose to talk more about their memory, motivated by their own sense of the seriousness of the crime. Likewise, it seems plausible that police press witnesses for more information when the crime involved a weapon. In either case, a witness to a crime with a weapon might end up reporting more details for reasons having nothing to do with the impact of the weapon's presence on what they actually remember. What about the lineup selections made by these victims? Here too there is a complication: in the studies of police records, the "accuracy" of lineup selections is assessed by asking whether the witness chooses the police suspect from the lineup. This is reasonable if we assume the police suspect is, in fact, the culprit. However, this may not always be the case, and, at the least, this will add variability to the data within each study. Worse, police investigators vary widely in how they use identification information. Police in some jurisdictions conduct identification procedures early in an investigation, as a means of narrowing the field of suspects. Police in other jurisdictions tend to wait until other evidence makes it reasonably likely that the person of interest is in fact the perpetrator, using the identification more as confirmation than as an investigative tool. In light of this variation in practice, the assumption that the suspect is the perpetrator (and thus using this assumption as a means of assessing memory accuracy) may be highly questionable in some investigations but quite plausible in others. This obviously makes it difficult to compare accuracy rates across the various studies and may explain why some real-world studies (e.g., Tollestrup et al., 1994) do find a weapon effect whereas other studies (Behrman & Davey, 2001; Valentine et al., 2003) do not. Finally, these real-world studies also have a further problem: in the laboratory, one can hold constant variables other than the one being examined and in this fashion create what researchers call a valid design (one that measures what it is supposed to measure) and a sensitive design (one that is likely to detect the effects of interest, without losing these effects against a backdrop of irrelevant variation in the data). Of course, this is not possible in the world outside of the laboratory. Crimes differ in their nature, their severity, and their duration; the delay between the crime and the identification varies; the number of times a witness has told and retold his or her account, prior to the identification, can range widely; and so on. As a result, the real-world data are, in technical terms, statistically noisier than the laboratory data, containing appreciably more variation, and hence less sensitive. It is inevitable, therefore, that some genuine, important, and reliable effects will be detected in the (more sensitive) measures of the laboratory but will not be detected in the (noisier) real-world data. This is not because these effects are absent in the real-world data, but instead because the nature of these data makes it difficult to document any effects. This, too, helps us understand why the real-world studies differ in their reports of weapon-focus effects. (We should note that this concern is by no means limited to the weapon-focus issue. It is striking, for example, that Behrman and Davey (2001), do find a contrast between "same-race" identifications and "cross-race"

90

REISBERG AND HEUER

identifications, in accord with the highly robust pattern often documented in laboratory data, but Valentine et al. do not find this effect. Thus there can be no question that the pattern of the real-world data-on virtually any issue-is far from crystal clear, largely for the methodological reasons just discussed.) In summary of the weapon-focus evidence, then, this phenomenon can certainly be documented reliably in the laboratory-both with measures of memory and with measures of eye movements. This phenomenon can also be documented outside of the lab-for example, in the identification rates observed in the survey by Tollestrup et al. of actual robbery cases. Other data do seem to challenge the weapon-focus idea, but, as we have now seen, these other findings are problematic-with concerns about whether memory accuracy has been correctly assessed and about the capacity of the experiments to detect the effect at issue. Overall, then, the weight of the evidence suggests that the weapon focus effect must be taken seriously-as a robust phenomenon in the laboratory, and as a pattern that can, in some cases, be documented outside of the laboratory as well. 3 With all of this said, there is still another question to be addressed-namely, what it is that causes the weapon-focus pattern. On the one side, it seems plausible that this pattern is a manifestation of the effect described by Easterbrook-with the witness's fear or anger causing a narrowing of attention, centered on the weapon. Some authors, however, have pointed out that the eyewitness may focus on the weapon simply because it is by far the most important and interesting aspect of the visual input. After all, what could be more important to a crime victim than to know whether he or she is in immediate danger? To this end, nothing in the scene is more important than knowing whether the weapon is cocked and pointed at him or her. On this logic, even an entirely calm witness

3 Astute readers may raise an objection to the argument we are offering here, but it is an objection that is easily dealt with. As we have just seen, the weapon-focus phenomenon has not been detected in several studies of actual police records. Perhaps, therefore, this is a phenomenon that the criminal justice system is well advised to ignore. After all, if we cannot consistently document the effect in the field, then why should we worry about its impact? The answer to this is straightforward. Consider the following analogy. There is no question that the cells within the human brain comrrunicate with each other via electrical signals, and these signals can be detected and measured, with suitable instruments, in the laboratory. However, imagine that we wanted to measure these signals, within the brain, in a person standing in the middle of an electrical power plant. This latter effort would fail, because the complex electrical fields produced by the power plant itself would interfere with our ability to measure the electrical signals within the brain. But, of course, this does not mean that the brain's signals play no role when someone is standing in the power plant; the brain cells work on the same principles no matter where they are. The power plant, in other words, does change our ability to measure the signals, but has little impact on the signals themselves, or on their consequences within the brain. The same logic applies to weapon focus. With other factors on the scene, sometimes augmenting weapon focus, sometimes working against it, it will be difficult to detect and measure weapon focus (or any other phenomenon) in the context of real-world crimes. This does not mean that these phenomena have no influence in this setting. The influence is still there and is part of the broader set of factors influencing what a witness remembers. Said differently, we should not confuse our inability to detect an effect with the effect being absent, or having no influence.

3.

MEMORY AND EMOTION

91

might still show the weapon-focus pattern, zooming in on the weapon simply because looking toward the weapon provides crucial information. As it turns out, evidence suggests that both of these mechanisms-one hinging on arousal, and one hinging on the weapon's visual importance-play a role in producing weapon focus. In several studies, weapon focus has been observed even in the absence of emotional arousal, indicating the weapon's potency as an attention magnet (Kramer, Buckhout, & Eugenio, 1990; Loftus et al., 1987; Maass & Kohnken, 1989). Other studies indicate that it may be the unusualness of the weapon, and not the threat, that produces weapon focus (Pickel, 1998). Still other studies, however, suggest that the strength of the weapon focus effect increases as arousal increases, indicating that arousal clearly plays a role (e.g., Peters, 1988; Steblay, 1992). These points, in turn, raise a further issue. We have noted that, in situations involving emotion (but no weapon), people usually show a pattern of memory narrowinggood memory for the event's center, but impaired memory for the event's periphery. We have so far suggested that this narrowing is produced by the emotionality itself; that is, of course, the Easterbrook proposal. However, studies of weapon focus remind us that there is another possibility-namely, that memory narrowing is observed because emotional situations often provide a specific stimulus that demands the witness's attention, and it is this attention magnet (i.e., a stimulus that the witnesses deem to be well worth focusing on), and not the emotion per se, that produces the narrowing. In this context, it is important to consider how exactly emotional remembering has been studied in the laboratory. In order to elicit emotional responses from research participants, laboratory procedures have almost invariably relied on gruesome or shocking stimuli. In one study, participants were shown a picture of a boy who had been shot and who is now bleeding from his eyeball; in another study, participants viewed a picture of a child whose legs had been severed in a car accident; in another study, participants were shown a picture of a woman whose throat had been slit, and so on. It seems plausible that these gripping sights seize participants' attention during the event, and this (and not the emotion per se) is what produces the memory narrowing. Said differently, participants in our studies may well have been influenced by the fact that the events they witnessed contained some specific salient attention-demanding object or action. Independent of emotion, witnesses focus on this attention magnet and so remember it well-but at the cost of remembering less about other aspects of the scene. Support for this proposal comes from a pair of experiments by Laney, Campbell, Heuer and Reisberg (2004). Their research participants viewed a stimulus sequence that contained no shocking, gory, or horrifying material. Instead, emotion was elicited via the participant's empathy with the characters depicted in the stimuli. In this situation, no memory narrowing was observed; instead, emotion seemed to improve memory for all aspects of the stimulus sequence-including both central and peripheral aspects. Where, then, does all of this leave us? First, as we noted in the previous section, emotion seems in many studies to promote memory for an event's gist or central content and memory for details that happen to be (spatially or temporally) associated with this central content. Second, emotion is generally associated with a pattern of memory narrowing, so

92

REISBERG AND HEUER

that memory for peripheral aspects of the scene (objects or actions that are neither crucial for the plot nor spatially central) is impaired. Third, this memory narrowing may not be a consequence of the emotion itself. Instead, the narrowing may be the result of the fact that emotional stimuli are often visually unusual or striking for some other reason. This gives these stimuli potency as attention magnets that grip the witness's attention during the event and dominate their subsequent recall. We emphasize, however, that this third point takes away none of the importance of memory narrowing. To be sure, there is no necessary linkage between emotion and memory narrowing. As we have seen, narrowing can sometimes be observed even in the absence of emotion (as it is for some cases of weapon focus). Conversely, some emotional remembering does not display narrowing (e.g., Laney et al., 2004). Even so, many emotional events (and certainly those of forensic interest) do contain attention magnets, and so memory narrowing will often be a feature of emotional remembering.

THE MEMORY IMPACT OF EXTREME EMOTION

What happens when a witness or victim is exposed to more extreme levels of emotion, including the levels we might call traumatic? For years, the answer to this question has been cast in terms of the proposal by Yerkes and Dodson (1908), namely that there is some optimal level of emotion for promoting memory and that poorer and poorer memory is observed as one moves farther away (in either direction) from this optimum. (If this relationship were graphed, with some measure of emotional intensity on the horizontal axis, and memory accuracy on the vertical axis, the result would be an inverted U, and thus the Yerkes-Dodson proposal is sometimes described as the inverted-U hypothesis.) This proposal is endorsed by many psychologists but nonetheless is probably wrong. As one large problem, Christianson (1992) noted that there is little direct evidence in favor of the Yerkes-Dodson pattern in emotional memory. Moreover, other evidence makes it clear that there are different types of memory, and circumstances that promote one sort of memory may not promote other sorts. As a result, there may not be a uniform relationship between level of emotion and memory; instead, the relationship may depend on what types of memory are being considered. In addition, no matter what type of memory we consider, the relationship between level of emotion and memory performance is likely to depend on other variables, so that the optimal level of emotion, for even a single type of memory, may change in the presence of other factors (cf. Deffenbacher, 1994; Deffenbacher et al., 2004; Fazey & Hardy, 1988; Yuille & Daylen, 1998). The most direct challenge for the Yerkes-Dodson proposal, however, comes when we consider how people remember emotionally extreme events, including the events we call traumatic-that is, events that are experienced as profoundly negative, which cause powerful changes in bodily arousal (including the changes that characterize biological stress-cf. Payne et al., 2004), and which threaten the life or well-being of the individual. On most construals of the Yerkes-Dodson function, memory for these events should be extremely poor. The data, however, indicate otherwise.

3.

MEMORY AND EMOTION

93

In a small number of cases (and consistent with the Yerkes-Dodson claim), traumatic experiences lead to a pattern known as psychogenic amnesia, 4 in which the victims or witnesses of the trauma apparently remember little or nothing about the horror they have experienced. Examples include amnesia for suicide attempts, apparent amnesia for criminal acts (including homicides), and amnesia for being raped (Arrigo & Pezdek, 1997). The much more common pattern, however, is that traumatic events lead to the opposite outcome-clear, complete, and vivid recollection, even years after the trauma was experienced (e.g., Bahrick, Parker, Fivush, & Levett, 1998; McNally, 2003; Peace & Porter, 2004; Peterson & Whalen, 2001; Pezdek & Taylor, 2001). Indeed, survivors of the Nazi death camps, rape victims, and soldiers who have been in horrific battles (to name just a few examples) often complain that they remember too much of their terrible experience and wish they remembered less. What produces this uneven pattern? Some authors have suggested that the key lies in the nature of the trauma. For example, one proposal is that repeated trauma is likely to be forgotten, whereas singular traumatic episodes will be remembered (e.g., Terr, 1994). A different proposal is that trauma involving betrayal is likely to be forgotten, whereas trauma of other sorts will be remembered (e.g., Freyd, 1996). However, the evidence favoring either of these views is unpersuasive so far (e.g., Shobe & Kihlstrom, 1997). Instead, the key may lie not in the nature of the traumatic experience, but within the person who is experiencing the trauma. It is likely that people differ (both psychologically and biologically) in how they react to trauma, and this may be why the clear majority of traumatic experiences are remembered extremely well, but others are not. (For discussion of what causes differences among people in how they recall traumatic events, see, for example, Cima, Merckelbach, Nijman, Knauer, & Hollnack, 2002; Deffenbacher et al., 2004; Heim, Owens, Polotsky, &Nemeroff, 1997; McCranie, Hyer, Boudewyns, & Woods, 1992; Morgan et al., 2004; Schacter, 1996; Yehuda, 1997; Yuille & Daylen, 1998.) In addition, many of the cases in which psychogenic amnesia is alleged may involve one or another biomedical factor, so that the amnesia is, in truth, not psychogenic. Thus, for example, in at least some cases, the now-forgotten event occurred while the person was heavily intoxicated (this is true for many of the cases in which murderers allege amnesia for their acts); it may therefore be the alcohol that is disrupting memory. In other cases (e.g., soldiers reporting amnesia for their combat), the individuals may be suffering from sleep deprivation, lack of food, concussions or other head injuries. (For more on these issues, see Kihlstrom, 2005; Kopelman & Kapur, 2001; McNally, 2003; and also the section below on stress.)

4 The Diagnostic and Statistical Manual (American Psychiatric Association, 1994) uses a different term for this sort of memory loss-namely dissociative amnesia, with the proposal that this amnesia results from a set of processes known as dissociative mechanisms. However, we will use the older and more general term, psychogenic amnesia, in order to keep our focus on the phenomenon, rather than on a hypothesized mechanism. As we use the term here, psychogenic amnesia refers only to cases of memory loss, typically associated with traumatic events, which cannot be attributed to the sorts of mechanical damage to the brain (e.g., damage to the hippocampus and surrounding regions) known in other cases to produce amnesia.

94

REISBERG AND HEUER

Just as pressing, we need to ask why trauma much more commonly have opposite effect-namely, the production of memories that are vivid, detailed, and long-lasting? Hypotheses about this point are often cast in terms of the relevant biology and, in particular, the process known as consolidation-a process that takes place across the hours (and perhaps days) following an experience, and which serves, in essence, to cement the memory in place (Hasselmo, 1999; McGaugh, 2000; Meeter & Murre, 2004). In most individuals, bodily arousal serves to promote consolidation, and, indeed, some authors speak of stress as producing an "overconsolidation of the memory trace" (e.g., Pittman & Orr, 1995, p. 81). Once again, however, we need to add a note of caution: traumatic memories do tend to be extraordinarily clear and detailed, but, as we have emphasized from the start, that does not mean these memories are uniformly accurate. At least some traumatic memories (e.g., some soldiers' battle memories) appear to be condensations of multiple episodes, coalesced into one truly horrific recollection. In addition, some traumatic memories seem to merge elements of the actual trauma with elements of what the person feared might happen (or, in some cases, hoped might happen). Thus traumatic memories, like all memories, cannot be understood as a perfect record of past events. (For parallel data in nontraumatic memories, with recollection blurring together various episodes, and mixing historical fact with expectations, fears, and hopes, see Neisser, 1981; for a broader discussion of errors and biases in memory for unemotional events, see Reisberg, 2005.) Thus, the extreme of emotion that we call trauma seems, just like emotion in general, to promote memory accuracy and longevity, but it certainly does not guarantee them. In short, then, the results described in this section are complex, and for forensic purposes this can produce a frustrating state of affairs. A witness who claims to recall her rape in excruciating detail, for example, must be regarded as credible; as we have seen, extremely clear, long-lasting memories for traumatic events are commonly observed. However, errors are still possible even within this incredibly detailed and compelling recollection. In addition, a witness who insists he remembers nothing of his traumatic ordeal is also credible. To be sure, the evidence suggests that this sort of amnesia is rare and is most likely when other factors (alcohol or drug abuse, head injuries, sleep deprivation) are present (Cima et al., 2002). Moreover, a person who displays amnesia for a trauma is likely to show other symptoms as well (e.g., cognitive difficulties that may extend well beyond the problems in recalling the traumatic event itself) and, in many cases, will spontaneously recover the memory, without intervention, within days (McNally, Clancy, & Barrett, 2004). As a result, the pattern of psychogenic amnesia can in some cases be confirmed from other evidence (or disconfirmed, if the person is dissembling). Nonetheless, both patterns-amnesia for trauma, and especially clear memory for trauma-can be documented.

EMOTION AND MEMORY: AN INTERIM SUMMARY

It may be helpful to pause and review the argument so far. Overall, the evidence suggests that emotion tends to promote memory for an event's gist and central materials. This

3.

MEMORY AND EMOTION

95

effect seems widespread-it is evident in the laboratory and outside of it, evident with many different sorts of emotional materials, and evident at a wide range of emotional in~ tensities. Indeed, even at the extreme level of emotion associated with trauma, emotion seems usually to promote memory accuracy and longevity. However, we have already seen reasons to wrap these claims in several layers of cau~ tion. First, we have several times flagged the fact that the data may be altered when we consider other types of emotion, and we will turn to that issue (at last) in the next sec~ tion. Second, we have repeatedly noted that the evidence cannot be understood as indi~ eating that emotion leads to uniformly accurate memories. It does not. No matter what the circumstances, memory errors are possible (and perhaps even likely), and some amount of forgetting will occur as time goes by. Thus, for example, victims of trauma tend to re~ member their horrible experience all too well, but with some errors or gaps in their recall (Payne et al., 2004; also see below); the same is true for people who experience more moderate levels of emotion. Third, the memory benefit produced by emotion applies, as we have discussed, pri~ marily to the gist of the emotional event and to details associated with that gist. In con~ trast, emotion appears to have the opposite effect, working against memory accuracy, for more peripheral aspects of the event-including aspects that, in many criminal cases, may be forensically crucial. This pattern of memory narrowing is often attributed to the emotion itself but, in truth, may be better understood as being caused by the presence of salient visual stimuli, within the event, that capture witnesses' attention. How exactly might attorneys or law enforcement professionals use this information? Quite simply, if a witness was emotional during an event, this should add to our confi~ dence that the witness's report of the event's gist and central materials is correct. (We will, however, add some complications to this claim in the next section. And, in any case, we must bear in mind that emotion is only one of the factors influencing memory accuracy, and so we would also need to consider other factors before trying to assess the risk of error for this, or any, witness.) Conversely, if the witness was emotional, this should decrease our confidence that the witness's report of the event's periphery is correct. Errors about peripheral details are likely (with the errors likely to involve reconstruction from other aspects of the witness's knowledge and expectations), and we also should not be surprised if there are gaps in the witness's recollection of peripheral information. One version of this overall pattern is the weapon~focus effect, with enhanced memory for the weapon (and the hand holding it) and diminished memory for other aspects of the scene (including the face of the person holding the weapon). However, other emotional events, not including a weapon, will yield much the same pattern. We should also note one further way that this information might be used, but here we sound a note of caution: In many legal contests, different witnesses offer different descriptions of how an event unfolded, and it is of enormous interest to know which ac~ count is more credible. One might read the discussion so far as suggesting that the more emotional witness is the one to be trusted, on the ground that he or she received the most benefit from emotion's memory~promoting effects. However, this conclusion is not warranted. A number of studies have compared memory in individuals who reacted more strongly to a stimulus with memory from individuals who reacted less strongly, and

96

REISBERG AND HEUER

these studies have yielded (at best) mixed results. There is certainly not a consistent pattern in which those people who report the most emotion are also those who show the most accurate memory. (For recent summaries, see Deffenbacher et al., 2004; Edelstein et al., 2004; Peace & Porter, 2004.) What underlies this mixed pattern? The problem is likely to arise from the difficulty of comparing emotionality from one individual to the next. The most obvious path to this comparison is via each person's self-report (so that, for example, one person says, "I was extremely upset," whereas another says only, "I was a little worried"). However, this self-report is of questionable value, because the differences may reflect only how people describe their emotions and not the emotions themselves. (Some people are likely to be more extravagant in how they describe their own emotions; others will be more cautious. Thanks to these variations, we cannot interpret the self-reports as directly revealing the level of emotionality.) A different option is to assess emotionality by means of observations of behavior: Is one person crying while the other is not? Is one person visibly pale while the other is not? These indices are likely to be influential for juries (and so the jury is likely to believe, for example, the witness who weeps as she recalls the crime). However, these assessments, once again, are of little value. Among other concerns, people differ in how expressive they are, in showing their emotions, and so we cannot interpret differences in expressiveness as directly revealing differences in emotionality. For these (and other) reasons, one cannot sustain the claim that events are better remembered by those individuals who are more emotional. Some studies suggest this is the case; others do not. However, this is entirely separate from the claim, well-rooted in evidence, that emotional events seem better remembered than neutral ones.

THE EFFECTS OF STRESS

Throughout this chapter, we have spoken of emotion in general terms, as if we needed to make no fine-grained distinctions among types of emotion. The implication, in other words, has been that the memory effects of fear are the same as the effects of anger, which in turn are the same as the effects of relief, and so on. At least some authors, however, challenge this perspective. For example, Levine and Pizarro (2004) argue specifically that the memory effects of anger differ from those of fear, because these two different emotions lead people to focus on (and thus to remember) different aspects of an emotional event. This is obviously an important claim, but the available evidence on this point is so far sparse, and hence we highlight this issue as a point in need of further research. (For a striking exemplar of work in this area, see Levine & Bluck, 2004.) As a related point, a number of authors have argued that we need a different type of theorizing to account for a special subset of emotional events-namely, those that cause stress. Stress is typically defined as a state characterized by high levels of physiological arousal, some sense of being threatened (and, with that, a sense of anxiety), and also a specific biological reaction-namely, activation of the so-called hypothalamic-pituitaryadrenal (HPA) axis, and then a set of hormonal effects that follow from this (cf. Payne

3.

MEMORY AND EMOTION

97

et al., 2004). This is, of course, a state that crime witnesses or victims will often be in, making the effects of stress important for this review. Without question, chronic stress-that is, stress that continues over weeks or longerdoes have a deleterious effect on memory. (This is true even if the chronic stress derives from factors in the person's life independent of the event being remembered.) In fact, the biological mechanisms behind this effect are well documented: As one aspect of the stress response, the adrenal glands release hormones called glucocorticoids, and, at high concentrations, these hormones disrupt the functioning of the hippocampus, a brain structure crucial for the establishment of episodic memories. If the stress continues, long term structural damage to the hippocampus can result. What about the memory effects of acute (short-term) stress? Many studies have examined this issue; some of these studies rely on human participants, others on animal subjects; some studies rely on circumstances that can be considered stressful; others use medications to simulate stress. In most cases, the results suggest that stress undermines memory, so that less is remembered about the stressful event than about a corresponding, but less stressful event (Payne et al., 2004). Two papers in particular, however, have highlighted the memory effects of stress, and it will be instructive to examine these papers in some detail: Morgan et al. (2004) studied a group of United States soldiers who were involved in military survival training; as part of the program, the soldiers had been deprived of food and sleep for 48 hours, and interrogated at length. One day later, the soldiers attempted to identify their interrogators (either from live lineups or photospreads); those who had been exposed to "high stress" interrogation (including physical confrontation) were much less likely to identify their interrogators than those exposed to "low stress" interrogation. The Morgan et al. paper has been widely cited as directly relevant to forensic questions, and the attention received by this paper is easily understood: The stress examined in their study was, after all, genuine, and the main measure concerned the identification of a "perpetrator." In addition, several features of the procedure make the result all the more persuasive: The soldiers were required to maintain eye contact with their interrogators (and punished if they did not), and, throughout the 40-minute interrogation, the instructors were constantly within a 2-5 foot distance from the soldiers. 5 These latter points make the low levels of identification accuracy (30% true-positive responses in the high-stress condition with a live lineup, 62% in the low-stress condition) all the more striking. Another recent study also suggests a memory impairment associated with stress, but this time based on a broader swath of evidence: Deffenbacher, Bornstein, Penrod and McGorty (2004) analyzed data drawn from a number of studies in which participants were under stress; studies were included in their meta-analysis if they "included either measures of cognitive anxiety or physiological activation or both'' (Deffenbacher, personal communication). Across these studies, stress seemed reliably to impair memory, measured

5W e

thank Charles Morgan for providing us with these extra details about the procedure.

98

REISBERG AND HEUER

either in terms of participants' ability to identify the other people who had been present during the stressful event, or in participants' ability to recall "crime-related details." Both the Morgan et al. and Deffenbacher et al. papers seem entirely persuasive, and, in light of these data, perhaps we need to qualify the claims made earlier in this chapter: Plausibly, emotional settings that do not cause stress will lead to an enhancement of memory (at least for the event's "center"), while emotional settings that do involve stress will lead to the opposite effect-impaired memory. Deffenbacher et al. (2004) describe their own meta-analysis in roughly these terms, distinguishing in particular two types of arousal. One type is involved in emotional events that are not stressful, and causes the witness to "orient" toward the stimulus, and thus improves memory. The second type, as they describe it, is produced by emotional events that are stressful, leads to a "defensive response" rather than orienting, and impairs memory. Payne et al. (2004) offer a related proposal for how stress influences memory: They argue that emotion in general involves activation of neural circuits centered on the amygdala, and this promotes memory for the gist or schematic structure of an event. Stress, they suggest, does not interfere with this mechanism, but has its own, separate, effects. Specifically, they argue that stress disrupts the functioning of the hippocampus, a structure that plays a key role in establishing the spatial and temporal setting of an event. Putting these pieces together, they argue that the stressful events (which are, of course, also emotional) will lead to accurate memories for the event's gist (thanks to the amygdala-based mechanism) but poor memories for an event's structure and sensory details (because of the hippocampal mechanism). (For evidence supporting the claim that the amygdala plays a crucial role in gist memory, but not memory for detail, see Adolphs, Tranel & Buchanan, 2005.) These claims are potentially of enormous importance in legal settings, for the simple reason that most crimes will evoke some amount of anxiety, and so most witnesses to crimes will experience some degree of stress. In addition, both the Morgan et al. and Deffenbacher et al. studies focus on a particular type of memory: eyewitness identification. This makes these studies of even greater interest for the legal system. Unfortunately, though, the status of these claims about stress is uncertain, and so the application of these arguments to courtroom settings would be premature. The main problem is that the results reported by Deffenbacher et al. and by Morgan et al. seem inconsistent with some of the data reviewed earlier in this chapter. Specifically, recall our earlier discussion of memory for trauma. Traumatic events surely involve a very high level of stress, and if stress impairs memory, then traumatic events should be poorly remembered. Yet, as we have seen, the opposite is typically the case: Traumatic events are remembered in a consistent fashion despite the passage of many months (or years), and, when checked, many traumatic memories tum out to be impressively accurate in both their gist and in their fine details. This is plainly at odds with both the Deffenbacher et al. and Payne et al. proposals, and seemingly inconsistent with the Morgan et al. findings. How can we reconcile these observations? The answer is by no means certain, but, given the centrality of these issues for eyewitness evidence, it may be useful to count through some of the suggestions that are either in the research literature already, or which flow naturally from the literature. This will, we hope, help to sensitize readers to

3.

MEMORY AND EMOTION

99

the issues and methodological points that are important for understanding the empirical evidence, and will also highlight gaps in the evidence that those in the judicial system should be aware of. One option for dealing with this mixed data pattern is that we might take the Deffenbacher et al. and Morgan et al. data at face value, but seek somehow to overturn the apparently-contrary findings concerned with memory for trauma. This is, for example, the path taken by Payne et al. (2004) who argued that traumatic memories are remembered only in a "fragmented" fashion-as a series of disjointed snapshots, rather than a coherent whole. On this basis, one could maintain the claim that stress in general is disruptive of memory-as reflected (obviously) in the Deffenbacher et al. or Morgan et al. data, but also as reflected in the fragmented quality of traumatic memories. This argument is appealing, but its empirical foundation has been challenged by several studies, with some authors arguing rather explicitly that traumatic memories do not have the fragmented quality alleged by Payne et al. (cf. Kihlstrom, 2005; Peace & Porter, 2004). Thus, we are skeptical that this argument will be productive. Alternatively, one could take the opposite tack-taking the evidence concerned with trauma memory at face value, and seeking to limit the conclusions we draw from the Deffenbacher et al. and Morgan et al. findings. Thus, for example, one could focus on the fact that, in the Morgan et al. study, the soldiers had been sleep- and food-deprived for two days prior to their interrogations. This deprivation state was the same for both the low-stress and high-stress interrogations, and so of course cannot by itself explain the contrast in how the two interrogation types are remembered. Nonetheless, the deprivation may have played a crucial role in shaping the data. Specifically, the biological duress may have potentiated the effects of interrogation stress, so that the impact of the interrogation stress may have been markedly (and perhaps qualitatively) different than it would have been in the absence of this biological vulnerability. On this view, the Morgan et al. data may tell us a great deal about how sleep-deprived individuals react to high- or low-stress, but this may not tell us how healthy, rested individuals would react to the same levels of stress. Said differently, the Morgan et al. result might inform us about the memory effects of stress in the presence of biological hardship, but not about stress in other circumstances. 6 (For more on the interplay among biological/medical duress, stress, and memory, see Kihlstrom, 2005; McNally, 2003; McNally et al., 2004.) A related argument is possible for the Deffenbacher et al. meta-analysis. Deffenbacher et al. note explicitly that whether an event is "stressful" is to a large extent determined by how a person interprets or perceives the event, so that the same event might be highly stressful for one individual, but not at all stressful for some other person. 7 Indeed, this point is crucial for their meta-analysis, because the analysis includes many studies in which the stress manipulation (the sight of a weapon, the threat of an injection, the 6To

make this point concrete, imagine a (hypothetical) study in which the participants were all inebriated, and then exposed either to low or high levels of stress. Such a study would tell us how drunken people respond to stress, but surely we should hesitate to draw conclusions from this about how sober people respond to stress. It seems plausible to us that a similar argument might apply to the Morgan et a!. study. 7We thank Ken Deffenbacher for helping us to understand these important points.

100

REISBERG AND HEUER

viewing of a theft) had a much larger memory effect for those participants who were-for reasons external to the experiment-particularly vulnerable to stress. In some of the studies, this vulnerability was assessed by observers; in other cases, it was assessed by means of standardized personality tests. Across various studies, though, the memory effects of stress were most visible-and, in some studies, only visible-in participants who were anxious, more neurotic, or in some other fashion more vulnerable to the stress. Deffenbacher et al. draw an important message from these findings: Most of the research cited in this chapter (and, indeed, most of the research on the topics we are discussing) has regarded stressfulness or emotionality as a factor inherent in a certain situation; thus we have asked, for example, how people remember "emotional events," or how their memories are shaped by "stressful settings." The data may force us, though, to a different type of analysis, in which we must consider both the situation and the person, acknowledging that whether an event is stressful, or how stressful it is, will depend on the person involved, and so how the event will be remembered will likewise depend on the person involved. Researchers of course understand this point, but nonetheless assume that people have enough in common with each other so that we can make claims based largely on the situation per se, essentially ignoring the differences from one individual to the next as if these provided only minor variations on the theme. Deffenbacher et al. challenge this perspective, and, indeed, their data provide reason to take their challenge seriously. (See Yuille & Daylen, 1998, for a similar argument.) This broad point must be highlighted, therefore, as an issue in need of future research scrutiny. At the same time, though, we note that this point about differences from one individual to the next may place limits on what we can conclude from the Deffenbacher et al. findings. In brief, we have already suggested that the Morgan et al. (2004) data, with soldiers failing to identify their interrogators, may tell us a great deal about the effects of stress when one is already under biological duress, but may tell us appreciably less about the effects of stress when one is not already under some medical strain. By the same token, many of the studies reviewed by Deffenbacher et al. may tell us a great deal about the effects of stress on those who, by their personality or background, are particularly predisposed to suffer from the stress, but may tell us less about the memory effects of stress for the broader population. This suggestion should not be understood as a criticism of the Deffenbacher et al. claims. Indeed, if they are correct, then the sort of claim we are looking for (e.g., a broad assertion about how stress influences memory) may be unattainable. Instead, we may need claims about stress that are 'tuned' to this population or that-people who are anxious overall, or people who are neurotic, and so on. More immediately, though, this reading of the findings on stress may explain the seeming conflict in the research literature. We would suggest (in line, we believe, with Deffenbacher et al.'s overall argument) that perhaps we should not be talking about the memory effects of stress per se. Instead, stress simply does not affect everyone in the same fashion, and seems, in particular, to impair memory largely for those people who are vulnerable to the stress (either for psychological or for biomedical reasons). For more resilient individuals, stress may have little memory effect. Cast in these terms, there is a clear parallel, and not a conflict, with the data on trauma memory, because here too

3.

MEMORY AND EMOTION

101

there are differences among individuals: Some people fail to remember traumatic events (presumably because they are ill-equipped to cope with the trauma) while most people do remember the trauma-accurately and in detail. Thus, in both cases, in the laboratory studies of stress and the field studies of trauma, we are seeing two types of response, and a conflict appears only if we emphasize the impairment side of things in one data set and the improvement effect in the other set. This account seems quite promising to us as a way to think about the various findings we have reviewed. We emphasize, though, that this proposal rests on some amount of conjecture, and, in addition, there are other factors we must consider as we try to understand the relation between the studies of stress and the studies of trauma. For example, it is noteworthy that most trauma memories involve events with considerable autobiographical significance, in contrast, say, to the studies in the Deffenbacher et al. review in which people are remembering a single episode in which they received an inoculation. As a result, traumatic events may be understood within a self-referenced framework, and rehearsed more often, in ways that directly aid memory. 8 Related, it is important that traumatic events often involve familiar people and places, and this, too, may help someone to remember the event. Thus it is possible that memory for traumatic events is, in fact, impaired by the stress inherent in the trauma, but that this disadvantage is offset by other factors in the situation that promote long-term retention. The suggestion, then, is that there may be two reasons why traumatic events are better remembered than one would expect, based on the observation that stress impairs memory. One reason lies in differences among individuals-with the possibility that stress produces little memory impairment for some people, and these are the people who are accurately remembering the traumas they have experienced. The second reason lies in factors external to the stress that may help people in remembering traumatic events. In addition, there is a third factor that we believe should be considered, because it is possible that stress changes what people remember, but not how much they remember. The proposal here begins with a point made early in this chapter-namely, that people are likely to remember what they paid attention to during an event, and almost certainly not going to remember things they paid no attention to. This point in turn requires that we ask what it is within an event that is attended, and we have already suggested that, in some emotional situations, a particular stimulus (such as a weapon or a gaping wound) will catch and hold the witness's attention; psychologists refer to this as "exogenous" control of attention-i.e., control of attention by the stimulus itself. In other settings, though, attention is under "endogenous" control: The witness herself decides what to pay attention to. And, specifically, in a stressful event, a witness may choose to focus on a particular object that is not crucial for the event, largely as a defensive means of not focusing on, or thinking about, the horror of the event itself. This latter possibility-that people in stressful environments deliberately focus their attention on some neutral stimulus-is of course consistent with the suggestion that what is distinctive about stressful events is that they produce a defensive, "avoidance" reaction. 8

Here, too, we acknowledge our debt to Ken Deffenbacher, who drew our attention to these important issues.

102

REISBERG AND HEUER

(Our proposal is also consistent with the suggestion, offered by many clinicians, that stressful events often produce a "dissociative" response.) Moreover, this proposal fits well with anecdotal reports from many crime victims. Women who have been raped, for example, sometimes report vivid memories of their attacker's shoes, or his belt buckle, or some detail in the physical setting. 9 To be sure, these observations are anecdotal, but, taken at face value, are what we would expect on the conjecture we are offering. The idea, then, is that participants in the various studies of stress may have been guarding themselves against the stress by focusing on some aspect of the event other than the faces of the people directly associated with the stress. As a result, memory was improved for these other (attended) aspects, but this was not examined in these studies. And, with attention focused elsewhere, it is unsurprising that face memory was impaired. Amid all of these complexities, let us pause to take stock. The simple fact is that we have an apparent conflict in the data, with some studies showing that stress undermines memory, and other studies showing that trauma (which surely is stressful) is associated with excellent memory. Plainly, some reconciliation of these findings is needed. We have considered several paths toward this reconciliation, but we note that all of the paths rest to some extent on conjecture. It therefore remains to be seen what the best way is to think about memory for stressful events and memory for trauma. With this backdrop, it may be helpful to review the claims that are uncontentious, and separate those from the proposals that are still tentative. First, as we have seen in earlier sections, emotion in the absence of stress seems to improve memory, at least for an event's central aspects; this pattern remains in place even for extremely intense emotion (i.e., trauma). Second, the presence of a weapon (or some other gripping sight) is likely to draw attention to this singular stimulus, with the consequence that memory is impaired for other aspects of the scene. Third, biological duress (in the form of blood loss, food deprivation, exhaustion, a head injury, and so on) is itself likely to impair memory (and other cognitive functions). Fourth, chronic stress is also likely to impair cognition (in part because of the long-term effects of the hormones released in response to stress). Fifth, there is no question that some people seem particularly vulnerable to stress, and their memories are very likely to be impaired by stress. These people can often be identified by means of standardized personality tests, with the vulnerability to stress being associated with the traits that psychologists label trait anxiety and neuroticism. Finally, acute stress does seem to impair memory for the faces of people participating in an event, and at least some other event-related details. These last points are established by both the Deffenbacher et al. and Morgan et al. findings, and so we can almost certainly set aside the "popular conception that most people would never forget the face of a clearly seen individual who had physically confronted them for more than 30 minutes" (Morgan et al., 2004, pg. 274). 9In other cases, rape victims report a specific effort toward focusing on the perpetrator's face, in order to aid the prosecution later. We believe that this heroic effort is rare, but, in any case, it is consistent with the point that we are making, with the victim choosing to focus attention in a specific way, with obvious consequences for memory.

3.

MEMORY AND EMOTION

103

What remains uncertain, however, is how best to explain the full pattern of these data, and how exactly to fit together the findings on stress and the findings on trauma. We have considered various claims, but obviously more research on this topic is urgently needed.

EMOTION AND MEMORY RETRIEVAL

Much of our discussion so far has centered on the creation of memories for emotional events. But, of course, victims and witnesses need to do more than just create the mem~ ory; they also need to be able to retrieve the memory later on. Otherwise, the creation of the memory would be akin to writing books that could not be read, or placing money in a bank account from which no withdrawals are possible. How does emotion influence subsequent retrieval? Our answer to this question begins with the fact that retrieval of information often requires some work or effortboth to search through memory, hunting for the relevant material, and then to make decisions about the material that comes to mind. Specifically, if a police officer's ques~ tion calls some fact to mind, the witness can still ask: "Why did this fact come to mind? Was the fact part of the to~be~remembered event? Or was the fact merely something I thought about, in relation to the event? Or perhaps something I imagined, in thinking back to the event later on?" These are referred to as source~monitoring decisions (because they concern the source of the remembered information); they are often made uncon~ sciously, but are still a crucial determinant of how the remembered event will be reported (e.g., Johnson, Hashtroudi, & Lindsay, 1993). In addition, memories arrive in one's thoughts with varying degrees of conviction ("I'm certain I recall this" vs. "I have a vague recollection ... ").As a result, one must ask, in essence, ''Am I certain enough to voice this recollection?" This, too, can sometimes be an effortful decision. (For more on these judg~ ments, see Pansky, Koriat, & Goldsmith, 2005). Emotion is, of course, a significant distractor in many circumstances and so poten~ tially can draw resources away from the processes just described. This may, in tum, help provide the explanation for two phenomena. First, a number of studies indicate that emotionality is associated with an increased tendency toward false~positive responsesan increased tendency to say that something is familiar when in truth it is not (Payne, Nadel, Allen, Thomas, & Jacobs, 2002; Windmann & Kutas, 2001). One (somewhat speculative) account of this pattern is that a focus on one's own emotion impairs source monitoring (Johnson, Nolde, & De Leonardis, 1996); this leaves witnesses less able to make accurate discriminations between real memories and fictions and hence more re~ liant on mere biases (in this case, a bias toward claiming familiarity with ideas, or faces, that are actually not familiar). If this proposal is confirmed, it has powerful implications for eyewitness memory. (For example, in the context of an eyewitness identification, a false~positive response is equivalent to identifying an innocent person as guilty.) Second, a larger number of studies suggest more directly that emotion can disrupt memory retrieval. This is reflected in the fact that emotion leads to poorer memory per~ formance if memory is tested immediately or at a very short delay, despite the fact that

104

REISBERG AND HEUER

emotion improves memory if memory is tested at longer delays (e.g., Kleinsmith & Kaplan, 1963, 1964; for a summary of related work, see Revelle & Loftus, 1992). This happens presumably because, with immediate testing, the emotion associated with the stimulus input is still present and disrupts retrieval. After a delay, the initial emotion has dissipated and so no longer has an effect on retrieval. This allows the encoding advantage, produced by emotion, to show itself. The suggestion that flows from these observations is that law enforcement would be best served by allowing witnesses to calm down, after an emotional event, before trying to interview them. In that way, the witness's emotion would interfere with neither the retrieval effort nor with source monitoring. However, a crucial complication must be added to this point. The passage of time is, by itself, a powerful factor in determining memory performance, and, on this basis, law enforcement officers would be well advised to interview witnesses as soon as possible after a crime, to minimize the retention interval and the memory loss associated with longer intervals. Thus, considerations of retention interval argue against delaying any interviews with witnesses, at the same time that the desirability of having the witness be calm and collected during the interview argues, in some circumstances, in favor of delay. How these points should be weighed against each other must be decided on a case-by-case basis. 10 EMOTION AND MEMORY RETRIEVAL: REPRESSION

There is one other (and highly controversial) way in which emotion is alleged to influence memory retrieval-through the mechanism of repression (or, in some settings, dissociation). The idea here is that the emotion associated with some events is so intense and so negative that it evokes mechanisms of self-protection. The result is an apparent amnesia for the emotional event-not because it is forgotten in any literal sense, but because the painful memories are denied access to consciousness. This is not the place to delve deeply into this difficult issue, but most scientific investigators agree that the evidence favoring the notion of repression is at best ambiguous 10 We might mention another potential complication here, but it is one easily set aside. A substantial body of research indicates that memory performance is best if someone is in the same psychological state during retrieval as they were during the remembered event (e.g., Balch, Bowman, & Mohler, 1992; Grant et al., 1998; Hershkowitz, Ohrbach, Lamb, Sternberg, & Horowitz, 2002; Smith, Glenberg, & Bjork, 1978; etc.), and this seems to suggest that witnesses and victims should be interviewed while they are still emotional, so that this aspect of retrieval is matched to the context of the initial episode. This obviously is contrary to the suggestion we are making here-that witnesses should be allowed to calm down before any interview. The resolution of this conflict, however, is straightforward: The evidence does say that retrieval is best if some aspects of the retrieval context are matched to the aspects of the learning context. It is not clear, though that emotional state is one of those aspects, and so the evidence is mixed on whether there is a retrieval benefit from being in the same emotional state that one was in during learning. Some studies have demonstrated mood dependency effects, such that one is better able to recall happy events if one is again happy during recall, better able to recall sad events if one is again sad during recall, and so on. But many other studies have failed to reproduce this finding, and the issue of when this mood dependency is observed, and when not, remains a matter of controversy (e.g., Balch, Myers, & Papotto, 1999; Eich, 1995; Kwiatkowski & Parkinson, 1994; Varner & Ellis, 1998; etc.). Thus, the value of context reinstatement (which is itself undeniable) turns out not to be an argument for interviewing witnesses while they are still emotional.

3.

MEMORY AND EMOTION

105

(for discussion, see McNally, 2003; Read & Lindsay, 1997). Perhaps more important, we have already discussed the fact that extremely emotional, even traumatic, memories are generally vividly remembered, which is the opposite of what we might expect from a mechanism like repression. This remains true even in cases in which the emotional event involves the sort of circumstances that seem most likely to trigger defense mechanisms. For example, several widely-cited studies have examined children's memory for a medical procedure that involved anxiety-producing and uncomfortable touching of the child's genitals; this experience, plausibly ripe for the triggering of defense mechanisms, was well remembered by most of the children (Brown et al., 1999; Ornstein, Baker-Ward, Myers, Principe, & Gordon, 1995; Pipe et al., 1997; Quas et al., 1999). Nonetheless, some of the phenomena that are marshaled in support of repression are real, even if rare. What is at issue, however, is how to explain these phenomena. For example, it is certainly possible to experience a deeply emotional event and not remember it. But there is no reason to assume that this reflects the operation of repression. Instead, we have already mentioned other factors (alcohol or substance abuse, head injury, sleep deprivation) that may account for some of many of these cases of apparent amnesia. In addition, some traumatic events are forgotten because the individual was very young at the time of the event and so was subject to the widespread pattern of childhood amnesia (cf. Shobe & Kihlstrom, 1997). In yet other cases, the event has simply been forgotten because of the passage of time (cf. Read & Lindsay, 2000). And so on. As we have mentioned, these are the exceptional cases, not the typical pattern; much more commonly, horrific events are remembered all too well. But even in these cases in which the trauma is forgotten, our point is that the forgetting can be understood within the framework we have already sketched in this chapter, with no need for any appeals to "repression." In the same fashion, consider claims made about recovered memories-that is, cases in which a person reports no recollection of the target event for many years, followed by a recovery of this long-lost and painful memory (often through the help of a therapist). These cases are commonly cited in support of the repression idea, with the argument that the repression has been lifted with the therapist's aid, allowing the long-hidden memory to come into view. For these cases, too, we would argue that the phenomenon itself should not be discounted, but can be understood in conventional terms (i.e., with no reference to repression or dissociation). For example, in some cases of (allegedly) recovered memory, the memory has probably never been lost; recovery merely refers to the person's new-found willingness to talk about this always remembered but never discussed episode. This willingness may emerge, for some individuals, in the supportive environment of therapy; it may emerge in other cases because the person becomes convinced, through other sources, that the memory will be taken seriously and so can, at last, be voiced. In still other cases, the recovery is probably best understood as a recategorization of the relevant memory: "I always knew that he did X, but until recently, I didn't think there was anything blameworthy in what he did, and it's only now, when I realize the action is culpable, that I'm talking about it." In yet other cases, the recovery needs to be understood differently, because someone is convinced he was amnesic about an event at some prior time (so that the memory, now

106

REISBERG AND HEUER

accessible, has been recovered), but what is being misremembered is, remarkably, the fact of amnesia itsel£ That is, the person did remember the event all along, and we can document this with other evidence, but the person mistakenly insists that the opposite is the case, and that there was a period of amnesia during which the memory was lost, followed by the memory's recovery (Schooler, Bendriksen, & Amadar, 1997). Finally, it is simply the case that memories can be inaccessible for some time, but then brought back to mind by some retrieval cue. This is not in any fashion evidence for repression; it is instead a reflection of the fact that some forgetting involves retrieval failure (an inability to find the event within the vast storehouse of memory), a pattern that applies to emotional memories just as it applies to any other memory (cf. Reisberg, 2005). The previous three paragraphs might be understood as a bit of academic fussinessfocused on a theoretical debate about what it is that causes memory loss and recovery, and, with that, the question of whether repression exists (or whether, as an alternative, the alleged phenomena of repression can be accommodated within the conventional frameworks of our field). In truth, however, what is at stake here is not just theory, but is instead of considerable pragmatic importance, because a better understanding of the mechanisms behind memory loss and recovery can help us determine the cases in which these memory claims are more plausible and the cases in which they are less so. Thus, for example, repression is allegedly capable of "hiding" the memory of an isolated episode, leaving other memories and other cognitive functions largely untouched. But, as we have seen, there is reason to be skeptical about all claims of repression, and thus reason to be skeptical about this particular expectation for memory loss. If, instead, we consider memory loss in terms of what else we know about psychogenic amnesia, we come to different expectations: Psychogenic amnesia typically emerges among other cognitive and emotional difficulties; it does not simply 'black out' a single episode, with no other effects. An allegation of memory loss that fits this profile, therefore, is credible and should perhaps be taken seriously. A suggestion of memory loss that does not at all fit this profile must be interpreted with considerable care. In the same vein, retrieval failure is most likely for distant memories; it is unlikely for events from the last day or two. Therefore, once we realize that retrieval failure (and not repression) is the likely cause of memory loss, we can use this fact about retrieval failure as a ground for deciding whether a particular case of alleged memory loss is plausible or not. Or, as a final illustration, retrieval-failures are often reversed when some longabsent cue or reminder enters the scene. (A clear example would be the flood of memories that becomes available when one returns to one's home town, and all the reminders it contains, after an absence of many years.) A memory "recovery" that fits this pattern is therefore credible; in contrast, a memory that is "recovered" with no new memory cues provided needs to be viewed with caution. In these and other ways, then, we can use a more refined understanding of memory "loss" and "recovery" as a means of assessing the plausibility of individual cases, and plainly some "recovered memories" will turn out to be credible, and others will not. Finally, it is a troubling yet undeniable fact that many recovered memories emerge in circumstances that contain factors known to produce false memories, raising the possibility that the recovered memories themselves might in some cases be false. For example,

3.

MEMORY AND EMOTION

107

recovered memories often involve distant events (i.e., years ago), often emerge after re~ peated suggestions that the remembered events did occur, often emerge after the person has engaged in exercises ("memory work") designed to promote recall, such as trying to visualize the remembered event, and so on. Each of these factors on its own has been documented as a powerful source of memory error (see Chapters 8, 9, and 10, this vol~ ume), and so the package of factors, as a whole, produces a substantial risk of error, thus demanding considerable wariness in our interpretation of these memories. In summary, then, recovered memories-like any memories-are open to error and need to be interpreted with caution. Indeed, the caution should be large, in light of the error~promoting factors that are often associated with the emergence memories. How~ ever, this must not be read as a blanket assertion that all recovered memories are false. Emotional events can be entirely forgotten. Events can be forgotten but then later re~ membered (e.g., because of retrieval failure). And, without question, motivation and self~service can play a role in shaping what we remember. (If, for example, we manage never to think about an event, because such thoughts would be painful, this is tanta~ mount to never rehearsing the memory, and this will speed the processes that contribute to forgetting.) On these grounds-and even with the need for caution-some of the basic facts about recovered memory are plausible (at least in some cases). Hence these cases should be regarded with some skepticism but certainly not dismissed out of hand.

EMOTION AND MEMORY IN CHILDREN

With only a few exceptions, the evidence described so far comes from studies-either in the laboratory or in the field-of adults. But, of course, children are also witnesses and victims of crimes, including crimes of abuse. This invites us to ask: Can the claims of~ fered so far be applied to children? Or are children's memories influenced by emotion in a fashion different from that found with adults? The answer to this question is straightforward. A number of studies have examined children's memory for emotional materials and compared this with their memory for neutral materials. Like adults, children generally show a memory advantage for the emo~ tional materials. In addition, and still like adults, children's memory for emotional events so that (for example) distressing medical procedures, sig~ seems to be rather long~lived, nificant injuries, and even abuse have been documented as well remembered years after the event. The studies examining the memory effects of stress also yield similar results for children and for adults; this is clear, for example, in the Deffenbacher et al. (2004) meta~analysis. Finally, and still like that of adults, children's memory for these emo~ tional experiences is best for the event's gist and central details; these are remembered more accurately than peripheral details (for an overview, see Edelstein et al., 2004). In addition, it should be noted that research on children's memory reliably shows an effect of age-older children are able to remember more and to report more clearly than younger children. This result, too, is observed in studies of children's emotional remem~ bering and may interact with emotion in some cases. For example, one study examined children's memory for stories that varied in their emotionality (Davidson, Luo, & Burden,

108

REISBERG AND HEUER

2001). Overall, children remembered more of the emotional stories than the neutral stories. Overall, older children remembered more than younger children. However, these effects interacted with each other: with the neutral stories, the older children had a clear memory advantage in comparison with the younger children. With emotional stories, this contrast was smaller, with the suggestion that the memory benefits of emotionality offset to some extent the disadvantages ordinarily shown by the younger children.

DOES EMOTIONALITY SIGNAL AN ACCURATE MEMORY?

Throughout this chapter we have examined research that takes as its starting point a particular emotional event and asked how this event will be remembered. In most criminal cases, however, the situation is different: The case begins with a memory of an event (or even an alleged event) and seeks to ask what the event must have been like, in light of the memory report. Rather than make predictions, based on the event, we seek to make retrospective inferences about the event, based on the memory. How should such inferences proceed? To make this concrete, imagine that a witness reports that he felt intense anger during a crime; perhaps the witness becomes angry again, merely in thinking back to the crime. Does this reported (and displayed) emotion indicate that the memory is trustworthy, and more likely to be accurate than a memory that contains no such emotionality? After all, if the witness is correct in reporting this emotion, apparently felt during the original event, then the arguments reviewed earlier in this chapter apply to this case: the witness was emotional during the event and therefore should have the memory advantages associated with emotionality. On this basis, we might well decide that the memory can be trusted and that the witness's report probably does provide a complete and correct description of the target event. Unfortunately, however, this conclusion would not be warranted, for several reasons. One is that some memories of emotional events can be entirely false, leading to reports of events that never happened at all. This is because techniques for implanting memories (via leading questions or other suggestive procedures) can create memories for emotional events either in adults or in children. Thus, in several studies, young children have been led to believe that a man put something "yucky in their mouths" (Poole & Lindsay, 1995, 2001); in other studies, children have been led to recall that someone touched their private parts (Bruck, Ceci, Francoeur, & Barr, 1995; Goodman, Bottom, Schwartz-Kenney, & Rudy, 1991). Similarly, adults have been led, through experimenters' suggestion, to remember frightful episodes, such as having been attacked by a dog (Porter, Yuille, & Lehman, 1999); shameful episodes, such as breaking a school rule and being punished for it (Lindsay, Hagen, Read, Wade, & Garry, 2004); or painful episodes, such as having been harassed, as a child, by a bully (Mazzoni, Loftus, Seitz, & Lynn, 1999). Or, in a very different vein, many individuals report vivid (and often terrifying) recollections of participation in Satanic abuse, or of having been abducted by aliens, even though, by

3.

MEMORY AND EMOTION

109

most measures, the plausibility of these memories is extremely low, and so the likelihood that these are false memories is very high (cf. McNally et al., 2004). The various examples just cited differ, of course, in many particulars. What the examples have in common, however, is that individuals are reporting, in each case, memories that we believe to be false but are also reporting memories for emotional events. In some cases, the emotionality is indicated by the content of these memories; in other cases, the participants themselves report the events as having been emotional. In all cases, however, it is clear that we cannot take the emotionality of the memory as an indicator of the memory's veracity. Second, even when we are sure a person is recalling an actual episode and sure the episode was emotional, this does not mean we can take the memory at face value-that is, we cannot assume the accuracy of the report. As we have argued throughout this chapter, emotion does seem to promote memory encoding and does slow the process of forgetting. However, we have repeatedly noted that this does not mean that emotion guarantees complete encoding or eliminates forgetting. Emotional memories can have er~ rors and gaps in them, and the number of these grows as time goes by. Indeed, at least some researchers have suggested that emotionality actually increases the likelihood of certain sorts of memory error (Heuer & Reisberg, 1990), but whether this turns out to be correct or not, it is clear that emotion is, by itself, no assurance of the veracity of memory. Third, it is also important that people sometimes misremember their own emotions. This is particularly likely if someone's current appraisal of a past event is different from their appraisal at the time the event occurred. In this case, the person's recollection of their original emotion is often biased by their current assessment and, more specifically, biased in a fashion that brings the (remembered) original emotion into better agreement with the current assessment (e.g., Levine, 1997; Levine, Prohaska, Burgess, Rice, & Laul~ here, 2001). Thus (for example) if an event did not make someone angry at the time it occurred, but now, in retrospect, it makes them angry, the person is likely to remember his or her reaction at the time as being angrier than it actually was. With this context, it is possible that an event that is now recalled as being emo~ tional may not have been emotional at the time that it occurred. In this case, the re~ membered event would not have benefited from the various effects we have described in this chapter, because most of these effects involve the impact of emotion-experienced during the event itself-on the encoding and consolidation of that event. If the event was not emotional at the time it occurred (whether it is now recalled in emotional terms or not), then these effects are irrelevant. (For more on the contrast between emotional~ ity during an event and what we might call "retrospective emotionality," see the com~ parison by Tollestrup et al. (1994) of memory in victims of robbery and memory in vic~ tims of fraud. In the fraud cases, the victims did not know a crime was being committed until sometime afterward, when the fraud came to light.) Worse, a change in the emotional appraisal of an event can, in some circumstances, impair memory accuracy. To see why, imagine an event that was not considered note~ worthy when it occurred, but which was later deemed consequential. One example might be an unremarkable glimpse of an individual visiting a store, followed by a discovery that

110

REISBERG AND HEUER

the individual was actually casing the store for a subsequent robbery. A different example might be a minor comment made by a child, with that comment later becoming far more important in the context of, say, allegations of child abuse. In cases like these, there is little reason to have encoded the initial event with great care and little reason to have paid special attention or to have rehearsed the event. Likewise, in these cases there was no arousal in place, during the event, to promote memory consolidation. As a result, little information about the event will be lodged in memory. Yet, after the fact, there is ample reason to want to recall the event, and this will spur efforts at memory reconstruction. Of course, without much initial encoding, information in memory will provide slim support for this reconstruction, leaving a large potential for schema-based inferences (which may or may not be correct) and intrusions from other, related episodes. For all of these reasons, investigators should be wary when a witness's emotional appraisal of a target event has changed-and all the more so if the target event seemed inconsequential at the time but was later judged to be more important. Likewise, investigators should be alert to the fact that the emotion that matters for memory is the emotion experienced at the time of the event. In many cases, there is little question about what this emotional state was. (If the witness was a victim of an armed robbery, for example, one can plausibly assume that the witness was at least somewhat afraid, and so on.) In other cases, we have only the witness's recollection that he or she was emotional at the time of the event, and this recollection (like any recollection) must be interpreted with caution.

CONCLUSIONS

The argument in this chapter has grown complex, and we wish we could provide a more straightforward account of the data. The fact is, though, that the data are messy, and for that reason perhaps less useful in forensic settings than anyone would wish. Indeed, in light of this complexity, part of our agenda in the chapter has been methodological, to help readers understand how arguments in this domain unfold. Even acknowledging these points, some aspects of the data are clear. Emotion promotes memory for an event's central materials, but emotional situations also make people less observant-i.e., less likely to notice, and less likely to recall, information that is more peripheral in an event. These claims seem to hold across a wide range of emotional intensities, so that, for example, traumatic events seem in most cases to be accurately remembered for many months (and perhaps years) after the traumatic event. At the same time, however, stress appears to be associated with a memory impairment, although it remains unclear whether this effect influences how much people remember, or instead what they remember. It also is unclear whether stress has these effects in general, or only in those individuals particularly vulnerable to stress for biomedical or psychological reasons. At many junctures, our claims about emotion have hinged on what it is that someone pays attention to within an emotional event, and, with that, what we count as the "center" of the episode. In some settings, the categorization of an event's aspects as "central'' or "peripheral" will be straightforward. If, for example, an event includes a weapon,

3.

MEMORY AND EMOTION

111

or a bloody wound, this is likely to seize the witness's or victim's attention, and thus to become the (well-remembered) center of the event. Likewise, if a crime involves a small number of perpetrators, it is their actions, and not the actions of other bystanders, that are likely to be central. In other cases, though, the event's "center" is determined by the witnesses themselves; as we noted earlier, attention is under both exogenous and endogenous controls. This latter fact-the influence of the witness's intentions or strategies, in choosing what to pay attention to-obviously complicates any efforts toward identifying the event's center, and thus in identifying what about the event will be remembered. In all situations, though, an attorney or law enforcement professional is well-served by asking what the witness was likely to be paying attention to. For example, did the witness have reason to look at, and think about, the criminal's face? Did the witness have reason to look at some other aspect of the scene? These questions are always pertinent for eyewitness memory, thanks to the logic, mentioned early on, that people remember what they pay attention to. However, these questions take on extra importance when emotion is on the scene, both because emotion makes it more likely that the attended aspects of the situation will be remembered, and less likely that other qaaspects will be.

ACKNOWLEDGMENT

We wish to express our gratitude to Ken Deffenbacher for his thoughtful commentary on many aspects of this chapter. We have acknowledged within the text several specific points that Ken called to our attention, but Ken's influence improved many other points in the chapter as well.

REFERENCES Adolphs, R., Tranel, D., & Buchanan, T. W. (2005). Amygdala damage impairs emotional memory for gist but not details of complex stimuli. Nature Neuroscience, 8, 512-518. American Psychiatric Association. (1994). Diagnostic and statistical manual for mental disorders (4th ed.). Washington, DC: Author. Arrigo, J. M., & Pezdek, K. (1997). Lessons from the study of psychogenic amnesia. Current Directions in Psychological Science, 6, 148-152. Bahrick, I., Parker, J., F., Fivush, R., & Levitt, M. (1998). The effects of stress on young children's memory for a natural disaster. Journal of Experimental Psychology: Applied, 4, 308-311. Balch, W., Bowman, K., & Mohler, L. (1992). Music-dependent memory in immediate and delayed word recall. Memory & Cognition, 20, 21-28. Balch, W. R., Myers, D. M., & Papotto, C. (1999). Dimensions of mood in mood-dependent memory.]ournal of Experimental Psychology: Learning, Memory and Cognition, 25, 70-83. Behrman, B. W., & Davey, S. L. (2001). Eyewitness identification in actual criminal cases: An archival analysis. Law and Human Behavior, 25,475-491. Brown, D. A., Salmon, K., Pipe, M.-E., Rutter, M., Craw, S., & Taylor, B. (1999). Children's recall of medical experiences. Child Abuse & Neglect, 23, 209-216. Bruck, M., Ceci, S. J., Francoeur, E., & Barr, R. J. (1995). "I hardly cried when I got my shot!": Influencing children's reports about a visit to their pediatrician. Child Development, 66, 193-208.

112

REISBERG AND HEUER

Buchanan, T. W., & Adolphs, R. (2004). The neuroanatomy of emotional memory in humans. In D. Reisberg & P. Hertel (Eds.), Memory and emotion (pp. 42-75). New York: Oxford University Press. Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events. Memory & Cognition, 20, 277-290. Cahill, L., & van Stegeren, A. (2003). Sex-related impairment of memory for emotional events with B-adrenergic blockade. Neurobiology of Learning and Memory, 79, 81-88. Canli, T., Zhao, Z., Brewer, J., Gabrieli, J.D. E., & Cahill, L. (2000). Event-related activation in the human amygdala associates with later memory for individual emotional response. Journal of Neuroscience, 20, RC99. Christianson, S.-A. (1992). Emotional stress and eyewitness memory: A critical review. Psychological Bulletin, 112, 284-309. Christianson, S.-A., & Hubinette, B. (1993). Hands up! A study of witnesses' emotional reactions and memories associated with bank robberies. Applied Cognitive Psychology, 7, 365-379. Cima, M., Merckelbach, H., Nijman, H., Knauer, E., & Hollnack, S. (2002). I can't remember, Your Honor: Offenders who claim amnesia. German Journal of Psychiatry, 5(3), 24-34. Conway, M., Anderson, S., Larsen, S., Donnelly, C., McDaniel, M., McClelland, A. G. R., et al. (1994). The formation of flashbulb memories. Memory & Cognition, 22, 326-343. Cooper, B.S., Kennedy, M.A., Herve, H., & Yuille, J. C. (2002). Weapon focus in sexual assault memories of prostitutes. International]ournal of Law and Psychiatry, 25, 181-191. Crombag, H. F. M., Wagenaar, W. A., & van Koppen, P. J. (1996). Crashing memories and the problem of'source monitoring.' Applied Cognitive Psychology, 10, 95-104. Daubert v. Merrell Dow Pharamaceuticals, Inc., 509 U.S., 113 S. Ct. 2786 (1993). Davidson, D., Luo, Z., & Burden, M. J. (2001). Children's recall of emotional behaviours, emotional labels, and nonemotional behaviours: Does emotion enhance memory? Cognition & Emotion, 15, 1-26. Deffenbacher, K. (1994). Effects of arousal on everyday memory. Human Performance, 7, 141-161. Deffenbacher, K., Bornstein, B. H., Penrod, S. D., & McGorty, E. K. (2004). A meta-analytic review of the effects of high stress on eyewitness memory. Law and Human Behavior, 28, 687-706. Dietrich, D. E., Waller, C., Johannes, S., Wieringa, B. M., Emrich, H. M., & Muente, T. F. (2001). Differential effects of emotional content on event-related potentials in word recognition memory. Neuropsychobiology, 43, 96-101. Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the organization of behavior. Psychological Review, 66, 183-201. Edelstein, R. S., Alexander, K. W., Goodman, G. S., & Newton, J. W. (2004). Emotion and eyewitness memory. In D. Reisberg & P. Hertel (Eds.), Memory and emotion (pp. 308-346). New York: Oxford University Press. Eich, E. (1995). Searching for mood dependent memory. Psychological Science, 6, 67-75. Fazey, J. A., & Hardy, L. (1988). The inverted-U hypothesis: A catastrophe for sport psychology. British Association for Sports Sciences Monograph No. 1. Leeds, England: National Coaching Foundation. Freyd, J. J. (1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge, MA: Harvard University Press. Goodman, G. S., Bottom, B. L., Schwartz-Kenney, B., &Rudy, L. (1991). Children's testimony about a stressful event: Improving children's reports. Journal of Narrative Life History, 1, 69-99. Grant, H. M., Bredahl, L. C., Clay, J., Ferrie, J., Groves, J. E., McDorman, T. A., & Dark, V. J. (1998). Context-dependent memory for meaningful material: Information for students. Applied Cognitive Psychology, 12, 617-623. Greenberg, D. L. (2004). President Bush's false "flashbulb" memory of 9/11/01. Applied Cognitive Psychology, 18(3), 363-370.

3.

MEMORY AND EMOTION

113

Gross, J. J. (2002). Emotion regulation: Affective, cognitive, and social consequences. Psychophysiology, 39, 281-291. Hamann, S. (2001). Cognitive and neural mechanisms of emotional memory. Trends in Cognitive Science, 5, 394-400. Hasselmo, M. E. (1999). Neuromodulation: Acetylcholine and memory consolidation. Trends in Cognitive Science, 6, 351-359. Heim, C., Owens, M., Polotsky, P., & Nemeroff, C. (1997). Endocrine factors in the pathophysiology of mental disorders. Psychopharmacology Bulletin, 33, 185-192. Hershkowitz, 1., Orbach, Y., Lamb, M. E., Sternberg, K. J., & Horowitz, D. (2002). A comparison of mental and physical context reinstatement in forensic interviews with alleged victims of sexual abuse. Applied Cognitive Psychology, 16,429-441. Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events: The accuracy of remembered minutiae. Memory & Cognition, 18, 496-506. Johnson, M., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin,

114,3-28.

Johnson, M., Nolde, S. F., & De Leonardis, D. M. (1996). Emotional focus and source-monitoring. Journal of Memory and Language, 35, 135-156. Jones, E. B., O'Gorman, J. G., & Byrne, B. (1987). Forgetting of word associates as a function of recall interval. British Journal of Psychology, 78, 79-89. Keltner, D., & Gross, J. J. (1999). Functional accounts of emotions. Cognition & Emotion, 13, 467-480. Kihlstrom, J. F. (2005). Trauma and memory revisited. In Uttl, B., Ohta, N., &Siegenthaler, A. L. (Eds.), Memory and Emotions: Interdisciplinary Perspectives. New York: Blackwell. (In press.) Kim, J. J., & Diamond, D. M. (2002). The stressed hippocampus, synaptic plasticity and lost memories. Nature Reviews Neuroscience, 3, 453-462. Kleinsmith, L. J., & Kaplan, S. (1963). Paired associate learning as a function of arousal and interpolated interval. Journal of Experimental Psychology, 65, 190-193. Kleinsmith, L. }., & Kaplan, S. (1964). Interaction of arousal and recall interval in nonsense syllable paired-associate learning. Journal of Experimental Psychology, 67, 124-126. Kopelman, M. D., & Kapur, N. (2001). The loss of episodic memories in retrograde amnesia: Single-case and group studies. In A. Baddeley, J. Aggleton, & M. Conway (Eds.), Episodic memory: New directions in research (pp. 110-131). New York: Oxford University Press. Kramer, T., Buckhout, R., & Eugenio, P. (1990). Weapon focus, arousal and eyewitness memory: Attention must be paid. Law and Human Behavior, 14, 167-184. Kwiatkowski, S., & Parkinson, S. (1994). Depression, elaboration, and mood congruence: Differences between natural and induced mood. Memory & Cognition, 22, 225-233. Laney, C., Campbell, H., Heuer, F., & Reisberg, D. (2004). Memory for thematically-arousing events. Memory & Cognition, 32, 1149-1159. LeDoux, J. (1996). The emotional brain. New York: Simon &Schuster. Levenson, R. W. (1994). Human emotions: A functional view. In P. Ekman & R. J. Davidson (Eds.), The nature of emotion: Fundamental questions. New York: Oxford University Press. Levine, L. J. (1997). Reconstructing memory for emotions. Journal of Experimental Psychology: General, 126, 165-177. Levine, L. J., Prohaska, V., Burgess, S. L., Rice, J. A., & Laulhere, T. M. (2001). Remembering past emotions: The role of current appraisals. Cognition & Emotion, 15,393-417. Levine, L. J., & Bluck, S. (2004). Painting with broad strokes: Happiness and the malleability of event memory. Cognition and Emotion, 18, 559-574. Levine, L. J., & Pizarro, D. (2004). Emotion and memory research: A grumpy overview. Social Cognition, 22, 530-554. Lindsay, D. S., Hagen, L., Read, J.D., Wade, K. A., & Garry, M. (2004). True photographs and false memories. Psychological Science, 15, 149-154.

114

REISBERG AND HEUER

Loftus, E., Loftus, G., & Messo, J. (1987). Some facts about "weapon focus." Law and Human Behavior, 11, 55-62. Maass, A., & Kohnken, G. (1989). Eyewitness identification. Law and Human Behavior, 13, 397-408. Mazzoni, G. A. L., Loftus, E. F., Seitz, A., & Lynn, S. J. (1999). Changing beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13, 125-144. McCranie, E. W., Hyer, L.A., Boudewyns, P. A., & Woods, M.G. (1992). Negative parenting behavior, combat exposure, and PTSD symptom severity: Test of a person-event interaction model.]oumal of Nervous & Mental Disease, 180,431-438. McGaugh, J. L. (2000). Memory-a century of consolidation. Science, 287, 248-251. McGaugh, J. L. (2003). Memory and emotion: The making of lasting memories. New York: Columbia University Press. McNally, R. J. (2003). Remembering trauma. Cambridge, MA: Harvard University Press. McNally, R. J., Clancy, S. A., & Barrett, H. M. (2004). Forgetting trauma? In D. Reisberg & P. Hertel (Eds), Memory and emotion (pp. 129-154). New York: Oxford University Press. Meeter, M., & Murre, J. M. J. (2004). Consolidation of long-term memory: Evidence and alternatives. Psychological Bulletin, 130,843-857. Morgan Ill, C. A., Hazlett, G., Doran, A., Garrett, S., Hoyt, G., Thomas, P., Baranoski, M., & Southwick, S. (2004). Accuracy of eyewitness memory for persons encountered during exposure to highly intense stress. International Journal of Law and Psychiatry, 27, 265-279. Neisser, U. (1981). John Dean's memory: A case study. Cognition, 9, 1-22. Neisser, U., & Harsch, N. (1992). Phantom flashbulbs: False recollections of hearing the news about Challenger. In E. Winograd & U. Neisser (Eds.), Affect and accuracy in recall: Studies of "flashbulb" memories (pp. 9-31). Cambridge: Cambridge University Press. Nickerson, R. A., & Adams, M. J. (1979). Long-term memory for a common object. Cognitive Psychology, 11, 287-307. Latley, K., &Jenkins, J. M. (1996). Understanding emotions. Cambridge, MA: Blackwell. Ornstein, P. A., Baker-Ward, L., Myers, J., Principe, G. F., & Gordon, B. N. (1995). Young children's long-term retention of medical experiences: Implications for testimony. In F. E. Weinert & W. Schneider (Eds.), Memory performance and competencies: Issues in growth and development. Mahwah, NJ: Erlbaum. Ost, J., Vrij, A., Costall, A., & Bull, R. (2002). Crashing memories and reality monitoring: Distinguishing between perceptions, imaginations and "false memories." Applied Cognitive Psychology, 16(2), 125-134. Pansky, A., Koriat, A., & Goldsmith, M. (2005). Eyewitness recall and testimony. In Brewer, N. & Williams, K. D. (Eds.), Psychology and law: An empirical perspective (pp. 93-150). N.Y.: Guilford Press. Payne, J.D., Nadel, L., Allen, J. J. B., Thomas, K. G. G., &Jacobs, W. J. (2002). The effects of experimentally induced stress on false recognition. Memory, 10, 1-6. Payne, J.D., Nadel, L., Britton, W. B., &Jacobs, W. J. (2004). The biopsychology of trauma and memory. In D. Reisberg & P. Hertel (Eds), Memory and emotion (pp. 76-128). New York: Oxford University Press. Peace, K. A., & Porter, S. (2004). A longitudinal investigation of the reliability of memories for trauma and other emotional experiences. Applied Cognitive Psychology, 18, 1143-1159. Peters, D. (1988). Eyewitness memory and arousal in a natural setting. In M. Gruneberg, P. Morris, & R. Sykes (Eds.), Practical aspects of memory: Current research and issues (pp. 89-94). New York: John Wiley & Sons. Peterson, C., & Whalen, D. (2001). Five years later: Children's memory for medical emergencies. Applied Cognitive Psychology, 15, 1-18. Pezdek, K., & Taylor, J. (2001). Memories for traumatic events. In M. Eisen, G. S. Goodman & J. S. Quas (eds.), Memory and suggestibility in the forensic interview (pp. 165-184). Hillsdale, NJ: Erlbaum.

3.

MEMORY AND EMOTION

115

Pickel, K. L. (1998). Unusualness and threat as possible causes of "weapon focus." Memory, 6, 277-295. Pipe, M.-E., Goodman, G. S., Quas, J., Bidrose, S., Ahlin, D., & Craw, S. (1997). Remembering early experiences during childhood: Are traumatic events special? In J.D. Read & D. S. Lindsay (Eds.), Recollections of trauma: Scientific evidence and clinical practice. New York: Plenum. Pitman, R. K., & Orr, S. P. (1995). Psychophysiology of emotional memory networks in posttraumatic stress disorder. lnJ. L. McGaugh, N. M. Weinberger, & G. Lynch (Eds.), Brain and memory: Modulation and mediation of neuroplasticity (pp. 75-83). New York: Oxford University Press. Poole, D. A., & Lindsay, D. S. (1995). Interviewing preschoolers: Effects of nonsuggestive techniques, parental coaching and leading questions on reports of nonexperienced events. Journal of Experimental Child Psychology, 60, 129-154. Poole, D. A., & Lindsay, D. S. (2001). Children's eyewitness reports after exposure to misinformation from parents. Journal of Experimental Psychology: Applied, 7, 27-50. Porter, S., Yuille, J. C., & Lehman, D. R. (1999). The nature of real, implanted, and fabricated memories for emotional childhood events: Implications for the recovered memory debate. Law and Human Behavior, 23, 517-537. Powers, P. A., Andriks, J. L., & Loftus, E. F. (1979). Eyewitness accounts of females and males. Journal of Applied Psychology, 64(3), 339-347. Quas, J. A., Goodman, G. S., Bidrose, S., Pipe, M.-E., Craw, S., & Ahlin, D. S. (1999). Emotion and memory: Children's long-term remembering, forgetting, and suggestibility. Journal of Experimental Child Psychology, 72, 235-270. Read, J.D., & Lindsay, D. S. (Eds.). (1997). Recollections of trauma: Scientific research and clinical practice. New York: Plenum. Read, J.D., & Lindsay, D. S. (2000). "Amnesia" for summer camps and high school graduation: Memory work increases reports of prior periods of remembering less. Journal of Traumatic Stress, 13, 129-147. Reisberg, D. (2005). Cognition: Exploring the science of the mind (3rd ed.). New York: W. W. Norton. Reisberg, D., & Heuer, F. (2004). Memory for emotional events. In D. Reisberg & P. Hertel (Eds.), Memory and emotion (pp. 3-41). New York: Oxford University Press. Reisberg, D., Heuer, F., McLean, J., & O'Shaughnessy, M. (1988). The quantity, not the quality, of affect predicts memory vividness. Bulletin of the Psychonomic Society, 26, 100-103. Revelle, W., & Loftus, D. A. (1992). The implications of arousal effects for the study of affect and memory. In S.-A. Christianson (Ed.), The handbook of emotion and memory: Research and theory (pp. 113-149). Hillsdale, NJ: Lawrence Erlbaum Associates. Roseman, I. J., & Smith, C.A. (2001). Appraisal theory: Overview, assumptions, varieties, controversies. InK. Scherer, A. Schorr, & T. Johnstone (Eds.), Appraisal processes in emotion: Theory, memories, research. New York: Oxford University Press. Schacter, D. (1996). Searching for memory: The brain, the mind, and the past. New York: Basic Books. Schmidt, S. R., & Williams, A. R. (2001). Memory for humorous cartoons. Memory & Cognition, 29,305-311. Schmolck, H., Buffalo, E. A., & Squire, L. R. (2000). Memory distortions develop over time: Recollections of the 0. J. Simpson trial verdict after 15 and 32 months. Psychological Science, 11 (1), 39-45. Schooler, J ., Bendriksen, M., & Amadar, Z. (1997). Recovered memories and false memories. Oxford: Oxford University Press. Shobe, K. K., & Kihlstrom, J. F. (1997). Is traumatic memory special? Current Directions in Psychological Science, 6, 70-74. Smith, S.M., Glenberg, A., & Bjork, R. A. (1978). Environmental context and human memory. Memory & Cognition, 6, 342-353. Stanny, C. J., &Johnson, T. C. (2000). Effects of stress induced by a simulated shooting on recall by police and citizen witnesses. American Journal of Psychology, 113, 359-386.

116

REISBERG AND HEUER

Steblay, N.J. (1992). A meta-analytic review of the weapon focus effect. Law and Human Behavior, I6, 413-424. Terr, L. C. (1994). Unchained memories: The stories of traumatic memories, lost and found. New York: Basic Books. Tollestrup, P. A., Turtle, J. W., & Yuille, J. C. (1994). Actual victims and witnesses to robbery and fraud: An archival analysis. In D. F. Ross &J.D. Read & M.P. Toglia (Eds.), Adult eyewitness testimony: Current trends and developments (pp. 144-160). New York: Cambridge University Press. Valentine, T., Pickering, A., & Darling, S. (2003). Characteristics of eyewitness identification that predict the outcome of real lineups. Applied Cognitive Psychology, I7, 969-994. Varner, L. J., & Ellis, H. C. (1998). Cognitive activity and physiological arousal processes that mediate mood-congruent memory. Memory & Cognition, 26, 939-950. Wessel, I., & Merckelbach, H. (1997). The impact of anxiety on memory for details in spider phobics. Applied Cognitive Psychology, II, 223-232. Wessel, I., & Merckelbach, H. (1998). Memory for threat-relevant and threat-irrelevant cues in spider phobics. Cognition & Emotion, I2, 93-104. Windmann, S., & Kutas, M. (2001). Electrophysiological correlations of emotion-induced recognition bias. ]oumal of Cognitive Neuroscience, 13, 577-592. Yehuda, R. (1997). Sensitization of the hypothalamic-pituitary-adrenal axis in posttraumatic stress disorder. In R. Yehuda &A. C. McFarlane (Eds.), Psychobiology posttraumatic stress disorder. Annals of the New York Academy of Science (Vol. 821). New York: New York Academy of Science. Yerkes, R., & Dodson, J. (1908). The relation of strength of stimulus to rapidity of habit-formation. ]oumal of Comparative Neurology of Psychology, IB, 459-482. Yuille, J. C., & Cutshall, J. L. (1986). A case study of eyewitness memory of a crime. ]oumal of Applied Psychology, 7I, 291-301. Yuille, J. C., & Daylen, J. (1998). The impact of traumatic events on eyewitness memory. In C. P. Thompson, D. J. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 155-178). Mahwah, NJ: Lawrence Erlbaum Associates.

4 The Effects of Delay on Long-Term Memory for Witnessed Events J. Don Read and Deborah A. Connolly Simon Fraser University

If an event is recognized as being of forensic interest and if it is reported to the authorities, eyewitnesses' descriptions of its details, including its environmental (spatial and temporal) context, the persons involved, and their conversations, comments, and actions, are usually gathered within minutes, hours, or days after the event. However, given the procedures and vagaries of most investigations and justice systems, subsequent eyewitness descriptions of the event may not be heard at inquiries or trials for months or years later. If, in contrast, the event is not so recognized or if it is not reported to the authorities, or if investigators require substantial time to develop the case, the first recall of its details may not occur for even longer periods of time. That is, witnesses are not always aware that they have, in fact, witnessed a criminal act, and some are reluctant to report a crime that they know has been committed. For example, fraud (e.g., passing of bogus checks) may not be recognized as such for a considerable period of time (months to years). Recall of other types of witnessed events is also often on the order of years; for example, recollection of child sexual abuse (CSA) by adult complainants (usually victims, sometimes witnesses) is delayed for a variety of reasons, including a failure to interpret the alleged events as criminal or abusive (Connolly & Read, 2003). Moreover, unlike the single events of, for instance, assault, robbery, or fraud, adults' recollections of some crimes (like CSA and spousal assault) may also include descriptions of multiple repeated and interrelated events. In rare cases, witnesses have been exposed to multiple independent crimes (e.g., bank robberies and terrorist attacks), and recall of each could be examined (e.g., Christianson & Hiibinette, 1993; Connolly & Price, 2005; Edery-Halpem & Nachson, 2004). This chapter focuses on the effects of the delay between an event and its recall, a period of time commonly called the retention interval. The chapter is somewhat broader

117

118

READ AND CONNOLLY

than other chapters in this handbook section on adult memory because we include some discussion of childhood events remembered by both adults and children. For some questions regarding the effects of retention interval on recall, the children's memory literature is much richer than it is for adults and, therefore, should be included. Ironically, it is life experiences from the adolescent and early adulthood years that are disproportionately represented in our autobiographical memories, but investigations of adolescents' memory for witnessed events are particularly few in number (Rubin, 1982). When reading the material on memory for events experienced during childhood (recalled during either childhood or adulthood), it is important to draw attention to important characteristics of children's memory. Ornstein, Ceci, and Loftus (1998) argue that to understand how adults recall childhood events, one must consider "age related differences in prior knowledge, in the strength and organization of underlying representations, [and] in the time course of forgetting" (p. 1031). Children and adults remember differently. Children's relatively impoverished and evolving knowledge base affects the organization, interpretation, and retrieval of memories for past events (e.g., Ornstein, Shapiro, Clubb, Follmer, & BakerWard, 1997). Children forget faster than adults, and this forgetting has been attributed, at least in part, to storage failures (Brainerd & Reyna, 1990, 1995; Reyna & Brainerd, 1995). Notwithstanding this, children's long-term memory for the gist and core details of distinctive events is quite impressive indeed (Bidrose & Goodman, 2000; BurgwynBailes, Baker-Ward, Gordon, & Ornstein, 2001; Fivush & Schwarzmueller, 1998; Peterson, 2002; Peterson & Whalen, 2001). For a more comprehensive review of these themes see Connolly and Read (2003). Because these cognitive developmental changes are particularly dramatic from birth to age six and the relevant literature is so vast, we cannot do it justice here and, therefore, restrict our discussion to studies of retention interval effects in which children were six years of age or older when the event occurred. The reader is referred to the chapter by Pipe, Thierry, and Lamb (this volume) and to reviews of the long-term retention of experiences by younger children (e.g., Westcott, Davies, & Bull, 2002). The chapter begins with the provision of general concepts in the discussion of long-term recollection of events, including both forensically relevant and more mundane autobiographical events. Not surprisingly, many legal professionals are interested in people's abilities to remember events after varying periods of time, and they often seek answers from cognitive psychologists regarding specific cases. However, the complexity and interrelationships of the myriad variables involved in the recollection of remote events make precise answers highly unlikely. Instead, this chapter discusses the empirical relationships between the passage of time and memory that have been examined as a function of many variables. In so doing, we organize the review of relevant research along several dichotomies: that is, we distinguish between single and repeated events, between events recalled once and events recalled multiple times, and between events experienced within or outside of the researchers' control. We conclude with recommendations to the professional community concerning the event and testimonial issues to

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

119

which attention should be given and to the procedures that improve the likelihood that quantitative and qualitative differences in recall provided by eyewitnesses can be under, stood within the context of the passage of time.

RELATIONSHIP BETWEEN THE PASSAGE OF TIME AND MEMORY: GENERAL CONSIDERATIONS

Many laypersons think of the effects of time upon memory to be much like those of sunlight upon a pat of butter: with time, the form of the butter changes and the pat eventually disappears. Sunlight and warmth require time to effect changes in the butter; however, these changes may be eliminated by simply storing the butter in a refrigerator or away from sunlight. Metaphorically speaking, if our memories could be similarly put on ice, we would not anticipate changes in their form. In fact, we often do minimize our memory losses by rehearsing and/or overlearning the details of a specific event or type of information. That is, the event or information is repeatedly re-experienced in precisely the same form. As a result, there is little surprise when someone recalls the overlearned facts of some event, context, or literature in considerable detail over a very long time period. Indeed, even at the longest retention intervals of several decades, some propor, tion of core knowledge or permastore (cf. Bahrick, 1998, 2000; Wixted, 2004) appears to remain available from well-learned material from 30 to 50 years later, for example: for elementary school readers (Read & Barnsley, 1977), the faces of high school classmates or inmates in a prisoner-of-war camp (Bahrick, Bahrick, & Wittlinger, 1975; Lindsay & Read, 2005; Wagenaar & Groeneweg, 1990), actors' roles in theatrical plays (Noice & Noice, 1999), numerical information (Healy & Sinclair, 1996), the names of the streets in one's childhood neighborhood (Schmidt, Peeck, Paas, & Breukelen, 2000), foreign language learning (Bahrick, 2000), and the daily regimen of a concentration camp by Holocaust survivors (Wagenaar & Groeneweg, 1990; Schelach & Nachson, 2001). However, because most events that are forensically relevant are not overlearned in this way, we do not focus in this chapter on overlearned material. Furthermore, we have chosen not to consider in detail those memory assessments of witnesses who, as a result of ex, periencing an event, have developed a psychological disorder such as Post-traumatic Stress Disorder, Dissociative Identity Disorder, or clinical depression. Although we re, turn to this issue briefly in the section on memory for repeated events, the chapter by Soraci, Carlin, Read, Pogoda, Wakeford, Cavanaugh, and Shin (this volume) speaks to the effects of a variety of impairments, psychological and otherwise, on witness event memory. That forgetting or information loss (the terms reduced availability or accessibility may be preferred) for witnessed events occurs with the passage of time is well documented by both experimental and nonexperimental (archival or case studies) research. For example, Turtle and Yuille's (1994) adult participants recalled about 43% fewer details of a video scenario after a 3-week delay than in an immediate interview. Similarly, Flin,

120

READ AND CONNOLLY

Boon, Knox, and Bull (1992) reported declining accuracy in cued recall for a staged event from 66% at immediate testing to 38% at 5 months for 6-year-olds, from 70% to 54% for 9-year-olds, and from 71% to 62% for adults. Furthermore, following a "pirate visit," 6-year-olds' cumulative recall of accurate details for the visit was reduced by 42% and cumulative errors tripled when interviews were completed 6 months after an immediate interview (La Rooy, Pipe, & Murray, 2005). Indeed, 11,000 reports of adults' autobiographical memories of childhood events before the age of 10 demonstrate the dramatically reduced probability of recall of events as a function of the length of the retention interval (Rubin, 2000) and other factors such as lifetime transition periods (cf. Belli, Schuman, & Jackson, 1997). In Rubin and Wenzel's (1996) meta-analysis of over 200 experimental studies involving episodic memories, the overall effect of delaying recall was to reduce the probability of recall as the interval increased in length, but the reduction was not linear (see also Wixted, 2004). In fact, they suggested that a power function best describes the retention function, with most forgetting occurring in the first 5-15 years following a target event, after which the absolute amount of additional forgetting per year is small (see also Anderson, Cohen, & Taylor, 2000; Rubin, 1982). Furthermore, of the information remaining after a long delay, general details can be highly resistant to forgetting. Broad generalizations, however, rarely do justice to the complexity of most phenomena, and memory for events is no different. Specifically, the ability of memory to withstand the passage of time also depends upon a host of factors, including the type of information of interest; that is, whether it has direct relevance for the meaning of the event. The more central the details of an event are to its understanding for an individual (and individuals differ in their assignments of meaning, as do children and adults), the more likely it is that these details will be available for later recall. So-called peripheral details, on the other hand, are much more likely to become unavailable or lost to recall over time. This distinction does not speak to the forensic importance of the details: some feature of a person's appearance may well have contributed little meaning to the event's interpretation and would be judged peripheral, but it may ultimately prove to be critical to the guilt or innocence of a suspect. Furthermore, the magnitude of these differential losses of central and peripheral details over time may be exaggerated as the levels of arousal, emotionality, and stress increase up to some point (Christianson, 1992a; Deffenbacher, Bornstein, Penrod, McGorty, 2004; see Reisberg & Heuer, this volume, for an extended discussion, the distinctions between and review of the relevance of these concepts to memory). Overall, the passage of time is much kinder to the central details than it is to peripheral details. Exactly which features are central or peripheral reflects a subjective parsing of the event, a subjectivity that has varied from one study to another and, in forensic contexts, from one point in time to another: that is, what is seen as peripheral immediately after an event may take on greater importance and become more central as the retention interval and the number of recalls increases (e.g., Poole & White, 1993; Pynoos, Steinberg, & Aronson, 1997; Terr, 1983). It is not known whether details that become central after an event is experienced benefit from the same memorial advantage as those interpreted as central at the time of the event.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

121

Autobiographical Memory Defined Most events of forensic interest are autobiographical events, and our mental representa, tions of them are considered to be autobiographical memories because each refers to a "specific event that took place at a particular time and place; it contains a detailed ac, count of the person's own personal circumstances at the time; the memory includes sensory imagery (visual, auditory, olfactory, or tactile); and the rememberer believes that the event actually happened" (Pillemer, 2001, p. 124). Conway and Pleydell,Pearce (2000) argue that autobiographical knowledge functions to define and maintain the "working self" and the realistic goals of this "working self." These goals influence what gets encoded (new autobiographical memories), what is remembered about the past, and how that information is remembered. This process of creating new autobiographical memories and remembering old ones in turn influences autobiographical knowledge.

Autobiographical Event Representation According to Conway and Pleydell,Pearce (and Conway, 1996), autobiographical knowl, edge exists at three levels of specificity: (1) lifetime periods that organize information around thematic and temporal knowledge (e.g., when I was in university, when the kids were young); (2) general events (e.g., both unique and general events that are more spe, cific and more homogeneous than lifetime periods and that may relate to attainment or failure to attain a goal, such as the activities of summer employment); and (3) event, specific knowledge or ESK (i.e., particular details associated with the single events). ESK is the level at which most errors tend to occur and the level with which this chap, ter is most concerned. In contrast, generally, witnesses remember that they experienced both unique and general events and they can usually locate them within a particular life, time period. Accuracy as to the specific date and temporal context of the event and all of its unique characteristics is, however, often much more difficult and subject to error. Indeed, from their review, Thompson, Skowronski, Larsen, and Betz (1996) argue that the process of locating autobiographical events in time is reconstructive (i.e., derived) and, therefore, will depend upon the recall of other event information and general auto, biographical knowledge as to the ways in which life events normally occur (see also Burt, Kemp, & Conway, 2001). On rare occasions, however, memory for temporal location may be reproductive (i.e., directly accessed) rather than reconstructive, because the date has special meaning for the person and can serve as an organizing landmark for one's life (e.g., a wedding, graduation, natural disaster). Such landmarks have also been shown to improve the dating accuracy of events that are temporally or thematically related to them (Shum, 1998). Witnessed events vary enormously in the extent to which their context and ESK will later be remembered because they also vary in the extent to which they were encoded and in the extent to which other subsequent events may have enhanced or impaired their recall. First, an event's representation or strength in memory is called a memory trace, and it is the trace that results from our cognitive interaction with the event. Memory

122

READ AND CONNOLLY

research over the last century has revealed that if two events vary in the strength of their original encoding, the event with stronger encoding (trace) will demonstrate a slower rate of forgetting than the weaker event, an empirical relationship known as ]ost's Law (Wixted, 2004). The next question, of course, is to ask what factors determine the strength of initial encoding of an event. Most memory theorists argue that what we call memory is really a by-product of cognitive processing: the greater the amount of processing, the richer or stronger the resultant encoding (trace) of memory. The importance of some cognitive processes to such a product is fairly obvious: for example, the extent to which one's attention is directed to an event (in turn, the opportunity to perceive is relevant to attention and may be restricted by variations in lighting, visibility, duration, and motion) is the foundation of developing a strong memory trace. How much attention an unfolding event receives also depends upon specific features, such as its novelty, sensory characteristics, distinctiveness from other similar events in a series, personal relevance or salience, emotional meaning and valence, and features of the witness, such as his or her motivation to remember the event. Second, the strength of a memory trace is also influenced by the post-event context, including further cognitive processing (such as rehearsal) as well as the impact of other similar events or related information (such as media depictions of the event). Emotional events appear to be followed frequently by voluntary recalls that are accompanied by rehearsal or rumination, discussion of the events in social contexts, and (if a public event) interest in its depictions in the media (with adults, Christianson, 1992a; Christianson & Loftus, 1991; Cohn, Mehl, & Pennebaker, 2004; Finkenauer et al., 1998; Luminet et al., 2004; Nachson & Zelig, 2003; Pennebaker & Harber, 1993; Pillemer, 1992, 2001; White, 2002; Yuille & Cutshall, 1998; and for children, Edelstein, Alexander, Goodman, & Newton, 2004). For adults, social discussion and elaboration of significant emotional events appears to continue for several weeks (Cohn et al., 2004; Pennebaker & Harber, 1993). For children, there is evidence that the parental style of reminiscing about the event with the child (i.e., extent to which the discussion is "elaborative") influences the child's later recall of traumatic experiences (Fivush & McDermott Sales, 2004). As a result of these post-event cognitive activities (sometimes called post-stimulus elaboration: see Christianson, 1992b) and, in contrast to the majority of our experiences, these remarkable events (cf. Yuille & Cutshall, 1998) may well survive the passage of time in some form, because they have benefited from the development of a strong memory representation. We emphasize "in some form" because research has consistently demonstrated alterations in the form and quantity of retrievable memory material across all types of events, participants' ages, and long retention intervals (cf. Davis & Loftus, this volume). For example, analyses of accounts of children's accounts of Hurricane Andrew after 6 years indicated that they had been influenced by the continuous discussion of the event (Parker, Bahrick, Lundy, Fivush, & Levitt, 1998). Like many disasters, Andrew had become a highly significant temporal marker (cf. Shum, 1998) such that years later residents in the Miami-Dade area described their lifetime experiences as either "pre- or postAndrew" (Parker, personal communication, June 2000). Furthermore, within a complex event, particular types of information also will have differential probabilities of being remembered; for example, memory for the verbatim

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

123

structure of conversations will, generally, be poorly retained over even very short reten~ tion intervals (see Davis & Friedman, this volume), whereas memory for faces may be available for a relatively longer time and then suffer losses (see Dysart & Lindsay, Volume II). Finally, other similar events or information from other sources that intervene between the target event and its recall may be incorporated in memory, producing changes in the ease and quality of recall of our memories over time that do not necessarily reflect increased accuracy. Thus, the long-term recollection of event details depends not so much upon the passage of time per se, but rather upon what happens in time. As a result, across time we see apparent gains, losses, and distortions. In addition to the effects of cognitive activities that may produce both apparent improvements and decrements in recall, a witness may be asked to recall an event repeatedly within a short period of time, and such retrieval practice and consecutive sampling of the memory representation may increase the cumulative amount of information recalled (cf. Roediger, McDermott, & Goff, 1997). Each of these activities may serve to make the memory representation appear richer after a delay than at the first recall. For example, the children who experienced the terror and devastation of Hurricane Andrew provided narratives of the event 6 years later that were, on average, over twice as long as their initial descriptions (Fivush, McDermott Sales, Goldberg, Bahrick, & Parker, 2004). In so doing, they did not repeat many aspects of their original descriptions, and they provided many new features that may have been accurate or inaccurate in regard to the specific child's experiences. Fivush et al. attributed these extraordinarily large changes to the processes of memory re~ construction by the children, reconstructions that incorporated discussion about the hurricane over the intervening 6 years and reflected their current lives, needs, and interests.

The Constructive and Reconstructive Nature of Memory "It is all too common for caterpillars to become butterflies and then to maintain that in their youth they had been little butterflies. Maturation makes liars of us all" (Vaillant, 1977, p. 197). Most memory theorists hold that autobiographical memory is constructive at encoding by incorporating past experience and existing expectations and reconstructive at retrieval by incorporating new information attended to over the retention interval and bringing one's own beliefs to bear at the time of recall (Bahrick, 1998; Bartlett, 1932; Conway & Pleydell-Pearce, 2000; Hyman & Loftus, 1998; Maughan & Rutter, 1997; Neisser, 1986a; Neisser & Libby, 2000; Pasupathi, 2001; Roediger et al., 1997; Ross, 1989; Rubin, Schrauf, & Greenberg, 2004; Schacter, 1996). Ross (1989) concluded that when people recall their autobiographical histories, they provide reports that are consistent with their current state and their implicit theories about the stability or consistency of that attitude/behavior (see also Maughan & Rutter, 1997). Conway and Pleydell-Pearce (2000) similarly concluded that "[t]he self, and especially the current goals of the self, function as control processes that modulate the construction of memories" (p. 264). This phenomenon has been illustrated in memory for family structure, exposure to combat-related stressors, emotional reaction to political events, and the witnessing of violence (for case studies demonstrating memory changes in children and

124

READ AND CONNOLLY

adolescents, see Pynoos & Nader, 1989; Pynoos, Steinberg, & Aronson, 1997). For example, Poole and White (1993) re-interviewed children who had observed a staged interaction between two people two years earlier. Some changes in the descriptions they gave of the event from the first to the second interview were, according to the researchers, changes in interpretation that reflected a social exchange that was much more socially desirable than what was first reported and much more appropriate (as a reflection of increased maturity) to the children's understanding of interactions 2 years later. Similarly, Davis et al. (2005) recently showed that memory for the specific content of comments made by actors within a staged event may change over a retention interval after the subject learns that the social interaction led to a violent outcome. Within the field of eyewitness testimony, research on the effects of retention interval has generally focused upon memory for people rather than events. Dysart and Lindsay (Volume II; see also Meissner, Sporer, & Schooler, Volume II) reviewed this literature and observed that within the generally short intervals used, the deleterious effects of delay have varied from minimal to substantial. Taken together, Dysart and Lindsay concluded that we probably should expect reductions in person identification accuracy (seen in decreased hits and/or increased false positives in lineup identifications) as a function of time. Some of the largest effects of delay upon person identification described by Dysart and Lindsay were reported in archival studies of suspect identifications in police lineups. However, the absence of "ground truth" (whether the police suspect is indeed the perpetrator) and lack of experimental control in archival studies make it difficult to know whether other variables correlated with delay (e.g., type of crime and available evidence) may instead be responsible for observed decrements. Unfortunately, for an understanding of the effects of retention interval on event memory, the central issue of person identification and its relative ease of testing has diverted eyewitness memory researchers away from the recollection of other event details and researchers such that the kinds of crimes for which event memory is particularly important have been relatively ignored. For example, in cases of CSA the question of who was present is not necessarily as important as what the person did while he was there. Similarly, who said what can be critically important in both civil and criminal litigation, but the application of research on speech-related interactions between speakers to forensic contexts has been limited (see Davis, this volume, for an extended discussion). These limitations for event details over retention intervals are all the more unfortunate because, considered together, the details of an event (the spatial and temporal context and specific objects, persons, their appearances, voices and conversations, actions and their sources and order) are more varied and richer than facial stimuli alone. Similarly, even when the interval between event and recall has varied, difficulties in interpreting the effects of delay arise because the events described in research studies are always different.

METHODOLOGICAL ISSUES IN INVESTIGATIONS OF THE EFFECTS OF RETENTION INTERVAL

Research investigating the recall of real-world events like fraud, assaults, and robberies is faced with numerous obstacles. First, an independent record of the event in question

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

125

usually does not exist, and so proxies for determining the accuracy and completeness of recall are needed. The reliability of retrospective recall has often been the subject of debate. For example, Brewin, Andrews, and Gotlib (1993) provided a review of memory for early family life and concluded that long-term memory for the core details of autobiographical events is quite accurate, particularly if the event was unique, consequential, or unexpected. Indeed, for certain kinds of central and integrated information, the effects of delay on forgetting may be trivial (see also Hardt & Rutter, 2004). On the other hand, Henry, Moffitt, Caspi, Langley, and Silva (1994) compared retrospective and prospective reports across many topics (for example, residence changes, injuries, and family conflict) and concluded that their results "failed to find substantial evidence of the validity of retrospective measures of subjective psychological states and family processes " (p. 99). The picture that emerges here, however, is consistent with that which is seen throughout the chapter. The validity of retrospective reports is much higher for accounts of central details, general information, and/or the presence/absence of events than it is for precise details or judgments and interpretations (see also Davis & Friedman, this volume). Second, to examine the effects of retention interval on the recollection of autobiographical events, researchers must consider the effects of intervening discussion and the reception of additional information from peers, parents, or the media upon later recall. Ideally, to study the effects of the passage of time on memory performance, researchers would prefer single recall conditions (with different participants) at different delays such that the effects of repeated testing are also controlled. Neither of these features, however, is likely to characterize recall of autobiographical memories in natural contexts. Not surprisingly, to gain control over these sources of distortion, researchers have often moved the investigation into the laboratory. Investigations of the effects of retention interval upon event memory may be characterized along three dichotomies: (1) whether a participant recalled the event once or more than once; (2) whether the event was unique or was one of a number of other similar events; and (3) whether the event was under the researcher's control. With respect to the first dichotomy, when separate and independent assessments of memory are made at each (of two or more) retention interval(s), the research design is referred to as between-subjects. On the other hand, if recall was requested from each witness at each of two or more intervals, the design is called within-subjects. From the perspective of gaining an understanding of the uncontaminated effects of the length of retention interval upon recall, the ideal research paradigm allows for comparisons of recall of a single incident at each of several points in time by independent groups of participants who had not been exposed to additional information. Only a minority of researchers use this between-subjects design, but a few studies have included both (e.g., Flin et al., 1992; Poole & White, 1991; Tuckey & Brewer, 2003). The second dichotomy asks whether the focus of study is a single one-time event (of greatest interest to eyewitness researchers) or is one or more events from within a sequence of repeated interrelated events. Regarding the former, witnesses have provided their recollections of single (and often extended) significant personal experiences (e.g., Parker et al., 1998). On the other hand, for the latter, there are studies of war veterans (e.g., Southwick, Morgan, Nicolau, & Charney, 1997) whose _accounts of repeated and specific battlefield events have been examined. Similarly, many child, adolescent, and adult survivors of child abuse report such a series

126

READ AND CONNOLLY

of events (e.g., Bidrose & Goodman, 2000; Connolly & Read, 2003; Lindsay & Read, 1994, 1995). For the most part, research has focused on unique rather than repeated events and, related to the final dichotomy, these unique events were often presented to the participants by the researcher. Nonetheless, many autobiographical events in which we have interest have occurred in natural contexts, are not subject to experimental control, and have often become part of archival records when they are forensically important or have been treated as case studies by researchers. In these latter cases, when ground truth is uncertain (e.g., real-world crimes or events for which no truly objective record exists), an initial or immediate recall is usually treated as a baseline measure of memory, and it is compared to a subsequent recall after the passage of a specific time period (e.g., Yuille & Cutshall, 1986). A similar methodology has characterized numerous studies of flashbulb memories, that is, memories for surprising and emotional events in which recall was requested about the circumstances in which one heard or learned the news about the event at one or more retention intervals. A more objective measure of accuracy has occasionally been gained by examination of memory for the facts of the event itself that were received at the same time as the news of the event's occurrence (Nachson & Zelig, 2003; Pezdek, 2003; Smith, Bibi, & Sheard, 2003). Often, however, rather than assessment of accuracy for the flashbulb event alone, the focus is upon consistency of its recall (e.g., Talarico & Rubin, 2003).

REVIEW OF RELEVANT RESEARCH

Our research review follows these dichotomies with investigations of single events followed by those of repeated events. Within each, we examine studies with single or multiple memory tests and, in turn, for both experimental (researcher-controlled) and case or archival events. Throughout we assume that reports are provided by witnesses who are motivated to recall the events honestly, something that may not always be true in real-world forensic circumstances, where the voluntary lies of commission or omission may be of benefit to witnesses, victims, or perpetrators. Single Events-Experimental Studies

Experimental studies that take place in a laboratory setting ensure researchers access to ground truth about an event because of the controlled presentation of a videotaped or staged crime scenario. A memory test follows at one or more intervals from 5 minutes to several hours, days, or, more rarely, years. When the data are collapsed across studies, the overall impact of increasing the length of the retention interval upon recollections of events has been to reduce performance in terms of both completeness and accuracy on free narrative, open-ended, or cued recall tests, with greater losses in completeness than in accuracy. To be sure, there is substantial variation in the magnitude of these effects such that, for some event descriptions from some individuals, we see relatively small changes over time in completeness or the accuracy of what is recalled. For other events and participants, we can see substantial losses in both.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

127

In laboratory studies with adults, the magnitudes of reductions in recall seen across retention intervals from 1 week to 5 months have been remarkably varied and are not easily attributed to the differences in the staged or video scenarios. For example, Turtle and Yuille (1994) included both an immediate test and a 3-week delayed test following a videotaped presentation of a crime and reported a 43% reduction in recall of details. Yarmey (1998) reported similar losses by adults over a 1-week interval for a perpetrator's characteristics, actions, and statements in a simulated bank robbery. On the other hand, Poole and White (1991) obtained an 18% overall reduction in accurate responses from immediate test to a test delayed by only 1 week for a staged social interaction. Yuille and Tollestrup (1990) reported no losses in accuracy or details for a staged event over 1 week. In the same vein of minimal losses by adults over time, Flin et al. reported a loss of 10% over 5 months in cued recall for a staged mundane event. It is found that children forget faster than adults, so we would expect these effects to be exacerbated in a child population (Brainerd & Reyna, 1990, 1995). Indeed, in the studies by Poole and White and Flin et al., children performed significantly worse after the respective delays of 1 week and 5 months. In two experiments, Tuckey and Brewer (2003) investigated the effects of length of delay upon the kinds of changes to adult eyewitness reports of details judged to be schema-consistent or schema-inconsistent (see also Davis & Loftus, this volume). By schema-consistent we mean congruent with their expectations about what typically happens during that kind of event, and, of course, schema-inconsistent is incongruent with such expectations. Participants recalled detailed information in both free recall and cued recall tests from video simulations of bank robberies. Four retention intervals were used (immediate, 4, 8, and 12 weeks), and the stimulus materials were selected to provide for questioning about schema-irrelevant, schema-consistent, and schema-inconsistent details in the film. Absolute levels of performance were not reported, but effect sizes indexed the losses (reflected in standardized scores) of participants across conditions as a function of interval. Retention interval (from 0 to 8 weeks) for the single-test groups had very large and significant effects on accuracy. The longer the delay between exposure and first test, the less the amount of accurate information reported. However, when performance on each initial interview (at 4 or 8 weeks) was compared with a second interview completed with all participants at 12 weeks, little change was detected, suggesting either that the reductions in trace strength, accessibility, and availability occurred relatively early and then stabilized (cf. Poole & White, 1991, 1993) or that the interim recall session inoculated the participant against additional forgetting up to 12 weeks. Finally, the greatest losses over time were for schema-irrelevant details followed by schema-inconsistent and schema-consistent information, a pattern anticipated by the reconstructive memory literature generally (see Davis & Loftus, this volume). Greenberg, Westcott, and Bailey (1998) used 5-minute and 1-week retention intervals and obtained results consistent with those of Tuckey and Brewer (2003). However, in their case, interest focused on the false recognition of schema-consistent central or peripheral details over the retention intervaL Based on common schemas or scripts for a robbery (cf. Holst & Pezdek, 1992), the authors developed a slide sequence that omitted schema-consistent central or peripheral actions (as determined by pilot work). They found that false recognitions of central details increased significantly from the 5-minute

128

READ AND CONNOLLY

to the 1-week interval. Taken together, these studies demonstrate that the retention interval provides the opportunities for the losses and changes in memory anticipated by a reconstructive memory perspective. The studies just discussed were completed in the controlled (and artificial) context of the laboratory. But experiments may also be completed in natural settings with a broader range of participants. Independent variables are also manipulated, but because field experiments are located outside the lab, they contain additional uncontrolled features that add to the ecological validity of the research. for example, variations in the ambient noise, lighting, and the numbers of extraneous people in the vicinity. Research participants represent a much broader range of ages and background (e.g., Read, Lindsay, & Nicholls, 1998; Yarmey, Jacob, & Porter, 2002), and, in some cases, specialized groups (such as police recruits) have provided data (Hulse & Memon, 2005; Stanny & Johnson, 2000; Yuille, Davies, Gibling, Marxsen, & Porter, 1994). In Yuille et al. (1994) English police recruits engaged in or observed a stressful or nonstressful role simulation ("stop-and-search" interaction) as part of regular training exercises. Recruits were videotaped to obtain an objective record of exactly what had happened. Using standard police interviewing techniques, Yuille et al. (1994) tested them either at 1-week and 12-week delays or only after 12 weeks. When scored for completeness, only those who experienced the stressful event recalled fewer details after 12 weeks than did those first tested at one week. When scored for accuracy, however, greater accuracy was observed at the 1-week versus the 12-week interval only for recruits in the nonstressful condition. In short, as the retention interval increased, reports from the stressed participants were less complete but maintained their level of accuracy. Over time the nonstressed participants' recall declined in accuracy but not in number of details. Hulse and Memon's (2005) police officers did or did not witness a shooting within a simulated firearms course. Recall by police officers who witnessed the shooting was more accurate. On the other hand, for Yuille et al., the recruits' roles (participant vs. observer) were largely unrelated to performance: only the numbers of details recalled significantly favored the participants. Similar to Yuille et al., Stanny and Johnson (2000) found no difference in police officers' recall of a simulated shooting as a function of their level of active engagement in the simulation. These results are somewhat contrary to the outcome of a recent meta-analysis that included 36 comparisons oflow and high stress levels for event details within both experimental and field studies by Deffenbacher et al. (2004). If the recruits had been successfully stressed by the simulations (as reflected in various physiological and self-report indicators required by the Deffenbacher et al. meta-analysis), lower rather than higher accuracy scores for the details of the scene and characters would have been anticipated from the group experiencing highest stress. Reisberg and Heuer (this volume) present a possible resolution to this conflicting pattern of results. The only other true field studies investigating retention interval effects on memory for witnessed criminal events are those of Yarmey (2004), in which short retention intervals up to 4 hours were used, but there were no effects on participants' recall of a target person's appearance. Other field studies by Yarmey (Yarmey et al., 2002) have

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

129

focused on earwitness identification (memory for voice and speech characteristics), and the reader is referred to Yarmey's chapter in Volume II of this handbook.

Archival, case Studies, and Flashbulb Memories As mentioned earlier, interpretation of the apparent effects of retention interval for naturally occurring events is particularly difficult because of the absence of objective records of the event, an expectation of uncontrolled rehearsal and discussion, and methodological differences in testing. Typically, report consistency is used as a proxy for accuracy on the assumption that inconsistency reflects changes in the event's memorial representation. However, the actual relationship between accuracy and consistency is weak (Fisher & Cutler, 1995; Brewer, Potter, Fisher, Bond, & Luszcz, 1999). Even inconsistencies that are apparent contradictions (e.g., reporting two different days as the temporal location of the event) may not reflect inaccuracy in the sense that one of the details was not experienced; rather, if the target event is an instance of a recurring event, the contradictions may reflect confusion between instances (Connolly & Price, 2005). As we shall see, consistency of reports appears to depend in part upon the timing of the first recall with greater consistency when recall is tested somewhat later rather than immediately. The scoring criteria for consistency vary across studies, but they usually represent comparisons of the numbers of unique units of information recalled (as defined by the researchers) and, for each unit, a subjective scoring as to whether its two reports are inconsistent, partially consistent, or completely consistent. The most frequently cited example of an archival real-world crime event for which there was a comparison of two retention intervals is that of Yuille and Cutshall (1986). The researchers interviewed 13 of21 people who witnessed a shooting of a thiefby a gun shop owner 4-5 months earlier. Accuracy was determined by the correspondence between each of the 13 police witness reports collected at the scene or 2 days later and a reconstruction of the crime by the researchers. Whereas much of the reconstruction was derived from police photos and other hard forensic evidence, a portion was based on the original police statements gathered from all witnesses, and it was these same reports that were then scored (at the short retention interval) for their match with the researchers' reconstruction. The problematic implications of this procedure for obtaining a pure measure of accuracy should be clear (see Brown, Scheflin, & Hammond, 1998; Christianson, Goodman, & Loftus, 1992; for further comments on this research). In any case, at the first interview, the witnesses's responses were judged to be 80% accurate in their recall of objects, people, and actions. At the second interview 4-5 months later, their accuracy levels remained at about the same level. Thus, on the surface and ignoring reservations about the validity of the verifying reconstruction, it appears that memory for this forensically relevant and arousing event was high and that there was, apparently, little impact of the passage of 5 months on their recall. Yuille and Cutshall suggest that the results question broad assumptions in the research literature about the fallibility of eyewitnesses and the detrimental effects of retention interval on memory for realworld events. To assess this claim however, we are uncertain about other data these results should be compared with. Certainly, if we compare this overall level of accuracy

130

READ AND CONNOLLY

(assuming veridicality of the reconstruction) with the outcomes of episodic list-learning studies for which delay has been extended to weeks or months (e.g., Ebbinghaus, 1885/ 1913; Wixted, 2004), it is impressive and did not exhibit the classic negatively decelerated forgetting curve (Rubin & Wenzel, 1996; Wixted, 2004) over the 5-month retention interval. However, compare this fatal shooting event with the following event described by Haber and Haber (1998). Again, 13 involved and threatened witnesses to a bank robbery provided immediate recall of the several robbers and their actions. In this case, accuracy could be objectively determined because a bank videotape of the event existed. However, unlike the Yuille and Cutshall results, the level of accuracy was extremely low. Indeed, no witness correctly described the sequence of actions, nor did any witness provide a consistent description of any of the robbers, and 7 of the 13 witnesses incorrectly selected a foil photo from a photospread, 4 of them selecting a photo of another witness in the bank at the robbery. Or, compare the data with Flin et al.'s (1992) mundane experimental event with an equivalent 5-month retention interval: despite the pedestrian nature of this staged event (a nurse giving a lecture demonstration of foot hygiene), adults' correct recall (objectively scored) and consistency of recall by adults approached that of Yuille and Cutshall at both delay intervals with overall accuracy rates of approximately 70%. These comparisons do not easily lead to the conclusion that the differential effects of retention interval are determined by the unique features of the events themselves. In a subsequent paper, Cutshall and Yuille (1989) again examined memory of witnesses to a variety of other single criminal events (shootings, stabbings, and bank robberies), who were first interviewed immediately or shortly after the event and again 4 to 24 months later. As before, accuracy was determined by matching witness reports to the researchers' reconstructions based on police evidence and witness reports. Overall, recall of the accurate event details was impressively high and was maintained across interviews and the long retention intervals experienced. Some numerical differences in judged accuracy were found between the violent (shooting) and nonviolent events (bank robberies), and Cutshall and Yuille attributed these differences to the degree of violence, concluding that the nonviolent events were not remembered well despite, in this case, accuracy levels over 80% (in comparison, the violent crimes accuracy scores ranged from 90%). Yuille and Cutshall reported errors at both interviews in both studies (in the 20% range), and, although the majority of these errors involved person descriptions (weight, height, etc.), others were categorized as errors in description of actions and other person or object details. Furthermore, changes in reports did occur between the two interviews, some of which were interpreted as reconstructions of a witness's perspective and mistaken recollection of actions over time. Christianson and Hiibinette (1993) provided multiple-choice questions to 58 witnesses to some 22 robberies and compared their responses after delays from 4 to 16 months to police file verification reports of the robberies, based on a variety of sources of information, including the original eyewitness reports. Accuracy ranged from 62% to 81% across retention intervals, questions, and types of witnesses (customers, employees, and victim/tellers). Unfortunately, performance was not assessed as a function of the length

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

131

of the retention intervals, however. As in the Yuille and Cutshall studies, no assessment was made of the impact of potential discussion and rehearsal by witnesses either before the initial interviews or across the retention interval (see Yarmey, 1998, for a discussion of the potential contamination among witnesses at real-world crime scenes and during the interview process). There are, of course, many other events that people experience that are unexpected, brief, surprising, and emotional but do not usually involve criminal activities or necessarily have forensic relevance. These are sometimes called flashbulb events, experiences that have been likened to eyewitness events because of their brevity and surprise, events in which someone received shocking or emotionally distressing news. The flashbulb literature now comprises at least 30 published studies and spans a wide variety of public events involving assassinations, deaths, natural disasters, and other tragedies. There is no question and much data, both objectively scored (e.g., Luminet et al., 2004) and anecdotally reported, that participants often had very strong emotional reactions to many of the events. For example, the death of Princess Diana generated overt displays of national mourning rarely seen for the death of any public figure (Kvavilashvili, Mirani, Schlagman, & Kornbrot, 2003; Ost, Vrij, Costal!, & Bull, 2002). For flashbulb events, access to ground truth concerning the circumstances in which a witness received the news is rare. As mentioned, researchers examine the consistency of participants' reports from one interview to another without regard for accuracy of memory across retention intervals. Like other archival research, the inter-interview delays have not been systematically manipulated, having varied from a few days to 3 years. Schmolck, Buffalo, and Squire (2000) speculated that differences in consistency obtained across studies relate to the length of the retention interval between interviews. If the interval is less than 1 year, generally researchers have reported good consistency. However, if the interval is more than 1 year, generally poor consistency has been reported. In the study by Schmolck et al., participants were asked about the 0. J. Simpson verdict 3 days after it was released and again either 15 months or 32 months later. Recollections at 15 and 32 months were coded as containing no errors, minor errors, or major distortions. Schmolck et al. (2000) reported that within 15 months of the Simpson verdict only 50% of the details matched the original reports taken 3 days after the verdict and that this was virtually halved again by the 32-month retention interval. Most interesting, of the reports that were inaccurate at 15 months, 66% were inaccurate because of"don't know" responses, whereas at 32 months, 88% were inaccurate because of major distortions. The retention interval between the event and the first recall may also influence consistency. Winningham, Hyman, and Dinnel (2000) had participants recall the same Simpson verdict but collected the first report either 5 hours or 1 week after its reporting by the media. A second interview occurred 8 weeks later. Based on their review of prior flashbulb studies, they speculated that consistency is greatest if the original report is collected, not immediately, but several days following the event, allowing two memory processes to occur: (1) immediate forgetting of minor details that will not appear in any subsequent reports, and (2) the early incorporation of new information available from other sources in the first few days following the event. Their results supported these

132

READ AND CONNOLLY

ideas and the possibility of a kind of consolidation period in which these post-event cognitive processes may operate: the earlier the first report, the lower the eventual consistency scores (cf. Lee & Brown, 2003). In a particularly thorough study that avoided virtually all of the weaknesses of prior flashbulb research, Talarico and Rubin (2003) examined recall of details for hearing about the September 11 World Trade Center terrorist attack at four retention intervals from 1 day to 32 weeks (with independent groups) and found systematic decreases in consistency and amount of detail recalled in the reports as a function of the length of the interval. More specifically, the percentage of consistent details decreased by about 35%, and inconsistent details increased by 100% across the 32-week period. These changes were mirrored in the recall of control (everyday) events over the same period; however, whereas vividness and confidence declined for control events, they did not do so for flashbulb events, despite their equivalent levels of inaccuracy. Previous research had failed to include critical comparisons with other non-flashbulb events. Memory for the emotional features of flashbulb events also has been explored by a few researchers (e.g., Levine, 1997; Levine & Safer, 2002; Luminet et al., 2004). Levine asked 227 Perot supporters about their emotional reaction to Perot's withdrawal from the 1996 U.S. presidential election. They completed the survey twice, once within 10 weeks of his withdrawal and again 5 months later. Although reports of emotions were largely consistent, there were systematic distortions in recall of their emotions, and these appeared to reflect the participants' current rather than retrospective (at time of his withdrawal) appraisal of events (see also Conway, 1996; Davis & Loftus, this volume; Ross, 1989). In another study, participants were asked about their emotional reaction (upset, surprised, uneasy, worried, angry) to the sinking of the MS Estonia on two occasions separated by 14 months. Correlations associated with ratings on each of the emotions as well as ratings of personal importance and of national importance were low (Christianson & Engelberg, 1999). In addition to the research just discussed, some general statements may be made about the effects of the passage of time on recall of flashbulb reception events. First, although for a few events reported details have been strikingly consistent (e.g., Neisser et al., 1996; Pillemer, 1984), inconsistency across a retention interval characterizes the majority of investigations (e.g., Christianson, 1989; Lee & Brown, 2003; Nachson & Zelig, 2003; McCloskey, Wible, & Cohen, 1988; Neisser & Harsch, 1992; Talarico & Rubin, 2003; Winningham et al., 2000), but it is neither more nor less than that observed for memories of everyday events (Talarico & Rubin; 2003). Furthermore, consistency appears to be related to the extent to which the participants were close to the event or for whom the event had personal consequences and, importantly, the extent to which rehearsal (rumination and discussion) of the events occurred (Er, 2003; Luminet et al., 2004; Neisser et al., 1996; Parker et al., 1998; Pezdek, 2003; Smith et al., 2003; Tekcan, 1999). For example, in a comparison of descriptions of hearing about the September 11 terrorist attacks, report consistency was greater for regions geographically closest to New York (Pezdek, 2003). The inconsistency may apply to both the memory for circumstances of hearing of the event (the flashbulb) and the factual aspects of the event itself. For example, Nachson and Zelig (2003) reported 25% and 36% losses in these two types

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

133

of memory over 11 months, respectively. Similarly, Smith et al. (2003) and Tekcan, Ece, Giilgoz, and Er (2003) have shown that event memory is lost more quickly than flash~ bulb details over 6 months. Second, the confidence with which the details have been provided has changed little over the delay, notwithstanding the reduction in actual accuracy (Talarico & Ru~ bin, 2003; Weaver, 1993; Wright, 1993). Third, compared with memories for everyday events, flashbulb memories are also accompanied by higher subjective vividness and more recollective phenomenology (Talarico & Rubin, 2003; Weaver, 1993). Fourth, it appears that if the initial report is collected not immediately, but several days after the event, subsequent reports demonstrate greater consistency over time with this report than with those collected more immediately (Schmolck et al., 2000; Winningham et al., 2000). Fifth, a longer interval between recall sessions leads to greater inconsistency across reports (Schmolck et al., 2000). Sixth, there are frequent examples of the conse~ quences of memory reconstruction of these kinds of events with the passage of time (e.g., Altmann, 2003; Neisser, 1986b; Neisser & Harsch, 1992; Pezdek, 2003; Smith et al., 2003; Wright, 1993) and with the changes in memory representation as a result of media depictions of the event (e.g., Nachson & Zelig, 2003; Tekcan, 1999; Wright, 1993). For example, Tekcan (1999) found that as exposure to media news items increased, consis~ tency of the flashbulb report decreased. Furthermore, in several cases and after several months, many respondents have reported witnessed events related to the flashbulb mem~ ory for which no visual record ever existed (44% recalled seeing the Paris crash of Princess Diana's car on television [Ost et al., 2002], and 55% recalled seeing film of the crash of an airliner with a building in Amsterdam [Crombag, Wagenaar, & van Koppen, 1996]) or for which no visual record had yet existed (e.g., 73% of one American sample and President George Bush reported seeing the impact of the first plane with the World Trade Center on September 11, 2001; Pezdek, 2003; Greenberg, 2004, respectively). No visual record of the first impact was made public until September 12.

Personal Flashbulb Memories The potentially contaminating effects of media attention in the flashbulb studies, of course, are results of their public nature. Emotional flashbulb events such as hearing of the loss of a parent or loved one are experienced by most people, but, because they are private events, they have only rarely been examined by researchers (e.g., Barclay & DeCooke, 1988; Christianson & Loftus, 1990). For example, within a diary study, Barclay and DeCooke (1988) compared recall over 3 months of a tragic death of a parent with everyday memories of two participants. Comparatively, the tragic event was remembered more vividly 3 months later, but it also demonstrated loss of many peripheral details while increasing in "gist" memory at the same time. We assume that the aftermath of pri~ vate experience includes rumination, discussion, and related social interaction, activi~ ties that would contribute to some reconstruction of the event and serve to maintain its vividness and the participants' confidence over the retention interval. Shaw, McClure, and Dykstra (this volume) detail the strong effects of a number of social variables upon the maintenance of and increase in confidence for event memory over time.

134

READ AND CONNOLLY

That rumination, vivid imagery, and discussion across a retention interval can establish false autobiographical memories for emotional events has also been shown in a laboratory study by Porter, Yuille, and Lehman (1998), for whom 56% of their participants developed either partial or complete false memories for childhood events such as serious illness, injury, or animal attacks that never occurred. For some, these memories developed by the end of the first of three interviews. As well, a number of equally serious and emotional "true" events provided by participants' parents were apparently completely forgotten over an approximate 15-year delay by a minority of participants (or their parents). In other work by Lindsay, Hagen, Read, Wade, and Garry (2004), a long retention interval provided the basis for establishing false memories for a childhood prank among 65% of participants (see also chapters in this volume by Soraci et al.; N euschatz, Lampinen, Toglia, Payne, & Preston; and Smith & Gleaves). Taken together, the research demonstrates that emotional memories are certainly not immune to the effects of reconstruction over the retention interval. A related category of single events examined at different points in time is those that not only are emotionally arousing or stressful but also qualify as traumatic because the individual may experience threats to their own life or the lives of others. With adults, these have been studied for recollections of wartime combat (Southwick et al., 1997) or policing (Karlsson & Christianson, 2003). With children and adolescents, these have included witnessing of natural disasters and of violence directed against others, including parents, siblings, and peers (e.g., Parker et al., 1998; Pynoos et al., 1997). There have been numerous recent discussions of whether these kinds of events might be considered so sufficiently different from other emotional memories that special memory mechanisms may be involved. Here, however, we consider the impact of the passage of time upon them. Unfortunately, there appear to have been no systematic investigations of the passage of time on these memories, and many of the reports are described only in psychiatric and psychological case studies. However, in their summary of the relationship between memory and trauma, Pynoos et al. emphasize that a number of events occurring over time can frequently modify or allow for reconstructions of memories for traumatic events by children and adolescents, influences upon memory that are frequently described in the reconstructive memory literature in general (e.g., Davis & Loftus, this volume; Ross, 1989). These are the presentation of new information; co-construction of the events with parents, peers, and siblings; intervention strategy fantasies that are associated with memory schematization of the event; reappraisals of the meaning of internal and external threats; and self-protective mental efforts to weaken certain details. In short, for these kinds of events, what occurs during a retention interval often has dramatic effects upon witness memory. With increases in the length of the retention interval, availability of accurate peripheral details of the original experience would be presumed to decline faster. Indeed, Pynoos et al. report that the reconstructive (rather than reproductive) nature of memory produced shortened narratives over repeated recalls over hours to months, with some peripheral features becoming more central over time. Finally, another related approach is directed toward exploring qualitative differences (and possibly, differing underlying memory mechanisms) between events of differing emotional valence (positive, neutral, negative). To do so, participants recalled self-

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

135

selected events on two occasions (2 weeks to 3 months apart), and the consistency of participants' retrospective and current judgments about them was examined (e.g., Peace & Porter, 2004). The length of the retention interval has not been varied, but Peace and Porter argue that the differential consistency scores across types of events suggests that subjectively distressing or traumatic events are more resistant to the passage of time. In our view, arguments that these differences reflect different underlying memory processes would be premature for a variety of reasons, including limitations generated by the researchers' sampling techniques of both participants and memories and a substantial overlap in consistency score distributions between the various memory types. Researchers are divided on whether an understanding of memory for trauma and highly emotional events requires unique memory mechanisms (Brewin, McNally, & Taylor, 2004; Brown et al., 1998; McNally, 2003; Pezdek & Taylor, 2002; Read, 1997; Shobe & Kihlstrom, 1997; Yuille & Cutshall, 1998). In general, however, the type and quantity of details re, ported about emotional events and neutral events volunteered by participants are differ, ent. Compared with remembering positive experiences, trauma is remembered with more central details generally (Berns ten, 2002; Berns ten & Rubin, 2002), but with less sensory (Bryne, Hyman, & Scott, 2001; Porter & Birt, 2001) and more emotional infor, mation (Bernsten, 2001; Porter & Birt, 2001). Also, compared with memory for neutral events, memory for central details of negative events is superior (i.e., more complete and less susceptible to forgetting), and memory for peripheral details of negative events is poorer (i.e., less complete and more susceptible to forgetting; Christianson & Safer, 1995; Engelberg & Christianson, 2002; Porter & Birt, 2001). One explanation for these differences is that attention is focused on the most salient aspects of negative events (including, of course, those details that directly elicited the emotion) that support strong memories for those details at the expense of memories for more peripheral details (Christianson & Safer, 1995; Reisberg & Heuer, this volume; Yuille & Cutshall, 1998). A second explanation is that memories of trauma are rehearsed, discussed, and/or thought about more often than positive events or neutral events with long,term consequences for availability in recall (Bernsten, 2001; Bryne et al., 2001; Neisser, 1986a, 2003; Porter & Birt, 2001). A third possibility is that because the specific events examined by the researchers are selected by the participants, biases in retrieval may yield events that differ on other variables (e.g., recency) that may underlie other differences obtained.

Repeated Recalls of Single Events Whereas the previous sections focused on single events recalled once (perhaps following the establishment of a baseline, as in studies that use consistency as a proxy for accu, racy), it is commonplace for witnesses and victims to provide several recalls of a target event, in one or more police statements, depositions, or affidavits, to testimony at a pre, liminary inquiry or examination for discovery, and finally to testimony given in court under direct and cross, examination. Under cross,examination, memory is tested repeat, edly as counsel seeks to uncover some fatal inconsistency in the reported details of an event and prior statements in the hopes of such a discovery. Certainly, the implicit and,

136

READ AND CONNOLLY

often, explicit message conveyed to the trier-of-fact is that testimonial inconsistencies denote unreliability of the witness's memory (cf. Brewer et al., 1999). We know from mock juror studies that such inconsistencies are detrimental to the credibility of witnesses (e.g., Berman &Cutler, 1996; Berman, Narby, &Cutler, 1995; Brewer et al., 1999), despite the fact that the statistical relationship between testimonial consistency and overall accuracy is weak (Brewer et al., 1999). Person identification researchers worry about the contaminating effects of viewing too many photos or of viewing a mugshot photo prior to the identification photospread (McAllister, Volume II; Read, Tollestrup, Hammersley, McFadzen, & Christensen, 1990); however, similar concerns are rarely expressed in regard to the multiple recalls of events within a legal proceeding. Little is known about the effects of multiple recalls of a single experienced event. We do know, however, from investigations of real-world cases where inappropriate pretrial interviewing and questioning techniques were used, that repeated interviews and questioning can seriously distort the testimony of both children and adults, so much so that determination of whether the described events actually occurred is in question (Bruck & Ceci, 1995; Kassin, 2005; Lamb, Orbach, Warren, Esplin, & Hershkovitz, this volume; Lindsay & Read, 1994, 1995; Poole & Lamb, 1998; Poole & Lindsay, 2002). Experimental studies have manipulated the number of recalls of a known event for two reasons: ( 1) to examine the benefits of multiple tests for achieving as complete a recollection as possible (e.g., Bornstein, Liebel, & Scarberry, 1998; Dent & Stephenson, 1979; Dunning & Stern, 1992); and (2) to examine theoretical perspectives about memory in regard to the memory phenomena of reminiscence and hypermnesia (e.g., La Rooy et al., 2005; Scrivner & Safer, 1988; Turtle & Yuille, 1994). Briefly, reminiscence refers to the observation of an absolute increase in the total number of details recalled once, whereas hypermnesia refers to the observation of gains in net recall (more new information is recalled than old information is forgotten). From the perspective of fundamental memory processes, the collection of several recalls seems a good idea. In most models of episodic and autobiographical memory, what we call a memory is assumed to be a transient mental construction based upon some nonexhaustive sampling of knowledge in response to a specific cue or request (cf. Conway, 1996; Roediger et al., 1997; Turtle & Yuille, 1994; Whittlesea, 1997). Furthermore, most models presume that event details are encoded along a number of dimensions (e.g., visual, auditory, verbal, conceptual, emotional, spatial) and that access to the memory representation depends upon cues relevant to these dimensions and upon the relationship between cues available at encoding and at retrieval (Fisher & Geiselman, 1992). Taken together, this means we anticipate variability in free recall across repeated tests and that later tests will yield details not previously recalled and omissions of details previously recalled. For other types of tests that provide cues (e.g., cued-recall, forced choice alternatives, recognition testing); however, considerably less variability across tests would be anticipated because respondents are not required to generate their own cues for sampling memory (e.g., Dent & Stephenson, 1979; Dunning & Stern, 1992). Inconsistency in witness reports has several meanings (cf. Brewer et al., 1999). As discussed under archival studies, it may mean simple variation in the amounts of infor-

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

137

mation reported across interviews, but no specific detail in one interview necessarily contradicts any other. It may also mean that a level of detail is given at one time but not at another, more an error of omission than of contradiction. However, if one detail contradicts another previously reported detail, it has been assumed that only one can be accurate, although both could be inaccurate. This may be true when the witness is reporting a unique event, but not so if the witness is reporting a particular instance in a series of similar instances (Connolly & Price, 2005). In this section we discuss contradictions in repeated reports of single events. Unlike the variability in the amount of information recalled across interviews, the reporting of contradictions appears to arise from the loss of availability of the first response, regardless of its original accuracy. For example, the respondent may have initially fabricated information during free recall to give the narrative greater coherence, or guessed as a response, or chosen an alternative from among those provided in a closed question. If contradictions arise because initial responses are forgotten or are temporarily unavailable (cf. Parks, 1999), longer delays between tests should increase the number of inconsistencies because increased time allows for increased forgetting and the generation of new responses. Increasing the number of tests or interviews, on the other hand, would likely serve to reduce overall cumulative inconsistencies (but not necessarily accuracy) because some prior responses would have a greater chance of being remembered and repeated (Roediger et al., 1997). Practical recognition of these potential variations is reflected by the fact that in real-world legal proceedings with long delays, witnesses are usually asked to "refresh" their memory by re-reading their prior statements before being examined in court. All experimental investigations of repeated testing have used staged or videotaped scenarios. An early study with 11-year-old children by Dent and Stephenson (1979) found that more information was recalled as interviews increased (from one to three), but these gains were restricted to testing over the short term (immediate, 24 hours, and 48 hours) compared with a longer term (2-week and 2-month delays). Furthermore, the increase was observed only for free recall and not for specific or cued-recall questions. Single-recall control groups were not included, so we are unable to definitively estimate the benefits or losses of prior recall opportunities. Scrivner and Safer's (1988) participants completed multiple recalls of a videotaped violent burglary scenario. Four time-limited free recall opportunities were administered, three in the first session and the fourth 48 hours later. Single-recall groups were not included. Their results demonstrated increased numbers of details recalled across tests for a variety of types of information in the film. Indeed, the gains were sufficiently large to demonstrate hypermnesia because the amount of new information participants recalled over trials exceeded what had been forgotten. Scrivner and Safer concluded that multiple recalls of witnessed events would indeed be useful in forensic circumstances because relevant information not initially recalled was often recalled later. In three experiments Dunning and Stern (1992) examined hypermnesia across three recalls of videotaped crime scenarios over a 1-week delay. Like Scrivner and Safer, they obtained the hypermnesic effect (about 10-20% increase in information recalled), but only when free recall was requested. With cued recall or a multiple-choice question format, no hypermnesia was obtained.

138

READ AND CONNOLLY

Turtle and Yuille (1994) included appropriate single-test recall opportunities (for a videotaped crime simulation) over a 3-week interval. Their free recall opportunities varied from one to four, but, unlike Scrivner and Safer's, were unlimited in length. The researchers anticipated that reminiscence (the total number of details recalled at least once) would increase over recall attempts, but that hypermnesia was likely an artifact of the time-limited tests used by Scrivner and Safer. The result was striking: reminiscence was demonstrated across tests, and one immediate recall test apparently was sufficient to support the recall of about 85% more information at 3 weeks compared with the single delayed test condition. This result matches well that obtained by Yuille and Tollestrup (1990) in which recollection of the details of a staged event was significantly better when an immediate recall preceded a second test 1 week later compared with a singletest group at 1 week. Multiple tests also led to significant increases in the numbers of errors provided, although their proportion of total information was small (about 5%). Rather than distributing recall opportunities for violent and nonviolent stimulus materials over retention intervals of 1-3 weeks, Bornstein et al. (1998) used multiple tests within a single session. Successive tests produced significant but modest gains in total recall (reminiscence only) over the three opportunities. As found by Turtle and Yuille, errors also increased significantly across recalls but were few in absolute numbers. In each one of these studies, the details recollected in free recall demonstrated considerable inconsistency because the specific details reported on any given test were frequently different from those reported on a previous test. Poole and White completed two experiments in which children (4-, 6-, and 8-yearolds) and adults recalled on two occasions an ambiguous staged social interaction, separated by 1 week in the first study (1991) and (with the same participants) by 2 years in the second study (1993). Their interest was in the effect of two recall sessions and, within each, of multiple (three) cued-recall tests administered one immediately after the other. Fortunately, the researchers also included appropriate single-test control conditions. In terms of recall of correct units of information, for adults there was a clear benefit of immediate testing upon the second recall 1 week later, with 39% more details recalled than in the single-session delayed group. The authors suggested that the earlier interview consolidated memories for the adults; however, a similar gain was not observed for the 4-8-year-old children, and, if anything, among the youngest children, the repeatsession participants recalled less than those in the delayed group. In contrast, Peterson and Tizzard-Drover (2005) more recently found support for consolidation effects only with the youngest in their sample of children admitted to hospital for illness or injury. The multiple-questions trial within sessions in Poole and White's first study demonstrated striking inconsistencies (contradictions) for children as well as adults when a series of yes/no questions were used to assess memory (cf. Parks, 1999). For example, up to 40% of the adults gave inconsistent answers to these three sets of questions in the first interview, but this figure dropped to 18% in the second session. Children also became more consistent over sessions, suggesting that, like the adults, they remembered their earlier answers to the questions. However, 2 years later, the same adults demonstrated good consistency within sessions (and with the first study), but the children did not, with approximately 50% changing responses across question presentations. A similar

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

139

procedure was used to assess repeated testing for a flashbulb memory. Eight months after the 0. ]. Simpson verdict, Bluck, Levine, and Laulhere (1999) had participants recall details of the event three times within a single session. A videotaped record provided for objective scoring. As with other repeated test studies, the cumulative amount of information saw an increase of almost 100% over the trials, but these gains were matched by an equivalent increase in error. In summary, in the short term, consistency can be quite good, but contradictions arise as both the number of tests and length of the retention interval increase, particularly for children. La Rooy et al. (2005) also tested children (age 6) on three occasions (immediate interview followed by two interviews 24 hours apart after 6 months). Considerable information was lost from the immediate interview to the second interview, and there were significant increases in errors. Comparing across the interviews, more correct information (reminiscence, but not hypermnesia) was recalled, but more errors were also reported. However, after both interviews at 6 months, cumulative correct recall approximated that available at the immediate interview 6 months earlier. Of the new information reported after the 6-month delay, only 56% was correct, an outcome consistent with other studies of children seen in hospitals for illness or injury by Peterson, Moores, and White (2001). Brewer et al. (1999) investigated consistency in two reports of a filmed robbery scenario over a period of 2 weeks. Two important findings were obtained: first, in terms of overall agreement (number of items reported consistently at both interviews, regardless of accuracy divided by total items recalled) between the interviews, the adult participants averaged about 45%, and an overall measure of consistency (degree of agreement) yielded about 85% (when the same details were reported in each interview). Second, the numbers of correct units of information declined (rather than increased, as seen in some prior studies) significantly from the first to the second session. Thus, we can see that, like the earlier work, exactly which details are reported in an interview varies substantially and, when the same detail is reported, its description is generally, but not always, consistent. Furthermore, a repeated interview does not always yield more information. Taken together, the studies on multiple interviews or tests provide several general findings. First, participants typically demonstrate considerable variability in the particular details given in free recall from one test to another, but in a cumulative sense, more correct information is reported over repeated tests. However, the magnitude of these gains generally decreases as the number of tests increases. Second, this inconsistency is accompanied by additional error, often small in adults, but substantial in children, and for both groups, the proportion of error in the newly reported information increases with the length of the retention interval. Third, the phenomena of reminiscence and hypermnesia are dependent upon particular kinds of testing procedures. When free recall is requested, we see evidence for each, but when specific, cued-recall, or multiple-choice questions are used, neither is observed. Furthermore, these benefits of multiple recalls are greatest when the multiple tests occur within close temporal proximity to each other: long intervals between them may provide for sufficient forgetting of details, or, depending upon the nature of the event and opportunities for post-event contamination, subsequent recalls may incorporate information from other sources (see Lamb et al., this volume). Finally, an interview or test soon after the event reduces or ameliorates the

140

READ AND CONNOLLY

effects of the passage of time across a subsequent retention interval of days and weeks. In a review of the effects of retention interval on forgetting of laboratory verbal materials, effects are predicted from Jost's Wixted (2004) suggested that these consolidation~like Law and support the idea that over time, memory traces that still remain become more resistant to the disruptive forces of decay and interference. How many interviews or tests are sufficient? Following the day-care investigations of the 1980s and 1990s (Poole & Lamb, 1998), there is widespread recognition that, at least for children and possibly for adults, more than a few interviews are probably too many (Fivush, Peterson, & Schwarzmueller, 2002; Lamb et al., this volume; Poole & Lindsay, 2002). To the extent that no new information is retrieved over the retention intervals, the gains of repeated testing show diminishing returns, and continued testing may generate frustration and increase the rate with which incorrect information is of~ fered (e.g., Poole & Lamb, 1998). With children in particular, as Lamb et al. (this volume) indicate, the purpose of repeating the same questions generates other difficulties, inducing respondents to construct or change responses. Similar effects may occur with adults (Poole & White, 1991; 1993). There has been great interest in the nature of repeated recall ofCSA memories in therapeutic contexts (e.g., Alpert et al., 1998; Lindsay & Read, 1994, 1995), but interest has focused on the recollection of false rather than true events.

Repeated Events By repeated events we refer to a stream of similar, related events. Criminal offences that occur repeatedly, such as child physical or sexual abuse, domestic violence, and criminal harassment, raise legal and cognitive difficulties that appear at first to be irreconcilable. On the one hand, it is contrary to fundamental justice to require a person to defend himself or herself against a vague and imprecise charge. The charge must be particular enough for the accused to raise a defense. When the charge under investigation involves repeated similar experiences, this often requires the complainant to report details of a particular instance or subset of instances of the offence. On the other hand, as described in more detail below, when several similar experiences are encountered, memory for the series of instances becomes more general and less particular. Despite its forensic relevance, there has been very little recent systematic study of memory for instances of repeated events and repeated tests of repeated events, and even less on the effect of a retention interval on such memory reports. One exception is that survey researchers have attempted to decompose a string of related events into the single events on which the series is constructed. One technique for doing so involves a type of repeated testing (e.g., Means & Loftus, 1991), but the research emphasis has not been upon retention interval. Furthermore, from the archival work on CSA, the longer the retention interval, the less likely it is that an allegation of CSA will be prosecuted and the less likely it is that there will be a successful conviction (Connolly & Read, 2005). Presumably these outcomes are based upon perceptions by prosecutors and triers-of-fact of the reliability of such memories after a very long interval.

4. THE EFFECTS OF DELAY ON LONG-TERM MEMORY

141

Based on limited research that has focused primarily on children, there are important differences related to content and linguistic structure, between reports of repeated events and reports of unique events. Perhaps one of the most forensically important implications of repeated experience with similar events is that instances become difficult to retrieve for both children and adults (Connolly & Lindsay, 2001; Farrar & Goodman, 1990, 1992; Fivush, 1984; Means & Loftus, 1991; Neisser, 1986a; Pearse, Powell, & Thomson, 2003; Price & Connolly, 2004). Moreover, compared with reports of unique events, reports of repeated events are less distinctly associated with a particular instance (Connolly & Lindsay, 2001; Price & Connolly, 2004), sometimes more suggestible about details that change across instances (Connolly & Lindsay, 2001; Connolly & Price, 2005; Price & Connolly, 2004; but see Powell et al., 1999; Powell & Roberts, 2002), and less consistent across reports (Connolly & Price, 2005; Banipal, Price, Gordon, & Connolly, 2005). In terms of the linguistic structure, compared with children who report a unique event, children who report an instance of a repeated event use the present tense more often, more impersonal pronouns (i.e., you), and a greater proportion of temporal markers (Fivush, 1984, 1997; Hudson & Nelson, 1986). One tentative conclusion that may be drawn from the limited research on repeated events is that increasing the length of the retention interval may strengthen the effects that have been attributed to repeated exposure to similar instances of the event. In particular, memory for repeated events may become more general and less particular over time. Slackman and Nelson (1984) reported that memory for a repeated event became more general and less specific over time. Furthermore, 6- to 7-year-olds recalled more details that changed across instances and correctly attributed more details to the target instance after 1 week than after 6 weeks (Powell & Thomson, 1996, 2002), and there were more intrusions from related experiences after 6 weeks than after 1 week (Powell & Thomson, 1996). For a more comprehensive review of research on experimental investigations of repeated events with children, see Pipe et al. (this volume). To the extent that repeated traumatic events demonstrate the same difficulties in specific event recall as the number of events increases (see Bifulco, Brown, Lillie, & Jarvis, 1997; Fivush, 2002; Pezdek & Taylor, 2002; Price & Connolly, 2004), the less viable is the argument that these kinds of events are uniquely retained in memory and not subject to the same reconstructive forces as other kinds of events. There has been some research with adults' memory reports of traumatic events that happened repeatedly; however, the interest was not in the repeated nature of the event, and so the data do not speak to that issue directly. However, it is the most relevant research available, so we review it here briefly. At the outset, we restricted this review to normal individuals recalling salient autobiographical events. For repeated events of impact, however, mental disorders can result, and the respondents may be diagnosed with particular mental disorders, such as Posttraumatic Stress Disorder (PTSD) and Dissociative Identity Disorder (DID). In both cases, the respondent will likely demonstrate memory complaints, such as an inability to recall important aspects of the event, re-experiencing symptoms like flashbacks or intrusive memories, distress about being unable to avoid thoughts about the traumatic event,

142

READ AND CONNOLLY

and, for DID, being unable to recall substantial components of current life. Case studies have examined these phenomena and diagnostic conditions, but, to our knowledge, the retention interval has rarely been the focus of interest by researchers (Brown et al., 1998; McNally, 2003). One serious obstruction to our understanding of the memory processes and capabilities of individuals who are seriously distressed by events and/or who have been diagnosed with serious mental disorder, including depression, is that they frequently participate in targeted therapeutic interventions that have recollection of the traumatic events as a necessary component of treatment. Clearly, if these individuals subsequently or concurrently become participants in research studies investigating the consistency of their recall over time, the therapeutic experience provides not only more (and uncontrolled) retrieval practice, but also opportunities for distortions to arise as a result of new contaminating information introduced by the therapist or other patients (cf. Lindsay & Read, 1994, 1995). Nonetheless, to explore the consistency of reports of traumatic events over long delays, we examine a few studies wherein psychological disorders are likely to be present. Southwick et al. (1997) had 59 U.S. Gulf War veterans answer a detailed questionnaire about highly traumatic combat experiences 1 month and again 2 years after returning home. On the delayed tests, 88% of participants changed at least one response and 61% changed two or more responses. Of the participants, 46% reported an incident at 1 month that was not reported at 2 years, and 70% reported an incident at 2 years that was not reported at 1 month. Roemer, Litz, Orsillo, Ehlich, and Friedman's (1998) sample of 460 soldiers completed a questionnaire within 1 year of returning to the United States and again 1-3 years later. The survey included measures of the frequency of particular peace-keeping stressors (e.g., performing dangerous duties, being fired at, seeing Somalis dying). There was a small yet significant increase in frequency scores on all questions from Time 1 to Time 2. A similar result was reported by King et al. (2000), who interviewed 2942 soldiers 5 days after they returned from the Gulf War. Participants answered 31 questions concerning exposure to war stressors. The same measures were administered 18 to 24 months later. Only 9% of the sample was completely consistent across time. Most of the inconsistencies involved changes from reporting no exposure to reporting exposure to specific stressors. Together, these studies provide evidence that memory for specific details associated with trauma is not fixed and may not even be particularly stable. In line with previous studies on repeated testing effects, however, the changes may possibly result from differential memory sampling of combat experiences over time and, to the extent that the event was discussed with others, incorporate new information that rendered details of the first interview inconsistent with those in a later interview. In the absence of single-test groups at early and late retention intervals, the answer is indeterminate. In summary, in spite of its forensic relevance, there has been very little research on reports of repeated events. The little research that exists supports the notion that reports of repeated events differ in several forensically relevant ways from reports of single events. For instance, compared with reports of single events, reports of repeated events are more general, less particular (in terms of content and syntax), sometimes more sug-

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

143

gestible, and less stable and consistent across repeated reports. There is evidence that at least some of these characteristics become more pronounced over time.

SUMMARY Despite the wide variation in the types of events studied, witnesses' initial responses to them, and variations in the research methodologies used, the overall impact of increasing the length of delay between event and recall is one of memory change. In the large majority of studies, those changes reflect a loss of access to information, and that loss is often accompanied by distortions in the reports provided, in the form of reduced amounts of information recalled, inconsistencies in the level of detail, or qualitative changes in the meaning and interpretation of aspects of the event or of the event itself. The time period between event and recall provides for a variety of post-event cognitive processes that, for any event and particularly for emotional events, may serve to maintain the memory representation or central aspects of it. The effects of testing, rumination, discussion, and rehearsal certainly fall into this category. The sword of cognitive processing is double-edged because these same protective activities may also serve to put the memory representation at risk to distortion and change, altering the memory representation such that it is no longer as faithful to the event as witnessed by the individual, particularly when the retention interval clearly exposes individuals to many and varied sources of information, as in media accounts of events. And the data from substantial other work in the field of memory leads us to believe that for witnessed autobiographical events, the longer the delay, the more likely it is that reconstructive forces will distort the recollection in a way that meets other needs and current conceptions of self. In addition to the standardized materials presented in laboratory work that is often constructed in such a way as to provide some level of emotional response and arousal in the participants, we also examined naturally occurring surprising, emotional, or threatening public events. Unfortunately, for the majority of these events, the absence of an objective record does not allow us to distinguish between accurate and inaccurate recollections of the event. Taken together, the results suggest the operation of the same fundamental memory processes in both emotional and mundane events. These fundamental properties of memory are the effects of cognitive processing at and following the event that serve to enhance the event's encoding in memory, constructive influences at the time of witnessing and reconstructive processes at retrieval reflected in changes and inconsistencies over time, losses of memory strength over time due to decay, and interference that may be protected to some degree by consolidation processes. As a result of the differential protective and detrimental consequences of post-event activities, it is not currently possible to predict for any event, or indeed, any interval, the exact amount and nature of memory change over time. The research literature suggests, however, that in the short term for witnessed events that are well encoded, the length of the interval is not likely to be dramatically influential in their recall. However, as encoding of the event is reduced and length of the delay increases, losses will be greater for peripheral than for central or "gist" information, and the quantitative pattern of these

144

READ AND CONNOLLY

losses is best described by a function for which greatest loss occurs relatively early, followed by much smaller losses of additional information over a much longer period of time. One such protective factor is the early recall of an event. Because the adjudication of facts in both civil and criminal matters virtually always requires multiple recalls or testimonies from the same witness, the initial test or recall opportunity may serve as a critical post-event activity that may have a protective function for. the memory representation against substantial loss over the following days, weeks, or months. Because any retrieval activity serves to reinforce both accurate and inaccurate contents of memory, such consistency across recalls or testimonies, it must be emphasized, is not synonymous with accuracy, and researchers suspect that there is a critical period during which the event is first consolidated, as an accurate or inaccurate representation of what happened. This early interval may also be a period in which there is considerable opportunity for contamination from other sources, and, if these sources are present, greater accuracy is likely to characterize the earliest possible recollection. It is also the case that with repeated testing or recalls, the research demonstrates that additional information about the event will generally be reported. Whereas our theoretical understanding of the fundamental processes of memory certainly anticipates these changes, for real-world (compared with laboratory) events in which recall opportunities are separated by long periods of time, the possibility of contamination from other sources such as the media adds considerable complexity to this expectation. The changes over recall opportunities also depend upon the type of interview, test, or recall opportunity: with free recall procedures, changes are anticipated; with directed specific or closed questions, however, similar changes are not anticipated. There are also situations in which a witness, or more likely a victim, has experienced multiple instances of the same kind of event. For these individuals, we anticipate that the ease of recall of any specific event from the series will decrease as the number of similar events increases. With a longer retention interval and a greater number of events, the more likely it is that a generic script memory will characterize the majority of single events. Some research suggests that with the use of time-line interview questions, recall for single events may be enhanced, but this is speculative and has not been investigated systematically.

RECOMMENDATIONS FOR LEGAL PROFESSIONALS AND INVESTIGATORS

Although many of the following recommendations are already components of best practices in the investigation, prosecution of, and defense against alleged events, the research reviewed here underscores their essential importance. Generally, when opportunities for adequate encoding existed, memory for the details of autobiographical events serves us very well: recall of information gathered within an unbiased interview is characterized to a much greater degree by accuracy rather than inaccuracy. However, because of the fundamental properties of human memory described above, all events are subject to error across the retention interval, and the variables discussed in this chapter are related to the overall rates of error.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

145

First, because there is usually no independent record or ground truth of naturally occurring witnessed events, the events' reconstructions must be based, at least in part, upon witness reports. It is essential that those reports be collected in such a way as to ensure their complete independence. Many crime scenes do not allow for such independence because witnesses may be friends or acquaintances who were together at the scene, who discuss the event prior to the arrival of police and the interview, who overhear other witness reports to the police, or read media accounts, or who discuss the event away from the scene and prior to a delayed interview. Such discussion should be anticipated. Delays from a few hours to several days certainly provide these additional opportunities for discussion and pursuit of the event in the media. Similar social interactions and further rumination about the event should be anticipated as the various phases of the legal process are initiated. At a minimum, witnesses should be canvassed prior to every interview about the extent to which they engaged in any of these activities. Second, although the first report may not be complete, it is more likely, when compared with later reports, that its components are accurate. Research demonstrates that when the first interview is delayed (on the order of a few days), its content appears to better match that of later reports, than does an immediate interview. This effect of consolidation on memory may well reflect early losses of relevant and probably peripheral detail that will not appear in any reports and/or new information gained from other sources. Therefore, apart from restrictions imposed by physical injury or psychological stress, reports should be collected by investigators as early as possible, but a short delay should not have a substantial effect on reports of central and gist details. Third, because additional testimony will be gathered, investigators need to be aware that with repeated interviews the contents will be reported with greater confidence, regardless of whether accuracy is maintained or reduced. Simply put, in the absence of contamination from external sources, increases in the accuracy level of recalled material is not anticipated over time. Furthermore, investigators should anticipate that, depending upon the distribution of interviews over time and the types of interview techniques, cumulative recall of details from a witness may increase; however, as the number of tests increases, the proportion of erroneous new information will also increase. Taken together, repeated testing will demonstrate inconsistencies in amount, level, and interpretation of details. Because research does not support a strong link between report inconsistencies and overall accuracy, the forensic significance of these inconsistencies may be quite limited. Nonetheless, to monitor changes in report quality, quantity, and confidence over time, we recommend complete videotaping of the interviews. Fourth, as to the number of interviews and statements, there is no evidence that more is better, either for adults or for children, and, for the latter, there is much evidence that more interviewing is worse. We recommend a small number of good and complete interviews (incorporating free narrative recall as well as follow-up general and more specific questions) within a fairly short period of time, because doing so will likely exhaust the information available and will be less likely to encourage the witness to recall information that was not encoded. Fifth, to date, neither beneficial nor detrimental effects of refreshing a witness's memory by reading over prior statements have been demonstrated. However, we anticipate greater benefits to consistency as the retention interval and the period between

146

READ AND CONNOLLY

statements increases in length. However, this should not be interpreted as necessarily reflecting accuracy. Sixth, the idea that memories for emotional events will be highly accurate and resistant to forgetting is seriously misleading. Error and reconstructive processes characterize all memory reports, and the witness's confidence in the report is a weak predictor of its accuracy (see Shaw et al., this volume). Whereas events of emotional impact may trigger more trace-protective processes, emotional events are also subject to error and reconstruction over the retention interval, particularly for peripheral details. As a result of attentional narrowing with increased arousal, stress, and fear, these details may have received little or no attention and therefore are permanently unavailable for recall. That they are reported in recall much later is more likely a reflection of reconstructive processes and the acquisition of new information from other sources. Seventh, whereas children do provide free recall reports that are comparable in their accuracy levels to those of adults (although the amount of reported detail may differ), children are more likely to be influenced by interview style and suggestive questioning, post-event social interactions with peers, and the meaning of the event to themselves. Finally, long-term retention of autobiographical memory has taken on particular forensic significance because more and more adults are reporting details of alleged events that happened in the distant past, such as child sexual abuse (Connolly & Read, 2003, 2005; Porter, Campbell, Birt, & Woodworth, 2003). In a small subset of these cases, the complainants report noncontinuous memory for the events, and for these, various psychological processes have been invoked to explain their nonavailability and subsequent recovery (see Smith & Gleaves, this volume). For the vast majority, however, the complainant testifies that she (or he) has always remembered the essential aspects of the disputed events. We anticipate that recall of these very long-term autobiographical events would demonstrate the same strengths and frailties as those memories investigated at much shorter intervals by researchers described in the present chapter. ACKNOWLEDGMENTS

The writing of this chapter was supported by funding from the National Sciences and Engineering Research Council of Canada to the first author. The authors thank Daniel Reisberg, Sarah Desmarais, and Heather Price for their comments on an earlier version of the chapter. The authors may be contacted at either [email protected] or [email protected]. REFERENCES Alpert, J. L., Brown, L. S., Ceci, S. J., Courtois, C. A., Loftus, E. F., & Ornstein, P. A. (1998). Final conclusions of the American Psychological Association working group of investigation of memories of childhood abuse. Psychology, Public Policy, and Law, 4, 933-940. Altmann, E. M. (2003). Reconstructing the serial order of events: A case study of September 11, 2001. Applied Cognitive Psychology, 17, 1067-1080. Anderson, S. J., Cohen, G., & Taylor, S. (2000). Rewriting the past: Some factors affecting the variability of personal memories. Applied Cognitive Psychology, 14, 435-454.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

147

Bahrick, H. P. (1998). Loss and distortion or autobiographical memory content. In C. P. Thompson, D. J. Herrmann, D. Bruce,]. D. Read, D. G. Payne, & M.P. Toglia (Eds.), Autobiographical memory: Theoretical and applied perspectives (pp. 69-78). Mahwah, NJ: Lawrence Erlbaum Associates. Bahrick, H. P. (2000). Long-term maintenance of knowledge. In E. Tulving & F. I. M. Craik (Eds.), The Oxford Handbook of Memory (pp. 347-362). New York: Oxford University Press. Bahrick, H. P., Bahrick, P. 0., & Wittlinger, R. P. (1975). Fifty years of memory for names and faces: A cross-sectional approach. Journal of Experimental Psychology: General, 104, 54-75. Bahrick, L. E., Parker,]. F., Fivush, R., & Levitt, M. (1998). The effects of stress on young children's memory for a natural disaster. Journal of Experimental Psychology: Applied, 4, 308-331. Banipal, R., Price, H. L., Gordon, H. M., & Connolly, D. A. (2005, May). Children's report consistency as a function of event frequency. Poster presented at the seventh annual meeting of the NorthWest Cognition and Memory Conference, Bellingham, Washington. Barclay, C. R., & DeCooke, P. A. (1988). Ordinary everyday memories: Some of the things of which selves are made. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered: Ecological and traditional approaches to the study of memory (pp. 91-125). New York: Cambridge University Press. Bartlett, F. C. (1932). Remembering: A Study of Experimental and Social Psychology, Cambridge: Cambridge University Press. Belli, R. F., Schuman, H., &Jackson, B. (1997). Autobiographical misremembering: John Dean is not alone. Applied Cognitive Psychology, 11, 187-209. Berman, G. L. & Cutler, B. L. (1996). Effects of inconsistencies in eyewitness testimony and mock-juror judgments. Journal of Applied Psychology, 81, 170-177. Berman, G. L., Narby, D.]., & Cutler, B. L. (1995). Effects of inconsistent eyewitness statements on mock-jurors' evaluations of the eyewitness, perceptions of defendant culpability and verdicts. Law and Human Behavior, 19, 79-88. Bernsten, D. (2001). Involuntary memories of emotional events: Do memories of traumas and extremely happy events differ? Applied Cognitive Psychology, 15, S135-S158. Bernsten, D. (2002). Tunnel memories for autobiographical events: Central details are remembered more frequently from shocking than from happy experiences. Memory & Cognition, 30, 1010-1020. Berns ten, D., & Rubin, D. (2002). Emotionally charged autobiographical memories across the life span: The recall of happy, sad, traumatic, and involuntary memories. Psychology and Aging, 17, 636-652. Bidrose, S., & Goodman, G. S. (2000). Testimony and evidence: A scientific case study of memory for child sexual abuse. Applied Cognitive Psychology, 14, 197-213. Bifulco, A., Brown, G. W., Lillie, A., &Jarvis, J. (1997). Memories of childhood neglect and abuse: Corroboration in a series of sisters. Journal of Child Psychology and Psychiatry, 38, 365-374. Bluck, S. Levine, L. J., & Laulhere, T. M. (1999). Autobiographical remembering and hypermnesia: A comparison of older and younger adults. Psychology and Aging, 14, 671-682. Bornstein, B. H., Liebel, L. M., & Scarberry, N.C. (1998). Repeated testing in eyewitness memory: A means to improve recall of a negative emotional event. Applied Cognitive Psychology, 12, 119-131. Brainerd, C.]., & Reyna, V. F. (1990). Can the age x learnability interactions explain the development of forgetting? Developmental Psychology, 26, 194-204. Brainerd, C.]., & Reyna, V. F. (1995). Learning rate, learning opportunities, and the development of forgetting. Developmental Psychology, 31, 251-262. Brewer, N., Potter, R., Fisher, R. F., Bond, N., & Luszcz, M.A. (1999). Beliefs and data on the relationship between consistency and accuracy of eyewitness testimony. Applied Cognitive Psychology, 13, 297-313. Brewin, C. R., Andrews, B., & Gotlib, I. H. (1993). Psychopathology and early experience: A reappraisal of retrospective reports. Psychological Bulletin, 113, 82-98.

148

READ AND CONNOLLY

Brewin, C. R., McNally, R. J., & Taylor, S. (2004). Point-counterpoint: Two views on traumatic memories and posttraumatic stress disorder. Journal of Cognitive Psychotherapy: An International Quarterly, 18, 99-114. Brown, D., Scheflin, A W., & Hammond, D. C. (1998). Memory, trauma treatment, and the law. New York: Norton. Bruck, M., & Ceci, S., J. (1995). Amicus brief for the case of State of New Jersey v. Margaret Kelly Michaels presented by committee of concerned social scientists. Psychology, Public Policy, and Law, I, 272-322. Bryne, C. A., Hyman, I. E., Jr., & Scott, K. L. (2001). Comparisons of memories for traumatic events and other experiences. Applied Cognitive Psychology, 15, S119-S133. Burgwyn-Bailes, E, Baker-Ward, L., Gordon, B. N., & Ornstein, P. A (2001). Children's memory for emergency medical treatment after one year: The impact of individual difference variables on recall and suggestibility. Applied Cognitive Psychology, 15, S25-S48. Burt, C. D. B., Kemp, S., & Conway, M. (2001). What happens if you retest autobiographical memory 10 years on? Memory & Cognition, 29, 127-136. Ceci, S. J., & Bruck, M. (1993). The suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113, 403-439. Ceci, S. J., Huffman, M. L. C., Smith, E., & Loftus, E. F. (1994). Repeatedly thinking about a nonevent: source misattributions among preschoolers. Consciousness & Cognition, 3, 388-407. Christianson, S.-A. (1989). Flashbulb memories: Special, but no so special. Memory & Cognition, 17, 435-443. Christianson, S-A. (1992a). Emotional stress and eyewitness memory: A critical review. Psychological Bulletin, 112, 284-309. Christianson, S.-A. (1992b). Remembering emotional events: Potential mechanisms. In S.-A. Christianson (Ed.), The handbook of emotion and memory: Research and theory (pp. 307-340). Hillsdale, NJ: Lawrence Erlbaum Associates. Christianson, S.-A., & Loftus, E. F. (1991). Remembering emotional events: the fate of detailed information. Cognition and Emotion, 5, 81-108. Christianson, S.-A., Goodman, J., & Loftus, E. F. (1992). Eyewitness memory for stressful events: Methodological quandaries and ethical dilemmas. In S.-A. Christianson (Ed.), The handbook of emotion and memory: Research and theory (pp. 217-241). Hillsdale, NJ: Lawrence Erlbaum Associates. Christianson, S.-A., & Engelberg, E. (1999). Memory and emotional consistency: The MS Estonia ferry disaster. Memory, 7, 471-482. Christianson, S.-A., & Hiibinette, B. (1993). Hands up! A study of witnesses' emotional reactions and memories associated with bank robberies. Applied Cognitive Psychology, 7, 365-379. Christianson, S.-A., & Loftus, E. F. (1990). Some characteristics of people's traumatic memories. Bulletin of the Psychonomic Society, 28, 195-198. Christianson. S-A., & Safer, M. A (1995). Emotional events and emotions in autobiographical memories. In D.C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 218-243). New York: Cambridge University Press. Cohn, M. A., Mehl, M. R., & Pennebaker, J. W. (2004). Linguistic markers of psychological change surrounding September 11, 2001. Psychological Science, 15, 687-693. Connolly, D. A., & Lindsay, D. S. (2001). The Influence of suggestions on children's reports of a unique experience versus an instance of a repeated experience. Applied Cognitive Psychology, 15, 205-223. Connolly, D. A., & Price, H. (2005, March). Consistency of reports across three interviews with a woman who experienced five armed robberies in the 1970's. Paper presentation at the American Psychology-Law Society Meetings, La Jolla, CA. Connolly, D. A., & Read, J.D. (2003). Remembering historic child sexual abuse. Criminal Law Quarterly, 47, 438-480.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

149

Connolly, D. A., & Read, J. D. (2005). Delayed prosecutions of historic child sexual abuse: Analyses of 2064 Canadian criminal complaints Manuscript submitted for publication. Conway, M.A. (1996). Autobiographical knowledge and autobiographical memories. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 67-93). Cambridge, MA: Cambridge University Press. Conway, M.A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261-288. Crombag, H. F. M., Wagenaar, W. A., & van Koppen, P. ]. (1996). Crashing memories and the problem of'source monitoring'. Applied Cognitive Psychology, 10, 95-104. Cutshall, J. L., & Yuille,]. C. (1989). Field studies of eyewitness memory of actual crimes. In D. C. Raskin (Ed.), Psychological methods in criminal investigation and evidence (pp. 97-124). New York: Springer. Davis, D., Lopez, P., Koyama, M., White, B., Knaack, D., Kusal, T., & Bailey, D. (2005, May).

Memory for threats in conversation enhanced by later knowledge of violence between participants. Paper presentation at the American Psychological Society, Los Angeles, CA. Deffenbacher, K. A., Bomstein, B. H., Penrod, S.D., & McGorty, E. K. (2004). A meta-analytic review of high stress on eyewitness memory. Law and Human Behavior, 28, 687-706. Dent, H. R., & Stephenson, G. M. (1979). An experimental study of the effectiveness of different techniques of questioning child witnesses. British]oumal of Social and Clinical Psychology, 18,41-51. Dunning, D., &Stem, L. B. (1992). Examining the generality of eyewitness hypermnesia: A close look at time delay and question type. Applied Cognitive Psychology, 6, 643-657. Ebbinghaus, H. (1885/1913). Memory: A contribution to experimental psychology [H. A. Ruger & C. E. Bussenious, Trans.]. New York: Columbia University Teachers College. Edelstein, R. S., Alexander, K. W., Goodman, G. S., & Newton, J. W. (2004). Emotion and eyewitness memory. In D. Reisberg & P. Hertel (Eds.), Memory and emotion (pp. 308-346). Oxford: Oxford University Press. Edery-Halpem, G., & Nachson, I. (2004). Distinctiveness in flashbulb memory: Comparative analysis offive terrorist attacks. Memory, 12, 147-157. Engelberg, E., & Christianson, S.-A. (2002). Stress, trauma, and memory. In M. L. Eisen, J. A. Quas, &G. S. Goodman, (Eds.) Memory and suggestibility in the forensic interview (pp. 143-163). Mahwah, NJ: Lawrence Erlbaum Associates Er, N. (2003). A new flashbulb memory model applied to the Marmara earthquake. Applied Cognitive Psychology, 17, 503-517. Farrar, M. ]., & Goodman, G. S. (1990). Developmental differences in the relation between scripts and episodic memory: Do they exist? In R. Fivush & ]. A. Hudson (Eds.), Knowing and remembering in young children (pp. 30-64). New York: Cambridge University Press. Farrar, M. ]., & Goodman, G. S. (1992). Developmental changes in event memory. Child Development, 63, 173-187. Finkenauer, C., Luminet, 0., Gisle, L., El-Ahmadi, A., van der Linden, M., & Philippot, P. (1998). Flashbulb memories and the underlying mechanisms of their formation: Towards an emotional-integrative model. Memory & Cognition, 26, 516-531. Fisher, R. P., & Cutler, B. L. (1995). The relation between consistency and accuracy of eyewitness testimony. In G. Davies, S. Lloyd-Bostock, M. McMurran, & C. Wilson (Eds), Law and criminal justice: International developments in research and practice. Berlin: De Gruyter. Fisher, R. P., & Geiselman, R. E. ( 1992). Memory enhancing techniques for investigative interviewing: the cognitive interview. Springfield, IL: Charles C. Thomas. Fivush, R. (1984). Learning about school: The development of kindergartner's school script. Child Development, 55, 1697-1709. Fivush, R. ( 1997). Event memory in early childhood. In N. Cowan (Ed.), The development of memory in childhood (pp. 139-161). U.K: Psychology Press.

ISO

READ AND CONNOLLY

Fivush, R. (2002). Scripts, schemas, and memory of trauma. In N. L. Stein, P. J. Bauer, & M. Rabinowitz (Eds.), Representation, memory, and development: Essays in honor of Jean Mandler (pp. 53-7 4). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., & McDermott Sales, J. (2004). Children's memories of emotional events. In D. Reisberg & P. Hertel (Eds.), Memory and emotion (pp.242-271). Oxford: Oxford University Press. Fivush, R., McDermott Sales, J., Goldberg, A., Bahrick, L., & Parker, J. (2004). Weathering the storm: Children's long-term recall of Hurricane Andrew. Memory, 12, 104-118. Fivush, R., Peterson, C., & Schwarzmueller, A. (2002). Questions and answers: The credibility of child witnesses in the context of specific questioning techniques. In M. L. Eisen, J. A. Quas, & Goodman, G. S. (Eds.), Memory and suggestibility in the forensic interview (pp. 331-354). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., & Schwarzmueller, A. (1998). Children remember childhood: Implications for childhood amnesia. Applied Cognitive Psychology, 12, 455-473. Flin, R., Boon, J., Knox, A., & Bull, R. (1992). The effect of a five-month delay on children's and adults' eyewitness memory. British Journal of Psychology, 83, 323-336. Goodman, G. S., Batterman-Faunce, J. M., Schaaf, J. M., & Kenney, R. (2002). Nearly 4 years after an event: Children's eyewitness memory and adults' perceptions of children's accuracy. Child Abuse & Neglect, 26, 849-884. Greenberg, D. L. (2004). President Bush's 'flashbulb memory' of 9/11/01. Applied Cognitive Psychology, 18, 363-370. Greenberg, M.S., Westcott, D. R., & Bailey, S. E. (1998). When believing is seeing: The effect of scripts on eyewitness memory. Law and Human Behavior, 22, 685-694. Haber, L., & Haber, R.N. (1998). Criteria for judging the admissibility of eyewitness testimony oflong past events. Psychology, Public Policy, and Law, 4(4), 1135-1159. Hardt, J., & Rutter, M. (2004). Validity of adult retrospective reports of adverse childhood experiences: Review of the evidence. Journal of Child Psychology and Psychiatry, 45, 260-273. Healy, A. F., & Sinclair, G. P. (1996). The long-term retention of training and instruction. In E. L. Bjork & R. A. Bjork (Eds.), Memory (pp. 525-564). Toronto: Academic Press Henry, B., Moffitt, T. E., Caspi, A., Langley, J., & Silva, P. A. (1994). On the "remembrance of things past": A longitudinal evaluation of the retrospective method. Psychological Assessment, 6, 92-101. Holst, V. F., & Pezdek, K. (1992). Scripts for typical crimes and their effects on memory for eyewitness testimony. Applied Cognitive Psychology, 6, 573-587. Hudson, J., & Nelson, K. (1986). Repeated encounters of a similar kind: Effects of familiarity on children's autobiographic memory. Cognitive Development, 1, 253-2 71. Hulse, L. M., & Memon, A. (2005). Fatal impact? The effects of emotional arousal and weapon presence on police officers' memories for a simulated crime. Paper Under Review. Hyman, I. E., & Loftus, E. F. (1998). Errors in autobiographical memory. Clinical Psychology Review, 18,933-947. Karlsson, I., & Christianson, S.-A. (2003). The phenomenology of traumatic experiences in police work. Policing: An International Journal of Police Strategies & Management, 26, 419-438. Kassin, S. M. (2005). On the psychology of confessions: Does innocence put innocents at risk. American Psychologist, 60, 215-228. King, D. W., King, L.A., Erickson, D. J., Huang, M. T., Sharkansky, E. J., & Wolfe, J. (2000). Posttraumatic stress disorder and retrospectively reported stressor exposure: a longitudinal prediction model. Journal of Abnormal Psychology, 109, 624-633. Kvavilashvili, L., Mirani, J., Schlagman, S., & Kornbrot, D. E. (2003). Comparing flashbulb memories of September 11 and the death of Princess Diana: Effects of time delays and nationality. Applied Cognitive Psychology, 17, 1017-1031. La Rooy, D., Pipe, M. E., & Murray, J. E. (2005). Reminiscence and hypermnesia in children's eyewitness memory. Journal of Experimental Child Psychology, 90, 235-254.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

151

Lee, P. ]., & Brown, N. R. (2003). Delay related changes in personal memories for September 11, 2001. Applied Cognitive Psychology, 17, 1007-1016. Levine, L. (1997). Reconstructing memory for emotions. Journal of Experimental Psychology: General, 126, 165-177. Levine, L. J., & Safer, M. A. (2002). Sources of bias in memory for emotions. Psychological Science, 11,169-173. Lindsay, D. S., Hagen, L., Read, J.D., Wade, K. A., & Garry, M. (2004) True photographs and false memories. Psychological Science, 15, 149-154. Lindsay, D. S., & Read, J.D. (1994). Psychotherapy and memories of childhood sexual abuse: A cognitive perspective. Applied Cognitive Psychology, 8, 281-338. Lindsay, D. S., & Read, J.D. (1995). Recovered memories of childhood sexual abuse: Scientific evidence and public, professional, and personal issues. Psychology, Public Policy and Law, 1, 846-908. Lindsay, D. S., & Read, J.D. (2005, in press). Adults' memories oflong-past events. In L. G. Nilsson (Ed.), Memory and society: Psychological perspectives. Routledge and Psychology Press. Luminet, 0., Curci, A., Marsh, E. J., Wessel, I., Constantin, T., Gencoz, F., & Yogo, M. (2004). The cognitive, emotional, and social impacts of the September 11 attacks: Group differences in memory for the reception context and the determinants of flashbulb memory. The Journal of General Psychology, 131, 197-224. Maughan, B., & Rutter, M. (1997). Retrospective reporting of childhood adversity: Issues in assessing long-term recall. Journal of Personality Disorders, 11, 19-33. McCloskey, M., Wible, C. G., & Cohen, N. ]. (1988). Is there a special flashbulb-memory mechanism? Journal of Experimental Psychology: General, 117, 171-181. McNally, R. ]. (2003). Remembering trauma. Cambridge, MA: Belknap Press of Harvard University Press. Means, B., & Loftus, E. F. (1991). When personal memory repeats itself: Decomposing memories for recurring events. Applied Cognitive Psychology, 5, 2297-318. Meissner, C. A., Sporer, S. L., & Schooler,]. W. (2006). Person descriptions as eyewitness evidence. In R. C. L. Lindsay, D. F. Ross, J.D. Read, & M.P. Toglia (Eds.), Handbook of eyewitness psychology: Volume II. Memory for people. Mahwah, NJ: Lawrence Erlbaum Associates. Nachson, I., & Zelig, A. (2003). Flashbulb and factual memories: The case of Rabin's assassination. Applied Cognitive Psychology, 17, 519-531. Neisser, U. (1986a). Nested structure in autobiographical memory. In D. C. Rubin (Ed.), Autobiographical memory (pp. 71-81). Cambridge: Cambridge University Press. Neisser, U. (1986b). Remembering Pearl Harbor: Reply to Thompson and Cowan, Cognition, 23, 285-286. Neisser, U. (2003). New directions for flashbulb memories: Comments on the ACP Special Issue. Applied Cognitive Psychology, 17, 1149-115 5. Neisser, U., & Harsch, N. (1992). Phantom flashbulbs: False recollections of hearing the news about Challenger. In E. Winograd and U. Neisser (Eds.), Affect and accuracy in recall: Studies of"flashbulb" memories (pp. 9-31). New York: Cambridge University Press. Neisser, U., & Libby, L. K. (2000). Remembering life experiences. In E. Tulving & F. I. M. Craik (Eds.), Oxford handbook of memory (pp. 315-332). Oxford: Oxford University Press. Neisser, U., Winograd, E., Bergman, E. T., Schreiber, C. A., Palmer, S. E., & Weldon, M.S. (1996). Remembering the earthquake: Direct experience vs. hearing the news. Memory, 4, 337-357. Noice, H., & Noice, T. (1999). Long-term retention of theatrical roles. Memory, 7, 357-382. Orbach, Y., Lamb, M. E., Sternberg, K. ]., Williams, J. M.G., & Dawud-Noursi, S. (2001). The effect of being a victim or witness of family violence on the retrieval of autobiographical memories. Child Abuse & Neglect, 25, 1427-1437. Ornstein, P. A., Ceci, S. ]., & Loftus, E. F. (1998). Adult recollections of childhood abuse: Cognitive and developmental perspectives. Psychology, Public Policy, and Law, 4(4), 1025-1051.

152

READ AND CONNOLLY

Ornstein, P. A., Shapiro, L. B., Clubb, P. A., Follmer, A., & Baker-Ward, L. (1997). The influence of prior knowledge on children's memory for salient medical experiences. InN. L. Stein, P. A Ornstein, B. Tversky, & C. Brainerd (Eds.), Memory for everyday and emotional events (pp. 83-106). Mahwah, NJ: Lawrence Erlbaum Associates. Ost, J., Vrij, A., Costall, A., & Bull, R. (2002). Crashing memories and reality monitoring: Distinguishing between perceptions, imaginations, and 'false memories'. Applied Cognitive Psyclwlogy, 16, 125-134. Parks, T. E. (1999). On one aspect of the evidence for recovered memories. American]oumal of Psychology, 112, 365-370. Parker, J. F., Bahrick, L., Lundy, B., Fivush, R., & Levitt, M. (1998). Effects of stress on children's memory for a natural disaster. In C. P. Thompson, D. J. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M. P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 31-54). Mahwah, NJ: Lawrence Erlbaum Associates. Pasupathi, M. (2001). The social construction of the personal past and its implications for adult development. Psychological Bulletin, 127, 651-672. Peace, K. A., & Porter, S. (2004). A longitudinal investigation of the reliability of memories for trauma and other emotional experiences. Applied Cognitive Psychology, 18, 1143-1160. Pearse, S. L., Powell, M. B., & Thomson, D. M. (2003). The effect of contextual cues on children's ability to remember an occurrence of a repeated event. Legal and Criminological Psychology, 8, 39-50. Pennebaker, J. W., & Harber, K. D. (1993). A social stage model of collective coping: The Lorna Prieta earthquake and the Persian GulfWar.]oumal of Social Issuers, 49, 125-145. Peterson, C. (2002). Children's long-term memory for autobiographical events. Developmental Review, 22, 370-402. Peterson, C., Moores, L., & White, G. (2001). Recounting the same events again and again: Children's consistency across multiple interviews. Applied Cognitive Psychology, 15, 353371. Peterson, C., & Tizzard-Drover, T. (2005). When initial interviews are delayed a year: Effect on children's 2-year recall. Law and Human Behavior, 29, 527-541. Peterson, C., & Whalen, N. (2001). Five years later: Children's memory for medical emergencies. Applied Cognitive Psychology, S7-S24. Pezdek, K. (2003). Event memory and autobiographical memory for the events of September 11, 2001. Applied Cognitive Psychology, 17, 1033-1046. Pezdek, K., & Taylor, J. (2002). Memory for traumatic events in children and adults. In M. L. Eisen, J. AQuas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 165-183). Mahwah, NJ: Lawrence Erlbaum Associates. Pillemer, D. B. (1984). Flashbulb memories of the assassination attempt on President Reagan. Cognition, 16, 63-80. Pillemer, D. B. (1992). Remembering personal circumstances: A functional analysis. In E. Winograd & U. Neisser (Eds.), Affect and accuracy in recall: Studies of"fiashbulb" memories (pp. 236273). New York: Cambridge University Press. Pillemer, D. B. (2001). Momentous events and the life story. Review of General Psychology, 5, 123-134 Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. Poole, D. A., & Lindsay, D. S. (2002). Children's suggestibility in the forensic context. In M. L. Eisen, J. AQuas, & Goodman, G. S. (Eds.), Memory and suggestibility in the forensic interview (pp. 355-382). Mahwah, NJ: Erlbaum Poole, D. A., & White, L. T. (1991). Effects of question repetition on the eyewitness testimony of children and adults. Developmental Psychology, 27, 975-986. Poole, D. A., & White, L. T. (1993). Two years later: Effect of question repetition and retention interval on the eyewitness testimony of children and adults. Developmental Psychology, 29, 844-853.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

153

Porter, S., & Birt, A. R. (2001). Is traumatic memory special? A comparison of traumatic memory characteristics with memory for other emotional life experiences. Applied Cognitive Psychology, 15, S101-S117. Porter, S., Campbell, M.A., Birt, A. R., & Woodworth, M. T. (2003). "He said, She said": A psychological perspective on historical memory evidence in the courtroom. Canadian Psychology, 44, 190-206. Porter, S., Yuille, J. C., & Lehman, D. R. (1998). The nature of real, implanted, and fabricated memories for emotional childhood events: Implications for the recovered memory debate. Law and Human Behavior, 23, 517-537. Powell, M. B., Roberts, K. P., & Ceci, S. J. (1999). The effects of repeated experience on children's suggestibility. Developmental Psychology, 35, 1462-1477. Powell, M. B., & Roberts, K. P. (2002). The effect of repeated experience on children's suggestibility across two question types. Applied Cognitive Psychology, 16, 367-386. Powell, M. B., & Thomson, D. M. (1996). Children's memory of an occurrence of a repeated event: Effects of age, repetition, and retention interval across three question types. Child Development, 67, 1988-2004. Powell, M., & Thomson, D. (2002). Children's memories for repeated events. In H. L. Westcott, G. M. Davies, & R. H. C. Bull (Eds.), Children's testimony: A handbook of psychological research and forensic practice (pp. 69-82). Chichester, UK: Wiley. Price, H. L., & Connolly, D. A. (2004). Event frequency and children's suggestibility: A study of cued recall responses. Applied Cognitive Psychology, 18, 809-821. Pynoos, R. S., & Nader, K. (1989). Children's memory and proximity to violence. Journal of the American Academy of Child and Adolescent Psychiatry, 28, 236-241. Pynoos, R. S., Steinberg, A.M., & Aronson, L. (1997). Traumatic experiences: The early organization of memory in school-age children and adolescents. In P. S. Appelbaum, L. A. Uyehara, & M. R. Elin Trauma and memory: Clinical and legal controversies (pp. 272-289). Oxford: Oxford University Press. Reyna, V. F., & Brainerd, C. J. (1995). Fuzzy-Trace theory: An interim synthesis. Learning and Individual Differences, 7, 1-75. Read, J.D. (1997). Memory issues in the diagnosis of unreported trauma. In J. D. Read & D. S. Lindsay (Eds.), Recollections of Trauma: Scientific evidence and clinical practice (pp. 79-1 00). New York: Plenum Press. Read, J.D., & Barnsley, R. H. (1977). Remember Dick and Jane: Memory for elementary school readers. Canadian]ournal of Behavioural Science, 9, 361-370. Read, J.D., Lindsay, D. S., & Nicholls, T. (1998). The relationship between accuracy and confidence in eyewitness identification studies: Is the conclusion changing? In C. P. Thompson, D. Herrmann, J. D. Read, D. Bruce, D. Payne, & M.P. Toglia (Eds.). Eyewitness memory: Theoretical and applied perspectives (pp. 107-130). Mahwah, N.J.: Erlbaum. Read, J.D., Tollestrup, P., Hammersley, R. H., McFadzen, E., & Christensen, A. (1990). The unconscious transference effect: Are innocent bystanders ever misidentified. Applied Cognitive Psychology, 4, 3-31. Roediger, H. L., McDermott, K. B., & Goff, L. M. (1997). Recovery of true and false memories: paradoxical effects of repeated testing. In M.A. Conway (Ed.), Recovered memories and false memories (pp. 118-149). Oxford: Oxford University Press. Roemer, L., Litz, B. T., Orsillo, S.M., Ehlich, P. J., & Friedman, M. J. (1998). Increases in retrospective accounts of war-zone exposure over time: The role ofPTSD symptom severity. Journal of Traumatic Stress, II, 597-605. Ross, M. (1989). Relation of implicit theories to the construction of personal histories. Psychological Review, 96(2), 341-357. Ross, M. (1997). Validating memories. InN. L. Stein, P. A. Ornstein, B. Tversky, & C. J. Brainerd (Eds.), Memory for everyday and emotional events (pp. 49-82). Mahwah, NJ: Erlbaum. Rubin, D. C. (1982). On the retention function for autobiographical memories. Journal of Verbal Learning and Verbal Behavior, 21, 21-38.

154

READ AND CONNOLLY

Rubin, D. C. (2000). The distribution of early childhood memories. Memory, 8, 265-269. Rubin, D. C., & Wenzel, A. E. (1996). One hundred years of forgetting: A quantitative description of retention. Psychological Review, 103, 734-760. Rubin, D. C., Schrauf, R. W., & Greenberg, D. L. (2004). Stability in autobiographical memory. Memory, 12, 715-721. Scrivner, E., & Safer, M.A. (1988). Eyewitnesses show hypermnesia for details about a violent event. Journal of Applied Psychology, 73, 371-377. Schacter, D. L. (1996). Searching for memory: the brain, the mind, and the past. New York: Basic Books. Shelach, L., & Nachson, I. (2001). Memory of Auschwitz survivors. Applied Cognitive Psychology, 15, 119-132. Schmidt, H. G., Peeck, V. H., Paas, F., & Breukelen, G. J.P. (2000). Remembering the street names of one's childhood neighbourhood: A study of very long-term retention. Memory, 8, 37-49. Schmolck, H., Buffalo, E. A., & Squire, L. R. (2000). Memory distortions develop over time: Recollections of the 0. J. Simpson verdict after 15 and 32 months. Psychological Science, 11, 39-45. Shobe, K. K., & Kihlstrom, J. F. (1997). Is traumatic memory special? Current Directions in Psychological Science, 6(3), 70-74. Shum, M. S. (1998). The role of temporal landmarks in autobiographical memory processes. Psychological Bulletin, 124, 423-442. Slackman, E., & Nelson, K. (1984). Acquisition of an unfamiliar script story form by young children. Child Development, 55,329-340. Smith, M. C., Bibi, U., & Sheard, D. E. (2003). Evidence for the differential impact of time and emotion on personal and event memories for September 11, 2001. Applied Cognitive Psychology, 17, 1017-1032. Stanny, C. J., &Johnson, T. C. (2000). Effects of stress induced by a simulated shooting on recall by police and citizen witnesses. American Journal of Psychology, 113, 359-386. Southwick, S. M., Morgan, C. A., Nicolaou, A. L., & Charney, D. S. (1997). Consistency of memory for combat-related traumatic events in veterans of Operation Desert Storm. AmericanJournalofPsychiatry, 154,173-177. Talarico, J. M., & Rubin, D. C. (2003). Confidence, not consistency, characterizes flashbulb memories. Psychological Science, 14, 455-461. Tekcan, A. I. (1999). Recall, consistency, and determinants of flashbulb memories. Dissertation Abstracts International: Section B: The Sciences and Engineering, 60 (1-B), 0835. Tekcan, A. I., Ece, B., Giilgoz, S., & Er, N. (2003). Autobiographical and event memory for 9/11: Changes across one year. Applied Cognitive Psychology, 17, 1057-1066. Terr, L. C. (1983). Chowchilla revisited: The effects of psychic trauma four years after a school bus kidnapping. American Journal of Psychiatry, 140, 1543-1550. Thompson, C. P., Skowronski, J. J., Larsen, S. R., & Betz, A. L. (1996). Autobiographical memory: Remembering what and remembering when. New Jersey: Erlbaum. Tizzard-Drover, T., & Peterson, C. (2004). The influence of an early interview on long-term recall: A comparative analysis. Applied Cognitive Psychology, 18, 727-744. Tollestrup, P. A., Turtle, J. W., & Yuille, J. C. (1994). Actual victims and witnesses to robbery and fraud: An archival analysis. In D. F. Ross, J.D. Read, & M.P. Toglia (Eds.), Adult eyewitness testimony: Current trends and developments (pp. 144-160). New York: Cambridge University Press. Tuckey, M. R., & Brewer, N. (2003). The influence of schemas, stimulus ambiguity, and interview schedule on eyewitness memory over time. Journal of Experimental Psychology: Applied, 9, 101-118. Turtle, J. W., & Yuille, J. C. (1994). Lost but not forgotten details: Repeated eyewitness recall leads to reminiscence but not hypermnesia. Journal of Applied Psychology, 79, 260-2 71.

4.

THE EFFECTS OF DELAY ON LONG-TERM MEMORY

ISS

Vaillant, G. E. (1977). Adaptation to life. Boston: Little, Brown. Wagenaar, W. A., & Groeneweg, J. (1990). The memory of concentration camp survivors. Applied Cognitive Psychology, 4, 77-87. Weaver, C. A., III. (1993). Do you need a "flash" to form a flashbulb memory? Journal of Experimental Psychology: General, 122, 39-46. Weaver, C. A., III, & Krug, K. S. (2004). Consolidation-like effects in flashbulb memories: Evidence from September 11, 2001. American Journal of Psychology, 117, 517-530. Westcott. H. L., Davies, G. D., & Bull, R. H. C. (2002). Children's testimony: A handbook of psychological research and forensic practice. Chichester, UK: Wiley. White, R. (2002). Memory for events after twenty years. Applied Cognitive Psychology, 16, 603-612. Whittlesea, B. W. A. (1997). Production, evaluation and preservation of experiences: Constructive processing in remembering and performance tasks. In D. L. Medin (Ed.), The psychology oflearningandmotivation (Vol. 37, pp. 211-264). New York: Academic. Wixted, J. T. (2004). On common ground: Law of retrograde amnesia. Psychological Review, 111, 864-879. Winningham, R. G., Hyman, I. E., Jr., & Dinnel, D. L. (2000). Flashbulb memories? The effects of when the initial report was obtained. Memory, 8, 209-216. Wright, D. B. (1993). Recall of the Hillsborough Disaster over time: Systematic biases of 'flashbulb' memories. Applied Cognitive Psychology, 7, 129-138. Yarmey, A. D. (2004). Eyewitness recall and photo identification: a field experiment. Psychology, Crime & Law, 10, 53-68. Yarmey, A. D. (1998). Person identification in showups and lineups. In C. P. Thompson, D. J. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 131-154). Mahwah: N.J.: Erlbaum. Yarmey, A. D., Jacob, J., & Porter, A. (2002). Person recall in field settings. Journal of Applied Social Psychology, 32, 2354-2367. Yuille, J. C., & Cutshall, J. L. (1986). A case study of eyewitness memory of a crime. Journal of Applied Psychology, 71, 291-301. Yuille, J. C., & Cutshall, J. (1998) The impact of traumatic events on eyewitness memory. In C. P. Thompson, D. J. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 15 5-1 78) . Mahwah: N.J.: Erlbaum. Yuille, J. C., Davies, G., Gibling, F., Marxsen, D., & Porter, S. (1994). Eyewitness memory for police trainees for realistic role plays. Journal of Applied Psychology, 79, 931-036. Yuille, J. C., & Tollestrup, P. A. (1990). Some effects of alcohol on eyewitness memory.Journal of Applied Psychology, 75, 268-273.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

5 Alibis in Criminal Investigations and Trials Tara M. Burke and John w. TUrtle Ryerson University

Elizabeth A. Olson University of Wisconsin, Whitewater

THE PSYCHOLOGY OF ALIBIS

Imagine you are at your desk working one afternoon, when two police officers suddenly appear and request that you answer a few questions. They begin with, "Where were you a week ago last Sunday, at 2 o'clock in the afternoon?" How quickly could you recall the relevant information? How accurate would you be? If you hesitate, how will you be per~ ceived? What if you provide your statement and then realize later that you were mis~ taken? How would (or should) your responses be interpreted by the police officers? When does your statement officially become an alibi? No matter what the specific criminal case, chances are that several people-whether innocent or guilty-will end up providing statements to the police that need to be in~ vestigated. When a statement is offered as evidence to implicate or exonerate an indi~ vidual, it becomes an alibi. Definitions of the term alibi suggest several different uses, ranging from a specific legal strategy in which a suspect tries to show that he or she could not have committed the crime because he or she was somewhere else at the time (Good~ erson, 1977), to a more general notion of an excuse or plea that one was somewhere other than the scene of a crime (Oxford English Dictionary, 1984). Although it would seem that psychology is in an ideal position to investigate many legal and memory issues related to alibis, systematic research in the area is relatively new (e.g., Burke & Turtle, 2003; Culhane & Hosch, 2004; Olson & Wells, 2004). This chapter reviews the existing research on alibis, discusses the general social psychological and autobiographical mem~ ory context, and outlines a specific conceptual framework for the psychological study of alibis first articulated by Olson (2002).

157

158

BURKE, TURTLE, OLSON

Terminology Few can argue that an alibi defense, if it can be properly supported and proved to be true, has a strong potential for exoneration. But there are many types of alibis, and it is useful to differentiate among them. A true alibi is one in which an innocent alibi-provider gives an accurate account of his or her whereabouts at the time of a crime. An untrue alibi, on the other hand, may be one of two types: fabricated or mistaken. A fabricated alibi is deliberately false, because of the guilt of the provider or because the provider is unwilling to reveal the truth for a different reason (e.g., he was actually with his mistress). A mistaken alibi is one given initially as fact because the innocent alibi-provider believes it to be true, but which later turns out to be false. We use the term alibi provider throughout this chapter to refer to the person who is asked to give an alibi-a more inclusive, descriptive term than the commonly used defendant or suspect. We likewise use the term alibi witness to refer to those people the alibi provider offers as witnesses to back his or her alibi-sometimes referred to as a corroborator or defense witness in the eyewitness literature (e.g., Culhane & Hosch, 2004; Lindsay, Lim, Marando, & Cully, 1986).

Alibis in the Legal System The role of an alibi in a criminal investigation is not always clear, and the rules regarding how and when an alibi can and should be used vary greatly across jurisdictions. In most jurisdictions around the world, if a defendant wishes to raise an alibi as part of his or her defense, the prosecution must be notified in advance, and although the courts may consider a late notice, failure to provide notice at all may result in the alibi statement being completely excluded (e.g., P. v Brown, 2003; Craig, 1996). The general rationale of the prior notice rule is that by providing notice of an alibi ahead of time, defendants will be less likely to change their stories based on the facts of the case presented in court or to surprise the prosecution with new evidence and therefore limit their ability to conduct a thorough investigation (Epstein, 1964). Given that the alibi itself is often extraneous to the specific details of the crime, it theoretically allows the defense a great deal of latitude when coming up with a story regarding its client's whereabouts; a perceived ease of fabrication may lead to abuse of this defense. Indeed, historically alibi evidence has been referred to as the "rogue's defense"-terminology that was common in the United States in the late 1800s, and which provides some indication of how alibis were often derided, both within the court system and in legal texts at the time (Gooderson, 1977). During a trial an alibi is supposed to be considered as another piece of evidence, rather than as the piece of evidence, although it is often very difficult for all of those involved in a case to determine whether or not the alibi is credible, if those corroborating the alibi should be considered trustworthy, and the extent to which any supporting physical evidence could have been fabricated (Eldridge, 1978). The Court of Appeals in New York state (Monroe County) has stated that an alibi is "simply evidence that will require an acquittal, if, when all the evidence is considered, reasonable doubt is raised as to the defendant's guilt" (P. v Victor, 1984). Even when an alibi is considered suspect, it is not to be used as the sole basis for conviction, particularly when it has not been proved false. A

5. ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

159

false or fabricated alibi would therefore speak to the low credibility of the alibi witness or the alibi provider, but would not speak directly to his or her guilt (Connelly, 1983). Despite this supposed distinction, Gooderson (1977) notes that the reality is somewhat different; given the negative connotations associated with the term alibi itself, it tends to become something which must be proved rather than just another piece of evidence.

The Current State of Affairs Regarding Alibis in Real Cases As mentioned above, there is currently very little research, and virtually no literature, on the psychology of alibis. In contrast, the eyewitness evidence literature is rich and varied, with complex methodology, with contributions from several branches of psychology (Wells, 1993; Wells et al., 2000). Until recently, alibis were most often mentioned in the context of this eyewitness research, as a manipulated variable rather than as the focus of the study. Sanders (1984), for example, noted that alibi witness statements, no matter how powerful, rarely overcome the damaging impact of direct eyewitness evidence, even when this evidence is false. Lindsay et al. (1986) examined the impact of alibi witness testimony compared with eyewitness identification evidence on juror verdicts and found that alibi witness testimony did nothing to lessen the impact of the eyewitness testimony unless the alibi witness was unrelated to the defendant. Devlin (1976) discussed the use of alibi evidence in Britain for cases involving eyewitness testimony. Devlin's classification system states that an alibi proved to be false can and should be used as incriminating evidence that supports any eyewitness testimony. If the details of the alibi are suspicious but not clearly false, or if the alibi itself simply cannot be proved either way, then this evidence is to be considered neutral with regard to the eyewitness evidence-that is, neither helpful nor damaging. The psychology of alibis is as potentially rich, and yet as barren of data, as the eyewitness evidence area was 30 years ago. And like eyewitness evidence, alibis rely at least in part on issues involving episodic and autobiographical memory. In the case of alibis, episodic memory is at issue not only for the alibi provider, but also for alibi witnesses. Questions about the accuracy of eyewitnesses and the conditions under which they are most and least accurate can also be asked about alibi providers and alibi witnesses. A major issue in eyewitness evidence concerns the perceived credibility and persuasive impact of eyewitnesses (Wells, 1984). Similar questions can be asked about alibi providers: What are the conditions that lead people to believe alibis, and how persuasive are alibis in general? These types of questions have been explored heavily in eyewitness research, but there is only a nascent research literature on alibis. Although there are parallels between eyewitness testimony and alibis, there are also significant differences that are likely to have implications for both memory and credibility. For instance, while many eyewitnesses usually realize at the time (or soon after) that a witnessed event is important to notice and recall, an innocent alibi provider is more likely to have to recall an event that was of no particular significance at the time. Interestingly, this could render the memories of the innocent alibi provider even less accurate in some ways than the memories of crime witnesses. Another major difference concerns

160

BURKE, TURTLE, OLSON

the issue of honesty. In the case of eyewitness evidence, the presumption is that the witness is making a good-faith effort to tell the truth, and errors are presumed to be the result of faulty perceptions or memory. In the case of alibis, on the other hand, there is no presumption of honesty. Whereas an eyewitness might be believed based solely on uncorroborated recollection testimony, alibis are likely to require some form of corroboration, such as physical evidence (e.g., a timed, dated receipt) or another person. Much of the impetus behind the recent interest in alibi evidence comes from the same source as some of the interest in eyewitness evidence-the research into wrongful convictions. When Connors, Lundregan, Miller, and McEwen (1996) examined the first 28 DNA exoneration cases in the United States as part of a National Institute of Justice Project, they included an analysis of the factors that led to the erroneous convictions. Although faulty eyewitness evidence was the culprit in the vast majority of cases, 7 of these 28 cases also listed a "weak alibi" or "no alibi" as a leading factor. Although these terms were not discussed anywhere in the literature, the summaries for each of these cases suggest that a "weak alibi" is one provided by someone who might be motivated to lie, such as someone related to the suspect. Most often this motivated other appeared to be the mother or girlfriend of the suspect. In fact, it appeared that a person with such an alibi was considered in the same category as someone with no alibi at all. When these initial 28 cases were expanded to include another 12 (Wells et al., 1998), another 4 were described as having either no alibi or a weak alibi. This issue of the alibi (or lack thereof) mentioned in 11 of the 40 cases was second only to mistaken eyewitness identification in terms of the factors listed that led to wrongful convictions. At the time of this writing, the Innocence Project website (www.innocenceproject.com) lists 157 exoneration cases. Although case summaries are provided for each of these, some are quite brief, and we cannot assume that all information (such as that relating to alibis) has been included. In addition, no category for alibi evidence is listed in the specific sections dealing with why the individuals were sentenced in the first placeinstead, it appears that in most cases an alibi that was not investigated or believed most often appears under "police and prosecutorial misconduct." That said, of these 157 cases, more than 25% of them note that an alibi offered by the (actually innocent) defendant was not believed. In one particularly notable case, the defendant was in jail when the crime occurred, but his picture was accidentally placed in a lineup and was subsequently chosen by the witness. The justice system's procedures for handling eyewitness evidence are richly documented in the psychological literature, and procedural recommendations arising from this scientific literature have been worked into guidelines for collecting eyewitness evidence (e.g., Technical Working Group for Eyewitness Evidence, 1999). However, the justice system's general attitude and procedures regarding alibis are far less well known; indeed, alibi evidence is possibly one of the least considered and discussed areas of crime investigation in law enforcement. The justice system therefore has little in the way of empirical evidence by which to guide its procedures in this area. For example, a company that offers training materials to police departments relies on intuitive, commonsense theories of memory and deception when recommending how police should elicit and evaluate alibis (Reid & Associates, 2000).

5.

ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

161

To get a better understanding of the number and type of cases where an alibi is listed as a factor in the eventual trial outcome, Turtle and Burke (2003) examined a subset of cases where specific rulings were available. This sample comprised 45 Supreme Court of Canada cases, 60 Provincial Courts of Appeal cases, 20 Provincial Court cases, and 50 U.S. Court of Appeal cases. In many of these cases (44-75%, depending on the court) eyewitness evidence against the accused was also a factor. Despite the fact that, as pre~ viously noted, most jurisdictions require timely notice of an alibi defense, the results of this review suggest it is an apparently common tactic for the defense to reveal an alibi late in the investigation or trial. It was also interesting to note that the majority of these crimes occurred during the day. It is perhaps a stereotypic belief that crime is more likely to occur in the middle of the night, when one might assume it would be difficult to come up with an alibi witness. However, the majority of these crimes took place between 8 A.M. and midnight, and in most of these cases (68-86%) an alibi witness was available. In all but two of those cases, the person was a friend, family member, or co~worker, raising the issue of their potential to lie on behalf of the accused (as we know from the DNA exoneration cases, this seems to be a persistent belief). Finally, although it would be beneficial for an accused to pro~ duce physical evidence in support of his or her alibi, only 2% (Canada Supreme Court) to 14% (U.S. Court of Appeal) of the cases involved physical evidence such as hotel re~ ceipts, theater tickets, and the like.

Psychological Issues Regarding Alibis Alibi issues can occur at any point along an investigative timeline-from the time of the crime itself, through the gathering of evidence, the evaluation of this evidence, and finally to the final trial outcome. Olson (2002) has proposed a model for the study of alibis that nicely captures each of these components and provides a framework in which to view the relevant psychological research (see Fig. 5-l). The model notes two general domains of alibis, namely the generation domain and the believability domain. The generation domain involves two phases. The first is the story phase, in which the alibi provider gives a verbal account of his or her whereabouts at the time of the crime. The second phase, the validation phase, involves corroboration of the story through the generation of physical or person evidence. The believability domain involves the third and fourth phases in the process. In the third phase, the evaluation phase, the truthfulness of the story itself is evaluated by anyone who may come into contact with it. In the final phase, the ultimate evaluation phase, a definitive determination is made as to whether the accused actually committed the crime. These latter two phases are not identical. One can find that the alibi story is not true in the evaluation phase and still ultimately believe that the accused is innocent of the offense (e.g., when one assumes that the accused mis~ remembered his or her whereabouts or lied to cover for someone else or to avoid admit~ ting an impropriety such as an extramarital affair). Conversely, one cannot believe the alibi is true and yet ultimately believe that the accused is guilty of the offense; in this direction the two are mutually exclusive. We describe each of these components in more detail below.

BURKE, TURTLE, OLSON

162

Story Phase

Validation Phase

Evaluation Phase

Ultimate Evaluation Phase

Other Evidence

Generation Domain FIGURE 5-1.

Believability Domain Model of the alibi process (adapted from Olson, 2002).

Generation Domain Alibi generation includes within it issues of memory, specifically, autobiographical memory for where the alibi provider was at a certain time and place. The emphasis in the autobiographical research area to date has been on issues such as memory for the early years of childhood, changes in memory across the life span, and the process by which memories are clustered based on similar features (e.g., Conway & Pleydell-Pearce, 2000; Neisser, 1982; Rubin, 1986; Wright & Nunn, 2000), but several additional factors are especially relevant for learning about alibis. Demographic variables, for example, may play a large role in a particular person's ability to generate a believable alibi. People who are unemployed, for instance, and who live alone are more likely to have what may be perceived as a weak alibi than are employed, married people. Interacting with memory and demographics in alibi generation is time-the longer the interval between when the person is being asked to provide a statement and the time for which they are being asked to provide an alibi, the less likely it is that the person will remember or will have physical evidence. Also important is the time frame for which an alibi must be provided: It is far easier to have a solid alibi for a specific, short time interval (e.g., 15 minutes) than to account for a long time span (e.g., 6 hours). Temporal tags in memory are those details that are directly related to the time of the event's occurrence. If these tags are encoded at the time of an event they can be extremely helpful in pinpointing exactly when, and from there also where, the event took place, and more importantly, whether that is the same time that the crime allegedly took place. Research by Skowronski, Betz, Thompson, and Shannon (1991) has shown that an increased quantity of details encoded in the memory of an event increases the accuracy of remembering the date it occurred, and participants who recalled more details of a diary-recorded event were better at remembering its date. These tags can range in specificity from

5. ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

163

glancing at a clock to watching a television show that runs at a certain time. Someone's alibi might be that he or she could not have committed an offense because his or her car would not start that day. In this instance, "If one recalls that snow was on the ground when one's car broke down, the range of dates on which that event was likely to have occurred is considerably narrowed" (Skowronski et al., 1991, p. 833). Although their results indicate that many people are able to successfully recall what day of the week the event occurred, they are often off by 7-14 days ("i.e., right day wrong week") when asked to state the specific date.

Story Phase The initial phase within alibi generation is the story phase, during which the alibi provider delivers the memory statement of what he or she was doing. This is the claim or plea that he or she was somewhere other than the scene of the crime, and it relies on episodic memory. Episodic memory plays an important role in the generation of the memory statement portion of an alibi. When a person is asked to relate events that occurred in the past, with few cues or prompts, they are being asked to use their recall memory. This is in contrast to recognition memory, where an individual is given all of the relevant information and merely asked to pick out the correct choice from a set of alternatives. For example, an eyewitness asked to describe everything he or she saw that night is being asked to use recall memory, whereas an eyewitness asked to pick the per, petrator out of a lineup is using recognition memory. Unfortunately, recall memory is prone to many different types of errors. If a person did in fact commit the crime, then it is reasonable to assume that he or she will have a good memory for the day in question but will try to concoct a false alibi to thwart the investigation. On the other hand, an innocent person who did nothing out of the ordinary on the day in question is less likely to have a clear memory of that day and may have to admit having no recollection of what he or she was doing. The predicament for an innocent person can be made even worse because it is often difficult to keep memory details pure from events that occurred subsequent to the original event (e.g., Loftus, 1991), and once incorrect details have found their way into a memory the effects may be long,lasting (e.g., Blank, 1998). As noted by Conway (1983), "Remembering in the real world occurs in the context of a person's life and is intrinsically bound up with prior knowledge, future goals, and per, sonal meanings" (p. 20). There appears to be an implicit assumption within the justice system that people facing potential imprisonment-indeed even victims and complainants-should be able to recollect their actions from an earlier date. One quality that makes an alibi unique among memory statements is that an alibi has a peculiar time,space component: The alibi must speak to both the time of the crime and the space the alibi provider was occu, pying. Any evidence offered in support of the alibi must also speak to both of those di, mensions. Most episodic memory statements do not need to be so specific, and the aver, age person does not recall his or her past with the spectre of imprisonment. The climate of disbelief also makes an alibi a unique type of memory statement. Unlike the situa, tion facing the alibi provider, the everyday memory accounts that people give are (a) not

164

BURKE, TURTLE, OLSON

subject to intensive scrutiny, (b) not demanding of hard proof, and (c) not elicited by a professional interrogator. McCloskey, Wible, and Cohen (1988) provide anecdotal evidence that those people counted on to support one's alibi may not ultimately be as helpful as initially assumed. In their study of flashbulb memories for the Challenger shuttle disaster, they relate that at an informal seminar after the study several participants tried to corroborate their own memories, and the people they had remembered being with claimed they were elsewhere that day. If these kinds of inconsistencies occur naturally with supposedly powerful flashbulb memories, one may wonder how much more vulnerable to inconsistency are innocent alibi providers' memories of what for them was likely an average or typically mundane day. Research by Turtle and Burke (2001) examined the extent to which individuals can accurately recall their recent past and their ability to support their claims. First-year psychology students were asked to recall where they had been either 2 weeks or 2 days prior to the study. They were then given 1 week to validate their stated alibis. Initially, the 2-day group was more confident in their claims, but by the second week (when they had to provide evidence to support it) confidence dropped significantly. As these participants quickly discovered, supporting evidence is often difficult to come by. As students, many of them initially offered as proof of their whereabouts a class schedule, and statements that their friends could confirm their alibi. It was only later that they realized a schedule indicating where they should be was not the same as proof that they were there, and that a statement from a friend was both more difficult to collect and more suspicious to an interviewer than they had assumed. This is consistent with the findings of Kurbat, Shevell, and Rips (1998) regarding everyday memory errors and what they term a "calendar effect," described as the tendency to anchor events to seasonal cycles. Many of the student participants quite logically relied on what their class schedules suggested they should have been doing and used this as a guide to their past actions. Skowronski et al. (1991) used the diary technique to study autobiographical memory, where participants recorded one event every day for both themselves and for a significant other of their choice over the course of 10 weeks. Afterward they were asked to make two judgments, one regarding their memory of specific events and another addressing the date the event occurred. In addition to rating their memory for the recorded events, participants also rated person typicality on a 7-point scale. They found evidence for what has been dubbed the memory inconsistency effect. This refers to behaviors that are inconsistent with what is normally expected from someone being more accurately recalled than behaviors that are highly consistent. According to this idea, atypical behaviors are better remembered because they are given more attention during the encoding stage of memory as an event not fitting a regular schema or category for a person. This effect was found for memories in the diary kept for the significant other; therefore, the alibi witness may be more likely to recall the alibi provider doing something out of the ordinary at the time in question. For example, if the person providing the alibi was actually on a roller coaster at a theme park on the night of the crime, this might be easily recalled and corroborated by a friend who knows the person is severely afraid of heights.

5.

ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

165

Validation Phase Next, the alibi goes through a validation phase, during which the memory statement is corroborated with physical or person evidence (or perhaps both). Alibi validation can be undertaken by the alibi provider or by others (police investigators, for example). According to Olson (2002), this phase is what determines the true strength of the alibi, for although the memory statement is certainly tied to the types of evidence one thinks he or she can produce, it is the actual production of the corroborating evidence that gives the alibi its base. This phase can contain unique difficulties for the alibi provider, because corroborating evidence can prove to be extraordinarily elusive, as discussed above. Physical evidence, especially that which the provider has no control over (like security tapes), can be destroyed or lost. Person evidence is equally fragile; any potential corroborator is subject to the same sins of memory as the alibi provider. A Taxonomy of Alibi Strength. In order to study alibis scientifically, Olson and Wells (2004) argue that there needs to be some way to impose order on the many forms that alibis take. Some alibis, for instance, are nothing more than mere pleas that one was elsewhere at the time of the crime, with nothing but the plea itself to substantiate the claim. Other alibis might be inherently imbued with irrefutable corroborative proof that does not rely at all on the trustworthiness of the alibi provider (e.g., the alibi provider was delivering a nationwide live television broadcast at the time of the crime). It would be helpful to create some kind of organizing structure that can be brought to bear on the various forms that alibis can take. Olson and Wells (2004) begin with a distinct categorization of the two types of evidence, physical and person. In terms of person evidence, there can be a motivated familiar other, such as family members, who could conceivably have a vested interest in the outcome of the investigation. Although they would be unlikely to make a mistake, they may be more likely to lie for the alibi provider. Next, a complete stranger can be called on by someone to corroborate an alibi, and though that person would not be motivated to lie, he or she would pose a greater risk of being mistaken. It appears that the best person to corroborate an alibi is a nonmotivated familiar other, a person unlikely to lie or to be mistaken. These categories parallel several of those created by Gooderson (1977), who contrasted family members as witnesses against "stock" alibi witnesses, "disreputable witnesses," and "disinterested witnesses." There is of course a fourth category-it is possible that no person evidence exists at all-the person was home alone. In terms of physical evidence, the categories are based on a perceived ease of fabrication: easy to fabricate, difficult to fabricate, or no physical evidence available. Individuals could more easily fabricate the details of a television program they stated they were watching at the time of a crime simply by looking it up later, whereas their image on a security videotape showing them to be elsewhere at the time of the crime is not as easily manipulated. Olson and Wells designed an experiment to test this taxonomy. Undergraduate students were asked to read a description of an armed robbery, followed by three different

166

BURKE, TURTLE, OLSON

alibis from three different suspects. Each alibi was manipulated in terms of the four levels of person evidence and three types of physical evidence. The relative believability of each type of evidence was rated by the participants. As they expected, the least believable alibi was one that provided no person and no physical evidence to back it up. The most effective alibi was one that provided physical evidence that is difficult to fabricate and person evidence in the form of a nonmotivated familiar other. Generally speaking, physical evidence tended to have a stronger effect on believability than was initially expected.

Believability Domain After the first two phases in the process, the alibi now consists of both the memory statement and the evidence produced to support that statement. The process now moves into the domain of believability of alibis: an alibi's strong potential to exonerate motivates the alibi provider to produce a credible alibi and motivates alibi evaluators to determine whether an alibi is truthful. The alibi itself is a part of the puzzle of the crime, and evaluators should be motivated to make an accurate evaluation, especially when the they are expecting to make serious decisions involving the alibi provider, such as police investigators deciding whether to put their investigative resources into one of a number of possible suspects. As in the generation domain, demographic variables may also play a significant role in the believability domain. For example, evaluators may perceive employed, distinguished members of the community as inherently more believable and likable than unemployed former criminals. This could have serious implications for how credible the evaluator deems the alibi to be. In the case of an alibi situation, the evaluator (e.g., a police investigator) would likely have some initial feelings toward the alibi provider based on a first impression. For example, Burke and Turtle (2002) found that an alibi provider described as "poor" was believed less often than was a "wealthy" alibi provider, despite the fact that the details of the alibi were identical in the two cases. Other research has found that the general appearance, race, or religious beliefs of an alibi provider may influence the believability of his or her testimony in a courtroom (see Pfeifer, 1990, 1999). Going back to the DNA exoneration cases, it is evident that the race of the exonerated is overwhelmingly non-white (predominantly African American and Latino), whereas the race of the victims in the majority of cases is white. If an individual has a preconceived notion of how people in a particular category are likely to behave (i.e., stereotypical beliefs), then these expectations may have an impact on future interactions. If the expectations are negative (e.g., poor people are less credible), this may influence the processing of subsequent information regarding this individual.

Evaluation Phase During this phase, the credibility of the alibi may be evaluated by many people, including police, judges, and laypersons. Police investigators are likely the first to solicit and examine the alibi, often at a very early stage in the investigation, when it is perhaps

5.

ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

167

less likely to be developed. Judges, lawyers, and laypersons generally will come into contact with an alibi after it has gone through some retellings. Each evaluator might approach the alibi from a different perspective and therefore may view it with a different goal in mind. For purposes of determining the relative believability of one alibi compared with another, asking an evaluator to take the role of a judge or a detective should make little difference. However, future research could explore possible evaluator-role differences in greater detail. Ultimately it is in this evaluation phase that psychology may have its greatest impact. Do evaluators differentiate between mistaken and fabricated alibis? Or do they, as Arthur Will (1896) suggested, use an untrue alibi as incriminating evidence against the accused? A broad characterization of the eyewitness literature is that evaluators (e.g., jurors, mock jurors) believe autobiographical memory to be better than it actually is (Lindsay, Wells, & Rumpel, 1981). Coupled with the evaluator's assumption that an eyewitness has no reason to lie, it is no wonder that mistaken identifications are the primary cause of conviction of the innocent-the only explanation in the evaluator's mind for the identification is that it was an accurate one (Wells et al., 1998). To the extent that evaluators overestimate the memory capabilities of individuals, we might expect alibi evaluators to readily assume that an untrue alibi, especially one that has been changed or modified by the provider, is the result of deliberate lying and therefore a sure sign of guilt. The idea of an honestly mistaken alibi is not a salient alternative explanation. There are several possible explanations for this. First, retrieval cues related to temporal information may not be readily available (Burt, Kemp, & Conway, 2004). It should therefore not be surprising that initial statements regarding one's whereabouts may be modified, as perhaps more retrieval cues are provided. However, the problem is that an initial alibi statement and its adjusted version may appear to police and jurors as two inconsistent statements. They are likely to assume an inability to retrieve a memory accurately and the need to subsequently alter a statement to be indications of dishonesty. The result is that the perception of inconsistency between alibi statements may reduce the credibility of an accused person (Burke & Turtle, 2003; Culhane & Hosch, 2004; Olson & Wells, 2004). For example, Brewer, Potter, Fisher, Bond, and Luszcz (1999) have found that inconsistent eyewitness statements are more likely to be viewed as inaccurate (and, conversely, consistent statements viewed as accurate), despite the absence of any real consistency-accuracy relationship in this study. In general, then, we might expect alibi evaluators to be insensitive to the idea that there are two kinds of untrue alibis, one indicating deception and the other signifying mere misremembering. It is therefore important that those involved in the evaluation of an alibi be educated regarding these potentially critical differences. It would of course be helpful if there were specific characteristics of the alibi provider or his or her statement that could reliably predict guilt or innocence. Porter and Yuille (1996) explored the idea that there might be reliable verbal cues that the suspect or witness displays during an interview or interrogation that indicate deceit. Sixty undergraduates invited to participate in the study were asked to take either money or a folder from a room, and then provide a statement to an interrogator regarding their activities. The truthfulness of the

168

BURKE, TURTLE, OLSON

participants was varied, so that some provided a truthful alibi, whereas others provided either a false or partially false alibi. Finally, some were asked to truthfully confess to the crime. The researchers hypothesized that there would be measurable differences (using a standardized assessment technique) in the statements of those being truthful compared with those who were not. They found that untruthful alibi statements contained less detail overall and were less coherent. They also found that participants who were not completely truthful in their alibis were less likely to admit that they could not completely recall all facts associated with the event in question. Although the issue of an alibi per se was not the focus of the study, the findings regarding the believability of these statements is certainly relevant to the alibi evaluation phase. Stromwell, Granhag, and Jonsson (2003) investigated the credibility of pairs of alibi providers also in the context of deception detection. They point out that the literature in this area, as with the broader literature on the detection of deception, is somewhat bleak in that those attempting to distinguish truth from fiction often fare no better than chance. They asked 120 participants to watch videotaped statements while assuming the role of a police detective in order to determine if pairs of suspects were telling the truth or fabricating their stories (providing false alibis). They note that it is not uncommon for multiple individuals to be suspected of a crime, and to provide alibis for one another, and therefore argue that the pairs manipulation is perhaps more ecologically valid than studies involving only one suspect. Interestingly, they chose a rather mundane event as the basis for their study-participants were asked to create an alibi around a free lunch that either did or did not happen. They argue that if one were trying to come up with a believable alibi one would be wise to stick with a familiar real-life event; by changing a few details, such as time and place, one could theoretically create a more credible alibi than trying to concoct something from nothing. Their results indicated that overall, the ability to reliably and accurately detect deception was quite poor. Besides the more explicit attempts to detect deception discussed above, other factors may also play a role in the credibility associated with a particular alibi. For example, Wegner, Wenzlaff, Kerker, and Beattie (1981) offer evidence of what they call the innuendo effect. This is most often seen in evaluations of people about whom the evaluator has seen damaging personal information disguised as hints or questions in newspapers or on television. Unbeknownst to the evaluator, this damaging information continues to exert influence on the evaluator's judgments of the target. Participants in their study who saw incriminating innuendo became as negative toward the target as did those who saw directly incriminating statements. Wegner and his colleagues (1981) noted that innuendo effects would likely surface in any communication context similar to their experimental situation, one with a "highly charged atmosphere of evaluation" and "lack of prior audience knowledge of the target." (p. 831). The alibi situation is an example of one of these communication contexts; that is, the alibi is expressed in an atmosphere of evaluation, and, depending on the evaluator, the audience (especially a judge or jury) generally has little or no prior knowledge about the alibi provider. The mere fact that the alibi provider has been singled out and asked for an alibi is in itself potentially damaging information about that person. In fact, the act of providing an alibi may be a no-win situation. If someone directly requested that you give an alibi, your suspect status is at

5. ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

169

least implicit in the request-an innuendo effect. The only thing that might make you seem even more suspicious is if you provide an alibi in the absence of a request to do so (Yandell, 1979). Suspicion and innuendo may cause an evaluator to look for information that supports these preexisting beliefs to the neglect of disconfirming evidence. For example, one way that an individual may organize information is in the form of a schema or mental template (e.g., Fiske & Linville, 1980). People tend to keep related information gathered from their own experiences stored together in their memories, so that they may quickly access it when needed. Thus, a student quickly develops a schema for what he or she may expect to observe when entering a classroom-a professor at the front of the room, other students around them, a requirement to stay quiet and take notes, and so on. This information contained in the schema is reinforced every time he or she enters a classroom. Information that easily fits with a preexisting schema is accommodated into the template. However, if on a particular day a relatively minor event occurs that does not fit with the "typical" scenario (i.e., a fire alarm is pulled at the start of the class), it may be ignored by the schema altogether and quickly forgotten. Such organizational frameworks are arguably beneficial for us. However, reliance on these templates may lead us to ignore important information or to see evidence that supports these preexisting views when such evidence does not actually exist. For example, an investigator with many years' experience may come to develop a schema for what a typical guilty person looks like, such as appearing anxious, avoiding eye contact, and so on. When faced with the task of eliciting a statement from an individual, then, the investigator may rely on his or her schema and be more likely to pay attention to guilty behaviors while ignoring potentially disconfirming evidence of innocence. There is an extensive psychological literature on this potentially damaging cycle induced by preconceptions about another person-what Snyder (1984) called behavioral confirmation bias. In the alibi context, if an investigator begins with the assumption that the alibi provider is not being truthful, he or she might ask questions in ways that reinforce this belief, while possibly overlooking evidence to the contrary. Once alibi providers begin to understand that they are not being believed, they may alter their own behavior, becoming more anxious and upset. These behaviors might then serve to further reinforce the "guilty person" schema that has been activated. An interesting yet unexpected finding of the Olson and Wells (2004) taxonomy study discussed above was that even in the strongest believability condition (i.e., where an alibi was corroborated by a reliable witness and by strong physical evidence) the highest average score for those judging the credibility of the alibi was only 7.4 out of a possible 10, suggesting that it takes a great deal to overcome the initial suspicion with which an alibi is met. Research findings suggest that the negative effects of behavioral confirmation appear to occur more often in situations of conflict where we may have expectations that the person with whom we must interact will be aggressive or noncooperative (Kelley & Stahelski, 1970), or when the perceiver holds any sort of power over the person being interviewed (Copeland, 1994; Goodwin, Gubin, Fiske, & Yzerbyt, 2000), such as when a police officer interviews a suspect. In fact, Kassin, Goldstein, and Savitsky (2003) found that when individuals being interrogated by police were presumed to be guilty, behavioral

170

BURKE, TURTLE, OLSON

confirmation was more likely to occur. Indeed it has been noted that the very presumption of guilt is what fuels most interrogations. That is, only when someone is sure that a suspect is guilty is it believed that they should be interrogated (Kassin et al., 2003). This may help to explain the finding that in at least 23% of the cases reviewed by Scheck, Neufeld, and Dwyer (2000) where the suspect was actually innocent, he falsely confessed to the crime. As noted by Mortimer and Shepherd (1999), when an officer presumes guilt on the part of a suspect, "The interrogative task becomes one of confirmation not investigation" (p. 303). In the case of an alibi, it may be that a person who cannot quickly and accurately recall his or her whereabouts might become more nervous and the interviewer may then perceive this nervousness as a sign of guilt. This presumption may then lead the interviewer to focus only on the aspects of the individual's alibi that support the notion of guilt and ignore information that may suggest innocence. Given that behavioral confirmation tends to occur in a preliminary meeting-thus setting the stage for all future encounters-it is essential that those asked to gather and evaluate alibi statements be made aware of the potential role behavioral confirmation could play in their interpretation of the evidence.

Ultimate Evaluation Phase Finally, in the alibi process, if a case makes it to trial and an alibi is presented in court, it undergoes an ultimate evaluation phase, where a judge and/or jury determines the ultimate truth of the alibi. This phase is distinct from the evaluation phase. In the evaluation phase, the evaluator may not be aware of all the evidence offered in support of the alibi or of all the other facts in the case. However, the ultimate evaluation phase entails a formal presentation of the alibi in the context of the entire case, and evaluators at this phase are aware of all evidence offered in support of the alibi as well as other facts of the case. Ultimate evaluation will likely be affected by other types of evidence surrounding the case, such as other physical evidence, eyewitness accounts, or the type of crime committed. Of course, even in the ultimate evaluation phase, evaluators may be wrong in their judgment of the alibi. Culhane and Hosch (2004) studied the impact of alibis on juror decision-making processes by manipulating the confidence of the eyewitness to the crime, the valence of the alibi witnesses' testimony, and the relationship of the alibi witness to the accused. Four hundred twenty participants read a scenario describing a convenience store robbery after which they were asked to render individual verdicts. They also rated the alibi witness on various attributes, including confidence, credibility, and persuasiveness, and were asked to indicate which factors affected them most in reaching their verdicts. Not surprisingly, defendants were more likely to be found guilty when alibi witnesses were known to them than when they were unknown. In fact, having someone close to the defendant (e.g., his girlfriend) corroborate the alibi proved to be no more helpful to the defendant than when he had no alibi at all. The majority of participants rated the alibi witness as more important in their decision than either the physical evidence or the eyewitness. If they acquitted the defendant, they also rated the alibi witness higher on all measures, including the above-noted credibility, persuasiveness, and confidence. It ap-

5.

ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

171

pears then that in order to help, alibi witnesses ideally should not have any kind of personal relationship with the defendant-if they do, their testimony may end up having no impact on the trial outcome at all. This is similar to the Devlin (1976) classification discussed previously-any alibi that cannot be proved either way should be considered neutral. But the exemplars for neutral alibis are the same as for those provided by someone close to the defendant and therefore are likely the most difficult to disprove. This results in somewhat of a catch-22 because familiar others are most likely to be our only alibi witnesses (Gooderson, 1977; Turtle & Burke, 2003). CONCLUSION

There are three critical areas where psychology may already be able to assist law enforcement in the evaluation of alibi evidence. First, those involved in the assessment of an alibi need to be sensitive to the fact that a mistaken or changed alibi may not be indicative of guilt. There is an important difference between intentional deception and unintentional misremembering. With the extensive literature on the frailty of memory, even as it relates to one's own life events, inconsistent or incorrect alibis should not be overly weighted as a sign of guilt. Second, an extensive research literature tells us of the sometimes beneficial, and sometimes problematic, role of schemas in our interpretations of events. While they may provide a quick shortcut for us in daily life, when people are trying to determine the guilt or innocence of an accused, such shortcut routes are not warranted-they may cause us to ignore potentially important information or look for things that we expect to see but which may not actually exist. Finally, we need to be aware of the role that our expectations of others have on their behavior. Behavioral confirmation can occur in mundane settings, but becomes more powerful-and problematic-in situations involving pressure and power. Although it is not always easy to ignore aspects of behavior we think are useful indicators of a person's underlying psychological state (e.g., being deceptive), at times it is necessary to take a step back and ensure that our own behavior is not unduly influencing the behavior of others around us. Further research on alibis is already under way, and it is hoped that it will shed more light on their role in criminal investigations and trials. Analyses of archival data should be an especially valuable route to learning more about real alibis in real cases, although sufficient details are often not available. Surveys and focus groups with police and other investigators are likely to yield more information about their beliefs and expectations regarding alibis. And laboratory research will likely continue to identify factors that influence people's perceptions of alibis. The findings in this area are limited at present, but creating a psychology of alibis is a challenge that more researchers are likely to accept. REFERENCES Blank, H. (1998). Memory states and memory tasks: An integrative framework for eyewitness memory and suggestibility. Memory, 6, 481-529.

172

BURKE, TURTLE, OLSON

Brewer, N., Potter, R., Fisher, R., Bond, N., & Luszcz, M.A. (1999). Beliefs and data on therelationship between consistency and accuracy of eyewitness testimony. Applied Cognitive Psychology, 13,297-313. Burke, T. M., & Turtle, J. W. (2002, June). Alibi evidence: Do suspect characteristics affect the evaluation of their claims? Paper presented at the fourth biennial conference of the Society for the Psychological Study of Social Issues, Toronto, Ontario, Canada. Burke, T. M., & Turtle, J. W. (2003). Alibi evidence in criminal investigations and trials: Psychological and legal factors. Canadian Journal of Police and Security Services, 3, 286-294. Burt, C. D. B., Kemp, S., & Conway, M. (2004). Memory for true and false autobiographical event descriptions. Memory, 12, 545-552. Connelly, Peter, J. (1983). Alibi: Proof of falsehood and consciousness of guilt. Criminal Law Quarterly, 25, 165-178. Connors, E. T., Lundregan, T., Miller, N., & McEwen, T. (1996). Convicted by juries, exonerated by science: Case studies in the use of DNA evidence to establish innocence after trial. Washington, DC: U.S. Department of}ustice, National Institute of Justice, NCJ 161258. Conway, M. (1983). In defense of everyday memory. American Psychologist, 46, 19-26. Conway, M., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261-288 Copeland, J. T. (1994). Prophecies of power: Motivational implications of social power for behavioral confirmation. Journal of Personality and Social Psychology, 67, 264-2 77. Craig, J. D. R. (1996). The alibi exception to the right to silence. Criminal Law Quarterly, 39, 227-249. Culhane, S. E., & Hosch, H. M. (2004). An alibi witness's influence on juror verdicts. Journal of Applied Social Psychology, 24, 1604-1616. Devlin, Honorable Lord Patrick. (1976). Report to the Secretary of State for the Home Depart-

ment of the Departmental Committee on Evidence of Eyewitness Identification in Criminal Cases.

London: Her Majesty's Stationery Office. Eldridge, J. E. (1978). Judges' comments on the evidence: Jury instructions disparaging the alibi. Tennessee Law Review, 45, 772-782. Epstein, D. M. (1964). Advance notice of alibi. The Journal of Criminal Law, Criminology, and Police Science, 55, 29-38. Fiske, S. T., & Linville, P. W. (1980). What does the schema concept buy us? Personality and Social Psychology Bulletin, 6, 543-557. Gooderson, R.N. (1977). Alibi. London: Heinemann Educational. Goodwin, S. A., Gubin, A., Fiske, S. T., & Yzerbyt, V. Y. (2000). Power can bias impression processes: Stereotyping subordinates by default and by design. Group Processes and Intergroup Relations, 3, 227-256. Jussim, L. (1993). Accuracy in interpersonal expectations: A reflection-construction analysis of current and classic research. Journal of Personality, 61, 63 7-668. Kassin, S., Goldstein, C., & Savitsky, K. (2003). Behavioral confirmation in the interrogation room: On the dangers of presuming guilt. Law and Human Behavior, 27, 187-203. Kelley, H. H., & Stahelski, A. J. (1970). The social interaction basis of cooperators' and competitors' beliefs about others. Journal of Personality and Social Psychology, 16, 66-91. Kurbat, M. A., Shevell, S. K., & Rips, L. J. (1998). A year's memories: The calendar effect in autobiographical recall. Memory & Cognition, 26, 532-552. Lindsay, R. C. L., Lim, R., Marando, L., & Cully, D. (1986). Mock juror evaluations of eyewitnesss testimony: A test of metamemory hypothesis. Journal of Applied Social Psychology, 15, 447--459. Lindsay, R. C. L., Wells, G. L., & Rumpel, C. (1981). Can people detect eyewitness identification accuracy within and between situations? Journal of Applied Psychology, 66, 79-89. Loftus, E. F. (1991). Made in memory: Distortions in recollection after misleading information. In G. H. Bower (Ed.), The psychology of learning and motivation (pp. 187-215). New York: Academic Press.

5.

ALIBIS IN CRIMINAL INVESTIGATIONS AND TRIALS

173

McCloskey, M., Wible, C. G., & Cohen, N.J. (1988). Is there a special flashbulb-memory mechanism? Journal of Experimental Psychology: General, 117, 171-181. Mortimer, A., & Shepherd, E. (1999). Frames of mind: Schemata guiding cognition and conduct in the interviewing of suspected offenders. In A. Memon & R. Bull (Eds), Handbook of the psychology of interviewing {pp. 293-316). Chichester, England: Wiley. Neisser, U. (1982). Memory observed: Remembering in natural contexts. San Francisco: Freeman. Olson, E. A. (2002). Where were you last night? Alibi believability and corroborating evidence: A new direction in psychology and law. Unpublished master's thesis, Iowa State University. Olson, E. A., & Wells, G. L. (2004). What makes a good alibi? A proposed taxonomy. Law and Human Behavior, 28, 157-176. Oxford English dictionary. (1984). Oxford: Clarendon Press. P. v Brown, 306 AD2d 12, 761 NYS2d 630 {1st Dept. 2003). P. v Victor, 62 NY2d 374, 477 NYS2d 97 (1984). Pfeifer, J. E. (1990). Reviewing the empirical evidence of jury racism: Findings of discrimination or discriminatory findings? Nebraska Law Review, 69, 230-250. Pfeifer, J. E. (1999). Perceptual biases and mock juror decision making: Minority religions in court. Social Justice Research, 12,409-419. Porter, S., & Yuille, J. C. (1996). The language of deceit: An investigation of the verbal clues in the interrogation context. Law and Human Behavior, 20, 443-458. Reid, J. E., & Associates, Inc. (2000, April). Eliciting and evaluating an alibi. Retrieved November 13, 2001, from http://www.reid.com/lnvestigatortips.html?serial = 17 Rubin, D. (1986). Autobiographical memory. Cambridge: Cambridge University Press. Sanders, R. ( 1984). Helping the jury evaluate eyewitness testimony: The need for additional safeguards. American Journal of Criminal Law, 12, 189-220. Scheck, B., Neufeld, P., & Dwyer, J. (2000). Actual innocence: Five days to execution and other dispatches from the wrongfully convicted. New York: Doubleday. Skowronski, J. J., Betz, A. L., Thompson, C. P., & Shannon, L. (1991). Social memory in everyday life: Recall of self-events and other-events. Journal of Personality and Social Psychology, 60, 831-843. Snyder, M. (1984). When belief creates reality. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 18, pp. 248-305). New York: Academic Press. Stromwell, L.A., Granhag, P. A., &Jonsson, A.-C. (2003). Deception among pairs: "Let's say we had lunch and hope they will swallow it!" Psychology, Crime & Law, 9, 109-124. Turtle, J. W., & Burke, T. M. (2001, June). Where were you on the night of. . .? Memory and other evidence to support alibis in criminal investigations and trials. Paper presented at the biennial meeting of the Society for Applied Research in Memory and Cognition (SARMAC), Kingston, Ontario, Canada. Turtle, J. W., & Burke, T. M. (2003, July). Alibi evidence in criminal investigations and trials: An archival analysis of Canadian and U.S. cases. Paper presented at the International PsychologyLaw Conference, Edinburgh, Scotland. Technical Working Group for Eyewitness Evidence. (1999). Eyewitness evidence: A guide for law enforcement. Washington, DC: United States Department of]ustice, Office of]ustice Programs. Wegner, D. M., Wenzlaff, R., Kerker, R. M., & Beattie, A. E. (1981). Incrimination through innuendo: Can media questions become public answers? Journal of Personality and Social Psychology, 40, 822-832. Wells, G. L. (1984). How adequate is human intuition for judging eyewitness testimony? In G. L. Wells & E. F. Loftus (Eds.), Eyewitness testimony: Psychological perspectives. New York: Cambridge University Press. Wells, G. L. (1993). What do we know about eyewitness identification? American Psychologist, 48, 553-571. Wells, G. L., Malpass, R. S., Lindsay, R. C. L., Fisher, R. P., Turtle, J. W., & Fulero, S.M. (2000). From the lab to the police station: A successful application of eyewitness research. American Psychologist, 55, 581-598.

174

BURKE, TURTLE, OLSON

Wells, G. L., Small, M., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A E. (1998). Eyewitness identification procedures: Recommendations for lineups and photospreads. Law and Human Behavior, 22, 603-64 7. Will, A P. (1896). A treatise on the law of circumstantial evidence, illustrated by numerous cases. Philadelphia: T. &J. W. Johnson & Co. Wright, D. B., & Nunn, J. A (2000). Similarities within event clusters in autobiographical memory. Applied Cognitive Psychology, 14, 479-489. Yandell, B. (1979). Those who protest too much are seen as guilty. Personality and Social Psychology Bulletin, 5, 44-4 7.

6 Internalized False Confessions Saul M. Kassin Williams College

As I wrote this chapter, my attention was drawn to a courtroom in South Carolina, where 41-year-old Billy Wayne Cope was just convicted for the rape and murder of his 12-year-old daughter Amanda. Cope awoke one morning to find the oldest of his three daughters face down, cold and lifeless in her bed. It looked as if she had been strangled to death. Cope's wife worked the night shift, so she was not home. Immediately he called 911, but when the police arrived they treated him more like a suspect than a griefstricken father. Based on an erroneous first impression that there was no sign of a forced entry and belief that Cope showed "too little emotion," the police interviewed him twice, sent him to the hospital for a physical examination, and then took him to the station for questioning that would begin late at night and extend into the early morning. For more than 24 hours, Cope vehemently asserted his innocence despite persistent charges and accusations (e.g., "I swear before God, standing right here ... I did not do anything to my daughter"). During that time, he waived his rights, volunteered to be examined, and five times offered to take a polygraph test: "So you have faith in the polygraph test?" he was asked. "Yes," he replied. The next morning, after spending the night in jail, without food or drink, bewildered, still separated from family and friends, and without counsel, Cope was administered a polygraph test by a police examiner who reported to him that he failed (in fact, a leading researcher who later scored the charts indicated that Cope had actually passed). Devastated by the result, Cope wondered aloud if a person could commit such a heinous act without knowing it-an idea suggested to him the previous night by his interrogators. According to the examiner, Cope broke down and admitted that "I must have done it." He then allegedly followed this admission with a full narrative story of how he molested and strangled his daughter, cleaned up, and went back to sleep. Cope spent the next two and half days in jail, alone, still lacking contact with family, friends, or an attorney. He then handwrote a second confession in which he said that he had sexually assaulted and killed Amanda within the context of a dream. At that point, the police took him back to the house, where he reenacted on videotape-and in vivid and gruesome detail-how he had awakened in the middle of the night, molested and 175

176

KASSIN

killed Amanda while in a dissociated state, suddenly realized what he had done, went back to sleep, forgot what had occurred the next morning, then once again recalled his actions. This reenactment was followed by a fourth, even more detailed, confession typed by one of the detectives and signed by Cope. Serving as an expert witness for the defense in this case, I testified that Cope's confessions were taken under highly stressful circumstances, that police investigators used interrogation tactics that put innocent people at risk, and that Cope's statements were filled with contradictions and factual errors. None of this meant that Cope was innocent or that his confessions were false. Shortly thereafter, however, DNA tests revealed that the donor of the semen and saliva found on Amanda's dead body was not Cope but a sex offender, who was new to the neighborhood, and who had broken into other homes, raping and killing other girls in the same way. One would surmise from this DNA exoneration that Billy Wayne Cope would have been released from jail, freed, and compensated. Yet just hours after the DNA results were received, the police told Cope's wife in an egregious lie that the semen was her husband's, wired her, and sent her to jail to try to get her husband to confess again, which he did not (she died of surgery complications shortly thereafter, believing that the semen was her husband's). When the DNA was later matched to James Sanders, a serial offender, the prosecutor-armed with a police-induced confession that now did not match the facts of the crime, and lacking any evidence whatsoever of a link between the two men-charged Cope with conspiracy, arguing that he had pimped his daughter out to Sanders. The only additional evidence at trial was presented by a female friend of Cope's late wife who was corresponding with the defendant. She presented two confessional notes allegedly received shortly before trial that Cope had sent to her from jail. But Cope denied writing these notes, which were penned on paper he had no access to and in a handwriting that was likely not his own. As for the witness, she had once before been charged with forgery in another matter. In short, there was no evidence of Cope's involvement other than his original confessions. Yet after only five hours of deliberation, a South Carolina jury voted to convict him. In criminal justice, confession evidence is a prosecutor's most potent weapon-so much so, as one prominent legal scholar put it, that its introduction makes other aspects of a trial "superfluous" (McCormick, 1972, p. 316). Confessions play a vital role in law enforcement and crime control. They are also a recurring source of controversy, however, in large part because people sometimes confess to crimes they did not commit, only to be exonerated later (Drizin & Leo, 2004; Gross, Jacoby, Matheson, Montgomery, & Patel, 2004; Gudjonsson, 2003; Kassin, 1997a; Leo & Ofshe, 1998; Scheck, Neufeld, & Dwyer, 2000). Confessions are proved false when it is later discovered that no crime was committed (e.g., the presumed murder victim is found alive); when extrinsic evidence shows that the confessor could not have committed the crime (e.g., he or she was demonstrably elsewhere at the time or too young to have produced the semen found on the victim); when the real perpetrator, having no connection to the defendant, is apprehended and inculpated in the crime (e.g., by guilty knowledge, ballistics, or physical evidence); or when scientific evidence affirmatively establishes the confessor's innocence (e.g., he or she is excluded by DNA test results on semen, blood, hair, or saliva).

6.

INTERNALIZED FALSE CONFESSIONS

177

Indeed, as the post-conviction DNA exoneration numbers accumulate, research shows that 15-25% of cases had included confessions in evidence (www.innocenceproject.org/). False confessions arise in different ways and for different reasons. By reviewing the wrongful convictions that have stained the pages of American legal history, and by drawing on social-psychological theories of social influence, Kassin and Wrightsman (1985) introduced a taxonomy of false confessions that distinguished among three types: voluntary, coerced-compliant, and coerced-internalized (see Kassin, 1997a; Wrightsman & Kassin, 1993). Voluntary false confessions are self-incriminating statements that are offered to police without external pressure. Coerced-compliant false confessions are those in which a suspect confesses to police in order to escape an aversive interrogation, avoid an explicit or implied threat, or gain a promised or implied reward. This type of confession is a mere act of public compliance by a suspect who knows that he or she is innocent but is highly stressed and comes to decide that confession is more cost-beneficial than denial, at least in the short term. Finally, coerced-internalized false confessions are statements made by an innocent but vulnerable person who, as a result of exposure to highly suggestive and misleading interrogation tactics, comes to believe that he or she may have committed the crime-a belief that is sometimes supplemented by false memories. Over the years, this classification scheme has provided a useful heuristic framework for the study and analysis of false confessions and has been adopted, critiqued, and refined by others (Conte, 2000; Gudjonsson, 1992, 2003; lnbau, Reid, Buckley, & Jayne, 2001; Kassin, 1997a; Kassin & Gudjonsson, 2004; Lassiter, 2004; McCann, 1998; Ofshe & Leo, 1997). Uniquely, confessions are incriminating statements made by crime witnesses who are not bystanders or victims but alleged perpetrators. Common sense tells us that regular eyewitness can make mistakes but that innocent people do not confess to crimes they did not commit. For this reason, there is little, and I would argue insufficient, systemic concern about the reliability of the memories that these latter witnesses report. This chapter focuses on the internalized types of false confessions, those characterized by a change in the suspect's beliefs and sometimes accompanied by the formation of false memories that support those beliefs. To understand the nature of these false confessions, how they occur and why, it is important to examine some documented cases to see what they have in common, describe the methods of police interrogation that induced these confessions, review basic theories of social influence effects on cognition that are of relevance to the problem, and describe forensically specific studies of the factors that put innocent people at risk.

THE WAREHOUSE OF INTERNALIZED FALSE CONFESSIONS

In looking at cases that involve possibly internalized false confessions, it is important to realize that proof consists of some combination of a suspect's self-reports; background information about the suspect, sometimes including his or her criminal background, IQ, and personality test scores; police reports that describe what the suspect said and did

178

KASSIN

during interviews and interrogations; audiotapes and videotapes of the process, if available; and a body of extrinsic evidence indicating the confessor's guilt or innocence. When it comes to the question of internalization in these cases, the depth of the belief change may be a matter of dispute. For example, Ofshe and Leo (1997) have questioned whether an innocent confessor's acceptance of responsibility is ever fully or permanently internalized. Instead they describe the effect as temporary, unstable, and situationally adaptive and the confessor as "neither certain of his innocence nor of his culpability" (p. 209). This difference of opinion raises a question that resembles prior debates among cognitive researchers over whether misleading post-event information overwrites and alters a witness's memory for the event (e.g., Loftus, Miller, & Burns, 1978; Belli, Lindsay, Gales, & McCarthy, 1994; Weingardt, Loftus, & Lindsay, 1995) or merely coexists in storage with an intact and still retrievable memory (e.g., Dodson & Reisberg, 1991; McCloskey & Zaragoza, 1985). Although the case study approach is inherently limited, making it impossible to measure or secure behavioral proof of internalization, I believe that this criticism is misplaced (see also Kassin & Gudjonsson, 2004). Certainly in the internalized false confession cases that have been identified, the beliefs that are formed appear to be more temporary than permanent, and the cognitive product is more one of uncertainty and inference than of full-fledged acceptance and internalization. In some cases, these newly formed beliefs are buttressed by false memories of varying detail, but in other cases they are not. Nevertheless, as will become evident shortly, some degree of acceptance was evident in numerous cases-as when false confessor Paul Ingram was "brainwashed" into thinking that he had committed horrific acts of sexual violence as part of a satanic cult (see Nathan &Snedeker, 1995; Ofshe & Watters, 1994; Wright, 1994). More important, albeit on a lesser scale, this phenomenon was also observed in a controlled laboratory experiment, which is later described in greater detail (Kassin & Kiechel, 1996). This type of internalization also bears close resemblance to well-documented suggestibility effects in children (Bruck & Ceci, 1999; Ceci & Bruck, 1995), the misinformation effects produced under hypnosis (McConkey & Sheehan, 1995), the creation of false memories for words in a list (Roediger & McDermott, 1995) and autobiographical experiences that did not occur (Loftus, 1997), the "thought reform" effects of indoctrination in prisoners of war (Lifton, 1956; Schein, Schneier, & Barker, 1961), and the so-called recovery of false trauma memories in psychotherapy patients (de Rivera, 1997). In the interrogation room, the typical result is a detailed confession not only of what occurred-but when, where, how, with whom, and even why. At the time that Kassin and Wrightsman (1985) identified and defined coercedinternalized false confessions, very little systematic research had been published on misinformation effects-and there was nothing on the creation of false autobiographical memories. The science offered little guidance. In his typically precocious manner, Munsterberg (1908) long ago wrote about a Salem witch confession involving "illusions of memory" in which "a split-off second personality began to form itself with its own connected life story built up from the absurd superstitions which had been suggested to her through the hypnotising examinations" (p. 14 7). After reading through a number of more recent cases, however, we noticed that this process by which innocent people come

6.

INTERNALIZED FALSE CONFESSIONS

179

to accept blame for crimes they did not commit followed a predictable course, as if it was scripted. Indeed, Gudjonsson and MacKeith (1982) suggested that this type of confes, sion was the product of "memory distrust syndrome," a form of source amnesia in which people develop a profound distrust of their memory, rendering them vulnerable to influence from external cues and suggestions. The South Carolina trial of Billy Wayne Cope clearly depicted the process. Even if one were to argue in the absence of supportive evidence that Cope had conspired to invite his daughter's rape and murder, his confessions-which tell stories about his own actions, which did not occur, while omitting all mention of an intruder-were factually incorrect. Yet after suffering through a highly aggressive interrogation and a night alone in jail, and following false feedback about a failed polygraph, Cope became confused, lost confidence in his own memory, wondered about a possible blackout, and concluded that "I must have done it." From there, he constructed a vividly detailed confession that fit the facts of the crime scene as the police knew them to be at the time (e.g., "Amanda was asleep on her stomach ... I started strangling her with my hands ... Then I fixed the doors in Amanda's bedroom so that they would lock."). Numerous other stories illustrate this phenomenon. In a classic case, 18-year-old Peter Reilly returned to his Connecticut home one night to find that his mother had been murdered. Reilly immediately called the police but was suspected of matricide. After gaining his trust, one detective told Reilly that he failed a lie-detector test, which was not true, and which indicated that he was guilty despite his lack of a conscious recollection. After hours of relentless interrogation, Reilly underwent a chilling transformation-from denial through confusion, self-doubt, and conversion ("Well, it really looks like I did it"), followed by the utterance of a full written confession ("I remember slashing once at my mother's throat with a straight razor I used for model airplanes ... I also remember jumping on my mother's legs."). Two years later, independent evidence regarding the timeline revealed that Reilly could not have committed the murder, and that the confession even he came to believe was false. Reilly was released from prison and was never retried (Barthel, 1976; Connery, 1977). In a third case, Paul Ingram-a religious Christian, and a deputy sheriff in Olympia, Washington-was accused of raping his daughter, sex abuse, and satanic cult crimes that included the slaughter of newborn babies. After two dozen interrogations, which extended for 5 months, Ingram was detained, hypnotized, provided graphic crime details, and told by a police psychologist that sex offenders typically repress their offenses. At one point, he was urged by his pastor to confess. Ingram eventually "recalled" his crimes, pled guilty, and served his full sentence of 20 years in prison until he was released in 2003. Yet there was no physical evidence to prove that some of the crimes to which he confessed had even occurred. Serving as a consultant for the state in this case, Ofshe (1992) concluded that Ingram was "brainwashed" into thinking he was part of a satanic cult. To demonstrate Ingram's potential for suggestibility, Ofshe manufactured a phony crime. Ingram denied the new charge at first, but after 24 hours he submitted a full confession-and embellished the story. This case is fully described in three books: Remembering Satan (Wright, 1994), Making Monsters (Ofshe & Watters, 1994), and Satan's Silence (Nathan & Snedeker, 1995).

180

KASSIN

In a fourth case, in California, 14-year-old Michael Crowe-and then his friend, Joshua Treadway-confessed to the stabbing death of Michael's younger sister Stephanie. At first, Michael vehemently denied the charge. Eventually, however, Michael conceded that he was involved: "I'm not sure how I did it. All I know is I did it." This change in belief followed from three separate, highly charged interrogation sessions during which Michael was told that his hair was found in Stephanie's grasp, that her blood was in his bedroom, that all means of entry to the house were locked, and that he had failed a voice stress lie test-all claims that were false. Failing to recall the crime, Michael was persuaded that he had a split personality, that "good Michael" blocked out the incident, and that he could imagine how "bad Michael" had killed her. The charges against the boys were later dropped when Richard Tuite, a drifter who had a history of violence and who was prowling the area that night, was found with Stephanie's blood on his clothing. Tuite was eventually prosecuted and convicted (Drizin & Colgan, 2004).

SOCIAL INFLUENCES OF INTERROGATION

To understand the inducement to confess, it is necessary to know what methods of social influence are used in the interrogation room. Techniques vary and are described in the manuals that are available to train law enforcement professionals. The most popular of these manuals is Inbau, Reid, Buckley, and Jayne's (2001) Criminal Interrogation and Confessions, which was first published in 1962, is now in its fourth edition, and forms the basis for the popular Reid technique. According to Inbau et al. (2001), police begin a two-staged process with an open, nonconfrontational interview designed to determine whether the suspect is telling the truth or lying and, hence, whether or not to proceed to interrogation. Despite substantial evidence to the contrary (see Meissner & Kassin, 2002; Vrij, 2000), investigators are trained in this method to believe that they can make judgments of truth and deception at high levels of accuracy by analyzing the suspect's verbal and nonverbal behavior. Thus, it is clear that interrogation is by definition a guilt-presumptive process, a theory-driven social interaction led by an authority figure that has already formed a strong a priori belief, confidently held but often erroneous, that the suspect is guilty. As in other domains of social interaction, this presumption of guilt paves the way for a range of cognitive and behavioral confirmation effects (Kassin, Goldstein, & Savitsky, 2003). As for the interrogation itself, lnbau et al. advise police to conduct the questioning in a special room at the station that is small, barely furnished, and soundproof. The goal is to isolate the suspect, denying access to familiar people and places, in order to increase the incentive to escape and to insulate the suspect from outside sources of information and support. Against this physical backdrop, the Reid approach to interrogation is a multistep procedure that begins when a detective confronts the suspect with a strong and unwavering assertion of guilt. This confrontational phase may last from minutes to several hours until the suspect falls into a state of despair. As part of this process, investigators are trained to interrupt all efforts at denial, label the suspect a liar, overcome all objections, and refuse to allow the suspect to mount a defense. This confrontational

6.

INTERNALIZED FALSE CONFESSIONS

181

phase may even be bolstered by the insinuation or outright presentation of incontro, vertible evidence, which may or may not be true-a tactic that significantly increases compliance (e.g., Kassin & Kiechel, 1996). As the stress of interrogation intensifies, and as the beleaguered suspect comes to realize that denial does not provide a means of escape, detectives begin to develop scripted themes designed to help psychologically jus, tify, minimize, or excuse the crime charged. Showing sympathy and understanding, and urging the suspect to cooperate, detectives offer moral justification and a face,saving alternative construal of the alleged guilty act (e.g., suggesting to the suspect that he or she was intoxicated, peer pressured, provoked, or acting in self-defense, or that his or her actions were accidental). Using these minimization techniques, detectives imply that the suspect's alleged actions were morally defensible, which encourages confession by the implication that leniency may be forthcoming (Kassin & McNall, 1991; Russano, Meissner, Narchet, & Kassin, 2005). Conceptually, this multistep procedure is designed to get suspects to incriminate themselves by increasing the anxiety associated with denial, breaking them down into a state of despair, and minimizing the perceived consequences of confession. In this way, confession appears as a rational, cost-effective means of escape. The detective thus gets the suspect to make a simple admission, then to recount the details of the crime, ulti, mately converting that statement into a full written confession. It is clear that these methods are used with some frequency. John E. Reid and Associates report that they have trained more than 300,000 law enforcement professionals in seminars on interviewing and interrogation (www.reid.com/index.html). Leo (1996) observed 182 videotaped and live interrogations at three police departments in California, in which 64% of sus, pects made self-incriminating statements. He found that the detectives used a mean of 5.62 different techniques per interrogation and that those described above were particularly common.

THE PROCESS OF INTERNALIZATION

The cases described earlier and others illustrate that there is a predictable, if not scripted, process that gives rise to internalized false confessions. In one form or another, the process contains five components: (1) There is a suspect who is rendered highly vulner, able to manipulation as a function of dispositional characteristics (e.g., young, nai:Ve, mentally retarded, suggestible, or otherwise impaired) and transient factors associated with the crime, custody, and interrogation (e.g., extreme stress, feelings of isolation, sleep deprivation, the influence of drugs). (2) Knowingly or unknowingly, the police confront the suspect with false but allegedly objective and incontrovertible evidence of his or her involvement-evidence in the form of a failed polygraph, an eyewitness, a fin, gerprint, a shoeprint, or a DNA sample. (3) Often with guidance from police, the sus, pect reconciles his or her lack of memory with the alleged evidence by presuming that he or she had blacked out, dissociated, repressed, or otherwise failed to recollect the event. (4) The suspect makes a tentative admission of guilt, typically using a language of inference rather than of direct experience (e.g., "I guess I did it," "I may have done it," or

182

KASSIN

"I must have done it" rather than "I did it"). (5) The suspect may convert the simple admission into a fully detailed confession in which confabulations of memory originate from his or her exposure to secondhand sources of information (e.g., leading questions, overheard conversations, crime scene photos, and visits to the crime scene), often facilitated by various imaginational exercises (e.g., "Think hard about how you would have . ."). done 1t Focusing on how police have persuaded innocent suspects to accept responsibility for a crime they did not commit and cannot recall, Ofshe (1989) identified a number of common interrogation tactics, such as exhibiting strong and unwavering certainty about suspect's guilt, isolating the suspect from all familiar social contacts and outside sources of information, conducting sessions that are lengthy and emotionally intense, presenting false but allegedly incontrovertible proof of the suspect's guilt, offering the suspect a ready physical or psychological explanation for why he or she does not remember the crime, and applying implicit and explicit pressure on the suspect, in the form of promises and threats, to comply with the demand for a confession. As profound a form of influence as this seems, the construction of an internalized false confession may not be unique. Reviewing de Rivera's (1997) analysis of people who recover false memories from childhood only later to retract these reports, Kassin (1997b) likened this process of police interrogation to that of the recovery of false memories of childhood abuse in psychotherapy patients. In both situations, an authority figure claims, often with certainty, to have privileged insight into the individual's past; the individual is in a heightened state of weakness and malleability; all interactions between the expert and individual occur within a private, highly insulated setting devoid of external social or reality cues; and the expert ultimately convinces the person to accept a negative and painful self-insight by citing objective symptoms of this truth and invoking such concepts as dissociation, repression, alcoholic blackout, or multiple personality disorder (for related analyses, see Kopelman, 1999; Ost, Costall, & Bull, 2001). Over the years, two conceptually distinct models-one focusing on self-perception and the other on the misattributions that result from faulty source monitoring-have been proposed to explain how people might come to believe that they were involved in a crime or some other act they did not commit. In the same year that the United States Supreme Court referred to modern police interrogations as "inherently coercive," Bern (1966) theorized that false confessions may result from a process of self-perception. Bern's self-perception theory states that to the extent that internal states are weak or difficult to interpret, people infer what they think or how they feel by observing their own behavior and the situation in which that behavior took place. Interested in the criminal-forensic implications of"when saying is believing," Bern suggested that making a false confession could distort a person's recall of his or her own behavior if that confession is emitted in the presence of cues that are normally associated with telling the truth. To demonstrate, Bern had subjects in a laboratory experiment perform a task that required them to cross out some words but not others from a master list. To establish two colored lights as discriminative stimuli, one for truths and the other for lies, he then asked subjects general questions about themselves and instructed them to answer truthfully when the room was illuminated by a green light and to lie in the presence of an

6.

INTERNALIZED FALSE CONFESSIONS

183

amber-colored light. Next, the experimenter announced several words taken from the initial task. After some words, he instructed subjects to lie and after others to tell the truth about whether they had previously crossed the word out-again, while in the presence of a green or amber light. In a third and final phase, subjects were asked for each word to recall whether they actually had or had not crossed it out. The result was that false statements made in the presence of a truth light produced more errors in the recall of actual performance than did false statements made in the presence of the lie light or none at all. Under conditions normally associated with truth telling, subjects came to believe the lies they had been induced to tell. Pondering the implications for criminal justice, Bern (1967) noted that "a physical or emotional rubber hose never convinced anyone of anything" and that "saying becomes believing only when we feel the presence of truth, and certainly only when a minimum of inducement and the mildest and most subtle forms of coercion are used" (pp. 23-24). It is important not to generalize without disclaimer from Bern's laboratory experiment to real-life police interrogations. Still, case studies and anecdotal reports indicate the existence of internalized confessions, and self-perception theory provides one possible explanation for the first phase of this phenomenon, the formation of a false belief. For example, Driver (1968) described a common tactic in which police ask the suspect to repeat his or her story over and over again and suggested that "If duped into playing the part of the criminal in an imaginary sociodrama, the suspect may come to believe that he was the central actor in the crime" (p. 53). Such a transformation in self-perception appears to have afflicted 13-year-old Jerry Pacek, who in 1958 confessed to and reenacted for police a woman's murder that he did not commit. When asked to recount the experience more than 30 years later, "Jerry said he confessed so many times, to so many people, his memory of what happened that week is just a blur" (Fisher, 1996, p. 187). Focusing more on consciousness and memory, Foster (1969) likened the process of interrogation to hypnosis, suggesting that it can produce a "trance-like state of heightened suggestibility" so that "truth and falsehood become hopelessly confused in the suspect's mind" (pp. 690-691). Consistent with this notion, Weinstein, Abrams, and Gibbons (1970) found that hypnotized subjects in whom a false sense of guilt was induced were less able than others to pass a polygraph-based lie detector test. Within a contemporary framework that takes into account the confusion that plagues internalized false confessors, Henkel and Coffman (2004) recently argued that the reality-distorting processes of police interrogation provide fertile ground for source-monitoring confusion and the formation of internalized false confessions. According to this account, a suspect who cannot recall the details of having committed a crime but has access to information about it faces a cognitive source-monitoring dilemma: to differentiate between memories that arise from direct personal experience and those that emanate from his or her own thoughts, dreams, and imagination exercises, or from external but secondhand sources of information (e.g., leading questions, overheard conversations, photographs of the victim, or live visits to the crime scene). A source-monitoring framework focuses on how people make attributions for the sources, contexts, or origins of their own memories (see Johnson, Hashtroudi, & Lindsay, 1993). As reviewed by Henkel and Coffman (2004), research has shown that real or

184

KASSIN

imagined objects, actions, or events are sometimes misattributed in context to direct perception or experience-and that this source confusion is most likely to occur when the imagined material is plausible, vivid, easy to imagine, the subject of repetition, and similar to objects, actions, or events previously experienced. This problem is evident in eyewitness situations in which an innocent person, familiar looking because he or she was seen in one situation (e.g., in prior mug shots; present as a bystander), is later confused in a witness's memory and "transferred" to another situation, say the crime scene, only to be mistakenly identified as the criminal in a lineup (Ross, Ceci, Dunning, & Toglia, 1994). Of relevance to the internalized false confessions that sometimes emerge during interrogation, research alluded to earlier in this section indicates the profound biasing effects on autobiographical memory of exposure to photographs of nonwitnessed events (Koutstaal, Schacter, Johnson, & Galluccio, 1999; Lindsay, Hagen, Read, Wade, & Garry, 2004), verbal misinformation (Loftus & Hoffman, 1989), reports of co-witnesses (Gabbert, Memon, & Allan, 2003), imaginational exercises (Mazzoni & Memon, 2003; Thomas & Loftus, 2002), dream interpretation (Mazzoni, Loftus, & Seitz, 1999), and sheer repetition (Begg, Anis, & Farinacci, 1992). All of these biasing techniques inflate the likelihood of illusory recollections compared with that found in appropriate control conditions. Indeed, imagination inflation is the descriptive term that has been coined to refer to increased levels of false memories following the use of imaginational exercises (Garry, Manning, Loftus, & Sherman, 1996). Interestingly, too, research shows that people are particularly susceptible to misinformation effects when the scenes they are trying to recall are negative and highly emotional (Porter, Spencer, & Birt, 2003). In short, there is reason to believe that innocent people under the influence of police interrogation may be at risk for source confusion and the formation of false memories.

INTERNALIZED FALSE CONFESSIONS IN THE LABORATORY

Until recently, there was no empirical evidence for this phenomenon. To be sure, eyewitness researchers had found that misleading post-event information can alter actual or reported memories of observed events (e.g., Loftus et al., 1978; McCloskey & Zaragoza, 1985)-an effect that is particularly potent in preschool children (Ceci, Ross, & Toglia, 1987; Ceci & Bruck, 1995) and adults under hypnosis (e.g., Dinges et al., 1992; Sheehan, Statham, &Jamieson, 1991). Other studies indicate that it is possible to spontaneously produce false "recollections" of words in a list (e.g., Roediger & McDermott, 1995) and implant isolated childhood experiences that were supposedly forgotten or repressed (Hyman, Husband, & Billings, 1995)-like being lost in a shopping mall (Loftus, 1993). But can people's memory for their own actions similarly be altered? Can people be induced to accept guilt for outcomes they did not produce? As noted earlier, the cases involving internalized false confessions appear to have had two factors in common: an innocent person whose memory is rendered vulnerable to manipulation and the presentation of false evidence. Kassin and Kiechel (1996) thus developed a laboratory paradigm designed to test the hypothesis that the presentation of

6.

INTERNALIZED FALSE CONFESSIONS

185

false evidence can lead individuals who are vulnerable to confess to a prohibited act they did not commit, to internalize responsibility for that act, and to confabulate details con~ sistent with that belief. Two subjects per session participated in this experiment (actu~ ally there was only one subject and a confederate). The confederate was to read a list of letters and the subject was to type these letters as quickly as possible on the keyboard of a personal computer. Before the session began, subjects were warned not to press the ALT key positioned near the space bar or else the computer would malfunction and data would be lost. After 60 seconds, the computer supposedly crashed, at which point a dis~ traught experimenter accused the subject of hitting the forbidden key. All subjects were innocent and all initially denied the charge. In each session, the subject's vulnerability was manipulated by varying the pace of the task, fast or slow. The second factor was the presentation of false evidence in the form of a confederate who told the experimenter that she did or did not witness the sub~ ject hitting the forbidden key. Three levels of influence were then assessed. To elicit compliance, the experimenter quickly handwrote a confession and prodded subjects to sign it. To measure internalization, he recorded the way subjects privately described the experience when away from the experimenter. As subjects left the lab, they met a wait~ ing subject, actually a second confederate, who presumably overheard the commotion. This confederate asked what happened. The subject's reply was then coded for whether he or she accepted the blame for what happened (e.g., "I hit a key I wasn't supposed to and broke the computer."). Finally, although the session appeared to be over, the exper~ imenter brought subjects back into the lab, re~read the letters they typed, and asked if they could reconstruct how and when they hit the ALT key. This was used to probe for evidence of confabulation, to see if subjects would concoct details to fit their newly formed belief (e.g., "Yes, here, I hit it with the side of my hand right after you called out the 'A'."). Afterward, subjects were carefully debriefed about the study-its purpose, the hypothesis, and the need for the use of deception. Overall, 69% of all subjects signed the confession, 28% internalized guilt, and 9% manufactured details to support their newly created false beliefs. More important were the effects of the independent variables. In the baseline slow pace/no witness group, 35% of subjects signed the confession but none exhibited internalization or confabula~ tion. Yet in the fast pace/witness group all subjects signed the confession, 65% internal~ ized guilt, and 35% concocted supportive details. In short, people were induced to con~ fess and internalize guilt for an outcome they did not produce. In some cases they even went on to support that newly created belief of what they did with a false memory of how they did it. As predicted, the risk is increased both by personal vulnerability and the presentation of false evidence, a trick often used by police and sanctioned by the courts. Indeed, in Frazier v. Cupp (1969), the U.S. Supreme Court considered a case in which police told the defendant that the person who provided his alibi had confessed, which was false, and it refused to exclude the resulting confession. Since that time, the Court has repeatedly declined to reconsider the issue (Magid, 2001). Follow~up studies using variants of this computer crash paradigm have replicated and extended this effect. In an experiment conducted in the Netherlands, Horselenberg, Merckelbach, and Josephs (2003) accused college students of crashing a computer by

186

KASSIN

hitting a prohibited key and obtained even higher rates of compliant false confessions, internalization, and confabulation-even when the subjects were led to believe that confession would bear a financial consequence. Redlich and Goodman (2003) also obtained high rates of compliance in this paradigm despite leading subjects to believe that they would have to return for 10 hours without compensation to reenter the lost data. Demonstrating an important limitation of this effect, Klaver, Gordon, and Lee (2003) found that the false confession rate declined from 59% when subjects were accused of hitting the ALT key, as in the original study, to 13% when they were accused of hitting the less plausible Esc key. Focusing on individual differences in vulnerability, other researchers observed particularly high false confession rates in response to false evidence among stressinduced males (Forrest, Wadkins, & Miller, 2002) and among juveniles, 12 to 16 years old, who are more vulnerable to the effect than adults (Redlich & Goodman, 2003). It is clear that some people are dispositionally more vulnerable than others to make and internalize false confessions under interrogative pressure. To assess individual differences in this type of vulnerability to interrogation, Gudjonsson (1984) devised a memoryrelated instrument. Known as the Gudjonsson Suggestibility Scale (two parallel forms were created, GSS 1 and GSS 2), the test involves reading a narrative paragraph to a subject, who then free-recalls the story, immediately and after a brief delay, and then answers 20 memory questions-including 15 that are subtly misleading. After receiving feedback indicating that he or she had made several errors, the subject is then retested, presumably for the purpose of obtaining a higher level of accuracy. Through this testretest paradigm, researchers can measure the extent to which subjects exhibit a general shift in memory as well as a tendency to yield to misleading questions in the first and second tests. Added together, these two scores are used to determine a subject's Total Suggestibility (see Gudjonsson, 1997). Indeed, Scullin and Ceci (2001) created a similar video-based test to measure individual differences in suggestibility among preschool children. As a general rule, individuals who score high on interrogative suggestibility also tend to exhibit poor memories, high levels of general anxiety, low self-esteem, and a lack of assertiveness. Among crime suspects, "alleged false confessors" (those who confessed to police but later retracted the statements) obtained higher scores, and "resistors" (those who maintained their innocence throughout interrogation) obtained low scores relative to the general population (Gudjonsson, 1991). Research also shows that suggestibility scores on the GSS increase as a function of prolonged sleep deprivation, a state that often plagues suspects who are detained and questioned late at night (Blagrove, 1996), and as a function of alcohol withdrawal, also a common problem in criminal justice (Gudjonsson, Hannesdottir, Petursson, & Bjornsson, 2002). As for a link between suggestibility and internalization, Sigurdsson and Gudjonsson (2001) compared personality test scores of prison inmates who self-reported that they had confessed falsely to police with those of other prison inmates. They found that those who seemed to have internalized guilt, at least in part, were significantly more suggestible, as measured by the GSS. Finally, it is important to distinguish among the possible cognitive outcomes of this social influence-based process of internalization. Right from the start, Kassin and

6.

INTERNALIZED FALSE CONFESSIONS

187

Wrightsman (1985) had defined internalized false confessions as those in which the sus, pect comes to believe at some varying level of certainty and for some varying period of time that he or she is guilty, and that this false belief may or may not be accompanied by an alteration in memory. A perusal of internalized false confession cases supports the conclusion that police,induced changes in a suspect's beliefs are more common and not necessarily followed by the manufacture of new "memories." This is an important distinction. Noting that people hold many autobiographical beliefs for events that they cannot recall, Scoboria, Mazzoni, Kirsch, and Relyea (2004) proposed that plausibility, belief, and memory represented a series of nested effects re, lated to autobiographical accounts-that an event must be seen as plausible before it is believed and that it must be believed before it can generate a memory. To test this hy, pothesis, these investigators compiled a list of 10 childhood events that varied in their plausibility (e.g., getting lost in a mall, losing a toy, having a tooth extracted, getting ab, ducted by a UFO). For each question, subjects rated how plausible it is that this event occurred to them before the age of six, the strength of their belief that it occurred, and their memory of that occurrence. On average, plausibility ratings were higher than belief ratings, which, in turn, were higher than memory ratings. More to the point, memories were nested within beliefs, and beliefs were nested within perceptions of plausibility. The relevance of this nesting hypothesis to the manifestations of internalized false confes, sions is clear. In the original computer crash study described earlier, Kassin and Kiechel (1996) found that more subjects signed a false confession (compliance) than believed they were guilty of hitting the prohibited ALT key (internalization)-and only a subset of those who believed they were guilty also generated false memories of how they did it (confabulation).

PROSPECTS AND PROPOSED SOLUTIONS

As DNA exonerations accumulate, raising serious and fundamental concerns about the reliability of police,induced confessions, it is necessary that police, prosecutors, defense lawyers, judges, and juries learn how to better assess this evidence. Kassin, Meissner, and Norwick (2005) videotaped male prison inmates confessing to the crimes for which they were incarcerated and concocting false confessions to crimes they did not commit. Neither college students nor police investigators were able to distinguish significantly between the true and false confessions. Archival analyses show that confessions tend to overwhelm alibis and other evidence of innocence, resulting in a chain of adverse legal consequences (Leo & Ofshe, 1998; Drizin & Leo, 2004). Indeed, some prosecutors will refuse to admit innocence in the presence of a confession even after DNA tests appear to exonerate the confessor. This is what happened in the tragic conviction of Billy Wayne Cope described earlier. Grief stricken and subjected to relentless interrogation, Cope confessed to the rape and murder of his daughter after he was told that he failed a polygraph he trusted. Afterward, DNA tests showed that the semen taken from the victim belonged to a serial sex offender, not to Cope. Rather than drop the charges,

188

KASSIN

however, the prosecutor took Cope to trial, persuaded a jury to convict him of conspiracy, and stated to the press afterward that the verdict vindicated police (Dys & Pettibon, 2004). The problem in judging police-induced confessions, including those that ultimately prove to be false, is that the statements typically contain vivid and accurate details about the crime. To a naive observer, false confessions appear voluntary, accurate, and the product of personal experience. As a matter of speculation, one might expect that judges and juries would be fooled more by internalized false confessions than by compliant false confessions because they are not retracted quickly or with confidence and because they result from a deeper, more profound, less intuitive form of social influence. There are two policy implications that follow from the problems that arise from internalized false confessions. The first concerns the interrogation practice of lying to suspects about the evidence, a form of trickery that is permissible (Frazier v. Cupp, 1969) and is frequently used (Leo, 1996). Laboratory experiments have shown that the presentation of false evidence increases the risk that innocent people, particularly those vulnerable to manipulation, will confess to acts they did not commit and even at times internalize blame for outcomes they did not produce (Kassin & Kiechel, 1996; Horselenberg et al., 2003; Redlich & Goodman, 2003). In light of this research as well as numerous false confession cases in which the presentation of false evidence was implicated (as when Billy Wayne Cope was told that he failed the polygraph, feedback that led him to question his own memory), the courts should revisit their approval of this interrogation practice, realizing the ways in which deception increases the risk of false confessions. A second implication concerns the full videotaping of interrogations. For judges, juries, and other decision makers, evaluating a confession should involve a three-pronged analysis. The first prong is to consider the conditions under which the suspect confessed and the extent to which coercive social influence techniques were used. The second is to consider whether the confession contains details that are accurate in relation to verifiable facts of the crime. An overlooked but necessary third prong concerns a requirement of attribution for the source of those details. A confession can prove guilt if and only if it contains information knowable only to a perpetrator that was not derivable from pictures, leading questions, and other secondhand sources. To accurately judge the probative value of confessions, then, fact finders must have access to a videotape recording of the entire interview and interrogation in order to assess voluntariness and trace the origin or source of accurate details. Currently, only five states (Minnesota, Alaska, Illinois, Maine, Wisconsin) and the District of Columbia have videotaping requirements. In many other police departments, however, the practice is conducted on a voluntary basis. Several years ago, a National Institute of Justice study revealed that many police and sheriff's departments conducted videotaped interrogations on their own and that the vast majority found the practice to be useful (Geller, 1993). More recently, Sullivan (2004) interviewed officials from 238 police and sheriff's departments in 38 states who voluntarily recorded custodial interrogations and found that they enthusiastically favored the practice. Among the self-reported benefits cited are that recording permits detectives to focus on the suspect rather

6.

INTERNALIZED FALSE CONFESSIONS

189

than taking copious notes; that it increases accountability; that it provides an instant replay of the suspect's statement, often revealing information that was initially overlooked; that it enables a more objective and accurate record than does a reliance on memory; and that it reduces the amount of time detectives spend in court defending their interrogation conduct. For these reasons, a mandatory videotaping requirement has many advocates among legal scholars, social scientists, and law enforcement professionals (Drizin & Colgan, 2001; Drizin & Leo, 2004; Gudjonsson, 2003; Kassin, 2004; Slobogin, 2003).

REFERENCES Barthel, J. (1976). A death in Canaan. New York: Dutton. Begg, I. M., Anis, A., & Farinacci, S. (1992). Dissociation of processes of belief: Source recollection, statement familiarity, and the illusion of truth. Journal of Experimental Psychology: General,

121,446-458.

Belli, R. F., Lindsay, D. S., Gales, M. S., & McCarthy, T. T. (1994). Memory impairment and source misattribution in postevent information experiments with short retention intervals. Memory and Cognition, 22, 40-54. Bern, D. J. (1966). Inducing belief in false confessions. Journal of Personality and Social Psychology, 3, 707-710. Bern, D. J. (1967, June). When saying is believing. Psychology Today, 1, 21-25. Blagrove, M. (1996). Effects oflength of sleep deprivation on interrogative suggestibility. Journal of Experimental Psychology: Applied, 2, 48-59. Bruck, M., & Ceci, S. J. (1999). The suggestibility of children's memory. Annual Review of Psychology, 50, 419-439. Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific analysis of children's testimony. Washington, DC: American Psychological Association. Ceci, S. J., Ross, D. F., & Toglia, M.P. (1987). Suggestibility of children's memory: Psycholegal implications. Journal of Experimental Psychology: General, 116, 38-49. Connery, D. S. (1977). Guilty until proven innocent. New York: G.P. Putnam's. Conte, R. (2000). The psychology of false confessions. Journal of Credibility Assessment and Witness Psychology, 14-36. de Rivera, J. (1997). The construction of false memory syndrome: The experience of retractors. Psychological Inquiry, 8, 271-292. Dinges, D. F., Whitehouse, W. G., Orne, E. C., Powell, J. W., Orne, M. T., & Erdelyi, M. H. (1992). Evaluating hypnotic memory enhancement using multitrial forced recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 1139-1147. Dodson, C., & Reisberg, D. (1991). Indirect testing of eyewitness memory: The (non)effect of misinformation. Bulletin of the Psychonomic Society, 29, 333-336. Driver, E. D. (1968). Confessions and the social psychology of coercion. Harvard Law Review, 82, 42-61. Drizin, S. A., & Colgan, B. A. (2001). Let the cameras roll: Mandatory videotaping of interrogations is the solution to Illinois' problem of false confessions. Loyola University Chicago Law Journal, 32, 337-424. Drizin, S. A., & Colgan, B. A. (2004). Tales from the juvenile confessions front. In G. D. Lassiter (Ed.), Interrogations, confessions, and entrapment (pp. 127-162). New York: Kluwer Academic. Drizin, S. A., & Leo, R. A. (2004). The problem of false confessions in the post-DNA world. North Carolina Law Review, 82, 891-1007. Dys, A., & Pettibon, S. (2004). Cope, Sanders found guilty. Rock Hill Herald, September 23, 2004. Fisher, J. (1996). Fall guys: False confessions and the politics of murder. Carbondale, IL: Southern Illinois University Press.

KASSIN

190

Forrest, K. D., Wadkins, T. A., & Miller, R. L. (2002). The role of preexisting stress on false confessions: An empirical study. Journal of Credibility Assessment and Witness Psychology, 3, 23-45. Foster, H. H. (1969). Confessions and the station house syndrome. DePaul Law Review, 18, 683-701. Frazier v. Cupp, 394 U.S. 731 (1969). Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can eyewitnesses influence each other's memories for an event? Applied Cognitive Psychology, 17,533-544. Garry, M., Manning, C., Loftus, E. F., & Sherman, S. J. (1996). Imagination inflation: Imagining a childhood event inflates confidence that it occurred. Psychonomic Bulletin and Review, 3, 208-214. Geller, W. A. (1993). Videotaping interrogations and confessions. National Institute of Justice: Research in Brief. Washington, DC: U.S. Department ofJustice. Gross, S. R., Jacoby, K., Matheson, D. J., Montgomery, N., & Patel, S. (2004). Exonerations in the United States 1989 through 2003. Unpublished report, April19, 2004. Gudjonsson, G. H. (1984). A new scale of interrogative suggestibility. Personality and Individual Differences, 5, 303-314. Gudjonsson, G. H. (1991). Suggestibility and compliance among alleged false confessors and resisters in criminal trials. Medicine, Science, and the Law, 31, 147-151. Gudjonsson, G. H. (1997). The Gudjonsson Suggestibility Scales manual. Hove: Psychology. Gudjonsson, G. H. (2003). The psychology of interrogations and confessions. A handbook. Chichester: John Wiley & Sons. Gudjonsson, G. H., Hannesdottir, K., Petursson, H., & Bjornsson, G. (2002). The effects of alcohol withdrawal on mental state, interrogative suggestibility, and compliance: An experimental study. Journal of Forensic Psychology, 13, 53-67. Gudjonsson, G. H., & MacKeith, J. A. C. (1982). False confessions: Psychological effects of interrogation. In A. Trankell (Ed.), Reconstructing the past: The role of psychologists in criminal trials (pp. 253-269). Deventer, the Netherlands: Kluwer. Henkel, L. A., & Coffman, K. J. (2004). Memory distortions in coerced false confessions: A source monitoring framework analysis. Applied Cognitive Psychology, 18, 567-588. Horselenberg, R., Merckelbach, H., & Josephs, S. (2003). Individual differences and false confessions: A conceptual replication of Kassin and Kiechel (1996). Psychology, Crime, and Law, 9, 1-18. Hyman, I. E., Jr., Husband, T. H., & Billings, F. J. (1995). False memories of childhood experiences. Applied Cognitive Psychology, 9, 181-197. Inbau, F. E., Reid, J. E., Buckley, J.P., &Jayne, B. C. (2001). Criminal interrogation and confessions (4th ed.). Gaithersburg, MD: Aspen. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin,

114,3-28.

Kassin, S.M. (1997a). The psychology of confession evidence. American Psychologist, 52, 221-233. Kassin, S.M. (1997b). False memories against the self. Psychological Inquiry, 8, 300-302. Kassin, S.M. (2004). Videotape police interrogations. The Boston Globe, Op-Ed, April26, 2004, p. A-13. Kassin, S.M., Goldstein, C. C., &Savitsky, K. (2003). Behavioral confirmation in the interrogation room: On the dangers of presuming guilt. Law and Human Behavior, 27, 187-203. Kassin, S.M., & Gudjonsson, G. H. (2004). The psychology of confession evidence: A review of the literature and issues. Psychological Science in the Public Interest, 5, 35-69. Kassin, S. M., & Kiechel, K. L. (1996). The social psychology of false confessions: Compliance, internalization, and confabulation. Psychological Science, 7, 125-128. Kassin, S. M., & McNall, K. (1991). Police interrogations and confessions: Communicating promises and threats by pragmatic implication. Law and Human Behavior, 15, 233-251.

6.

INTERNALIZED FALSE CONFESSIONS

191

Kassin, S.M., Meissner, C. A., &Norwick, R. J. (2005). "I'd know a false confession ifl saw one": A comparative study of college students and police investigators. Law and Human Behavior, 29,211-227. Kassin, S. M., & Wrightsman, L. S. (1985). Confession evidence. In S. M. Kassin & L. S. Wrightsman (Eds.), The psychology of evidence and trial procedure (pp. 67-94). Beverly Hills, CA: Sage. Klaver, J., Gordon, R. V., & Lee, Z. (2003). Differential effects of minimization and maximization interrogation techniques and the role of plausibility in an experimental false confession paradigm. Paper presented at the International and Interdisciplinary Psychology & Law Conference, Edinburgh, Scotland. Kopelman, M.D. (1999). Varieties offalse memory. Cognitive Neuropsychology, 16, 197-214. Koutstaal, W., Schacter, D. L., Johnson, M. K., & Gallucio, L. (1999). Facilitation and impairment of event memory produced by photograph review. Memory & Cognition, 27, 478-493. Lassiter, G. D. (Ed.) (2004). Interrogations, Confessions, and Entrapment. New York: Kluwer. Leo, R. A. (1996). Inside the interrogation room. The Journal of Criminal Law and Criminology, 86, 266-303. Leo, R. A., & Ofshe, R. J. (1998). The consequences of false confessions: Deprivations of liberty and miscarriages of justice in the age of psychological interrogation. Journal of Criminal Law and Criminology, 88, 429-496. Lifton, R. J. (1956). "Thought reform" of Western civilians in Chinese prisons. American Journal of Psychiatry, 110, 732-739. Lindsay, D. S., Hagen, L., Read, J.D., Wade, K. A., & Garry, M. (2004). True photographs and false memories. Psychological Science, 15, 149-154. Loftus, E. F. (1993). The reality of repressed memories. American Psychologist, 48, 518-537. Loftus, E. F. (1997). Memories for a past that never was. Current Directions in Psychological Science, 6, 60-65. Loftus, E. F., & Hoffman, H. G. (1989). Misinformation and memory: The creation of new memories. Journal of Experimental Psychology: General, 118, 100-104. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4, 19-31. Magid, L. (2001). Deceptive police interrogation practices: How far is too far? Michigan Law Review, 99, 1168-1210. Mazzoni, G., Loftus, E. F., & Seitz, A. (1999). Changing beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13, 125-144. Mazzoni, G., & Memon, A. (2003). Imagination can create false autobiographical memories. Psychological Science, 14, 186-188. McCann, J. T. (1998). A conceptual framework for identifying various types of confessions. Behavioral Sciences and the Law, 16, 441-453. McCloskey, M., & Zaragoza, M. (1985). Misleading postevent information and memory for events: Arguments and evidence against memory impairment hypotheses. Journal of Experimental Psychology: General, 114, 1-16. McConkey, K. M., & Sheehan, P. W. (1995). Hypnosis, memory, and behavior in criminal investigation. New York: Guilford Press. McCormick, C. T. (1972). Handbook of the law of evidence (2nd ed.). St. Paul, MN: West. Meissner, C. A., & Kassin, S.M. (2002). "He's guilty!": Investigator bias in judgments of truth and deception. Law and Human Behavior, 26, 469-480. Munsterberg, H. (1908). On the witness stand: Essays on psychology and crime. Garden City, NY: Doubleday. Nathan, D., & Snedeker, M. (1995). Satan's silence: Ritual abuse and the making of a modem American witch hunt. New York: Basic Books.

192

KASSIN

Ofshe, R. J. (1992). Inadvertent hypnosis during interrogation: False confession due to dissociative state; misidentified multiple personality and the satanic cult hypothesis. The International]ournal of Clinical and Experimental Hypnosis, 4, 125-156. Ofshe, R. J., & Leo, R. A. (1997). The social psychology of police interrogation. The theory and classification of true and false confessions. Studies in Law, Politics and Society, 16, 189-251. Ofshe, R. J., & Watters, E. (1994). Making monsters: False memories, psychotherapy, and sexual hysteria. New York: Scribner. Ost, J., Costall, A., & Bull, R. (2001). False confessions and false memories: A model for understanding retractors' experiences. Journal of Forensic Psychiatry, 12, 549-579. Porter, S., Spencer, L., & Birt, A. R. (2003). Blinded by emotion? Effects of the emotionality of a scene on susceptibility to false memories. Canadian Journal of Behavioural Science, 35, 165-175. Redlich, A. D., & Goodman, G. S. (2003). Taking responsibility for an act not committed: Influence of age and suggestibility. Law and Human Behavior, 27, 141-156. Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. Ross, D. F., Ceci, S. J., Dunning, D., & Toglia, M. P. (1994). Unconscious transference and mistaken identity: When a witness misidentifies a familiar but innocent person. Journal of Applied Psychology, 79, 918-930. Russano, M. B., Meissner, C. A., Narchet, F. M., & Kassin, S. M. (2005). Investigating true and false confessions within a novel experimental paradigm. Psychological Science, 16, 481-486. Scheck, B., Neufeld, P., & Dwyer, J. (2000). Actual innocence. Garden City, NY: Doubleday. Schein, E. H., Schneier, I., & Barker, C. H. (1961). Coercive persuasion: A socio-psychological analysis of the "brainwashing" of American civilian prisoners by the Chinese Communists. New York: Norton. Scoboria, A., Mazzoni, G., Kirsch, I., & Relyea, M. (2004). Plausibility and belief in autobiographical memory. Applied Cognitive Psychology, 18, 791-807. Scullin, M. H., & Ceci, S. J. (2001). A suggestibility scale for children. Personality and Individual Differences, 30, 843-856. Sheehan, P. W., Statham, D., &Jamieson, G .A. (1991). Pseudomemory effects and their relationship to level of susceptibility to hypnosis and state instruction. Journal of Personality and Social Psychology, 60, 130-137. Sigurdsson, J., & Gudjonsson, G. H. (2001). False confessions: the relative importance of psychological, criminological and substance abuse variables. Psychology, C-rime and Law, 7, 275-289. Slobogin, C. (2003). Toward taping. Ohio State Journal of Criminal Law, 1, 309-322. Sullivan, T. P. (2004). Police experiences with recording custodial interrogations. Northwestern University School of Law, Center on Wrongful Convictions. Thomas, A. K., & Loftus, E. F. (2002). Creating bizarre false memories through imagination. Memory & Cognition, 30,423-431. Vrij, A. (2000). Detecting lies and deceit: The psychology of lying and the implications for professional practice. London: Wiley. Weingardt, K. R., Loftus, E. F., & Lindsay, D. S. (1995). Misinformation revisited: New evidence on suggestibility of memory. Memory and Cognition, 23, 72-82. Weinstein, E., Abrams, S., & Gibbons, D. (1970). The validity of the polygraph with hypnotically induced repression and guilt. American Journal of Psychiatry, 126, 1159-1162. Wright, L. (1994). Remembering Satan. New York: Alfred A. Knopf. Wrightsman, L. S., &Kassin, S.M. (1993). Confessions in the courtroom. Newbury Park, CA: Sage Publications.

II POTENTIAL SOURCES OF DISTORTED EYEWITNESS STATEMENTS AND POSTDICTORS OF STATEMENT ACCURACY

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

7 Internal and External Sources of Misinformation in Adult Witness Memory Deborah Davis University of Nevada, Reno

Elizabeth F. Loftus University of California, Irvine

At 8:30PM on July 17, 1996, Trans-World Airlines flight 800 crashed into the ocean. Reports consistent with the theory that a missile attack may have caused the crash soon spread among witnesses, investigators, and the media. Despite physical evidence pointing to another cause, discussion of the missile theory of the crash persisted for years. How did so many witnesses become convinced they had seen a nonexistent missile? And why did their reports continue to convince investigators, the media, and interested observers despite accumulating evidence of another cause? Some analyses suggest that post-event suggestion in the form of widespread speculation about causes of the crash may have brought about the initial potentially false reports consistent with the involvement of missiles. Within days of the crash, the media further fueled the missile theory by publishing seemingly supportive evidence and witness accounts. For example, the New York Daily News ran a story with the headline "Was TWA Jet Shot Down? Missile Attack, Bombing Probed." Other stories speculated about whether a bomb or possibly a surfaceto-air missile more likely caused the crash. Witnesses claiming to have seen a missile were interviewed extensively, leading other witnesses to come forward with similar stories. Their stories in turn encouraged investigators to process evidence with the missile theory as a primary hypothesis, perhaps biasing interpretation of that evidence toward support of the theory. This highly publicized TWA case provided a host of excellent real-life illustrations of the many biasing influences confronting witnesses and others in our judicial system. We address a number of these in this chapter in the context of our discussion of sources of misinformation in adult witness memory reports. Specifically, we review internal and 195

196

DAVIS AND LOFTUS

external influences that cause (a) selective memory or selective failure to remember, (b) false memories for things not actually witnessed or experienced, and (c) distortion or alteration of memories for things actually witnessed or experienced. We consider two general sources of these memory failures: (1) schematic and inferential processing and (2) sources of specific misinformation. We begin our review with consideration of the effects of schematic and inferential processing processes on memory and then turn to additional internal and external sources of specific misinformation. Perhaps the most central feature of organic information processors is the constructive nature of perception and comprehension. We add to or alter what we perceive or experience to render it comprehensible; we engage in this constructive processing at every level, ranging from simple sensory perception (for example, see Hoffman, 1998, for a fascinating review of constructive processes in visual perception) to processing of complex social events, oral and written discourse and narratives, and conversations; and we engage in constructive processing in response to stimulus features of the information being processed, as well as internal and external stimuli and events both preceding and following the target stimulus (for reviews see Davis & Friedman, this volume; Green, Strange, &Brock, 2002; Taylor &Crocker, 1981; Wyer &Srull, 1989; Wyer, 2002, 2004). Hence, original perceptions and judgments as well as later memory and judgments include a mixture of actual features of the original information or event and altered, distorted, or added features resulting from schematic/inferential processing and/or influential information external to the event. Predictable judgmental biases and distortions in memory result. Schematic processing involves relatively automatic (unconscious or uncontrolled) effects of the activation of specific concepts. Inferential processing, on the other hand, involves more deliberate and deliberative controlled processing of information in order to explain or disambiguate incoming information, or to select and plan responses to it (e.g., Graesser, Olde, & Klettke, 2002). We begin with consideration of these processes.

SCHEMATIC PROCESSING

Schematic processing is among the most pervasive sources of predictable errors in memory. Schemas are organized knowledge structures that include beliefs and expectations concerning the nature, characteristics, and behaviors or functions of objects, people, events, and other cognizable entities. Schemas selectively direct attention to relevant and useful information; facilitate the processes of perception, recognition, and understanding of incoming information; direct integration of new information with old; lend structure and meaning to experiences and events; guide information searches and retrieval; and provide the foundation for evaluative processes, problem-solving, anticipation of the future, setting goals, and making and carrying out plans. Without schematic processing, we would not know how to reduce the sea of input to manageable chunks, how to interpret what we see, what to expect of the world around us, or how to plan effective actions. However, although schematic processing is necessary and generally helpful, it also results in systematic errors in perception, judgment, and memory, including errors of the three categories we consider here: selective memory/forgetting, false mem-

7.

SOURCES OF MISINFORMATION

197

ories for events that did not occur, and distortions/alterations in memory for those that did occur (for reviews see Hastie, 1981; Kunda, 1999; Taylor & Crocker, 1981; Wyer & Srull, 1989; Wyer, 2002, 2004). A wide variety of schemas affect processing, including (a) category schemas telling us what the members of particular social or object categories are like; (b) person schemas involving perceptions and expectations of particular individuals; (c) self-schemas consisting of how we think about ourselves; (d) role schemas specifying characteristics and likely behaviors of those occupying specific social roles; (e) event schemas (or scripts) containing expected components and sequences for events, including crimes; (f) causal schemas containing theories of what causes what and how (for example, linking specific motives to specific crimes; Davis & Follette, 2002; 2003); and (g) procedural schemas specifying how specific tasks are executed, among others. We even have spatial schemas involving placement and motion that produce predictable errors of motion perception and spatial memory (e.g., Hubbard, 1996). Of particular importance to the legal arena is narrative processing and construction, which may be viewed as an overarching story structure involving a collage of relevant schemas and scripts that together inform the overall narrative. Essentially narratives may be constructed to explain complex events such as a crime, and may include a variety of subscripts and person, category, and causal schemas of various kinds involving particular relevant persons and episodes (both current and historical) within the overall narrative. Generally, we tend to think of and try to understand complex events in terms of narratives, as illustrated, for example, by the "story model" of jury decision making (e.g., Hastie & Pennington, 1995; Pennington & Hastie, 1986; 1992). For those who construct or listen to them alike, such narratives are particularly persuasive (e.g., Spiecker & Worthington, 2003) and influential for judgments and decisions of all sorts, from product choice to juror verdicts. Because they involve multiple schematic influences, they profoundly influence memory (see reviews in Green, Strange, & Brock, 2002; Hastie & Pennington, 2000; Wyer, 2004; Wyer, Ada val, & Colcombe, 2002).

Schema Activation To understand when schemas are likely to affect processing and memory, we must understand how they are activated. Schemas are sometimes activated by bottom-up processes, whereby properties of observed objects or situations themselves activate the schema (such as when seeing a man in a police uniform activates the "police" category schema, the "getting a traffic ticket" script, or the "why the man smoking marijuana ran'' causal schema). Schemas may also be activated by top-down processes in which a schema is activated by something other than properties of what is being observed (e.g., media accounts of the "missile theory"). Top-down activation and application of a schema can result in any of the following ways (for reviews see Fiske & Taylor, 1991; Hastie, 1981; Kunda, 1999; Taylor & Crocker, 1981). First, relatively stable features of observers may result in chronic activation of particular schemas and in their use (whether relevantor not) for processing incoming stimuli. Schemas important to the person's self concept (e.g., honesty as central to the self

198

DAVIS AND LOFTUS

concept of clergy) or to categorization/evaluation functions important to professional functioning (e.g., honesty as central to police interrogators' evaluation of suspects) will tend to be chronically activated and widely applied to incoming stimuli. A policeman, for example, might mistakenly interpret innocent behavior as suspicious because of chronic activation of crime/criminal-related schemas (see Thomassin & Alain, 1990, for illustration). Recovered memory therapists (who tend to chronically evaluate clients in terms of sexual abuse) tend to process, interpret, and search incoming information for abuse-relevant content (see Davis, Loftus, & Follette, 2001; Loftus & Ketcham, 1994). Similarly, the greater importance of personal appearance to women has been offered as one explanation for the superiority of women's memories for the appearance of others (i.e., descriptions of clothing/personal physical appearance vs. perpetrator identifications) (see Horgan, Mast, Hall, & Carter, 2004, for a recent review and demonstration). Even social roles can activate category-based schematic processing. Social power appears to encourage category-based perception of subordinates, whereas lower power positions appear to encourage individuated processing. For example, Richeson and Ambady (2003) found that placing whites in higher power roles relative to blacks increased racial bias, whereas Richeson and Ambady (2001) found similar effects of roles on automatic gender biases. Power can also affect other automatic associations, as illustrated by Bargh and his colleagues (Bargh, Raymond, & Pryor, 1995), who found that placing males in higher power roles relative to female subordinates tends to activate sexual thoughts (presumably the result of automatic power-sex associations). Second, temporary influences such as acute concerns, interests, or processing goals, or features of the context in which a stimulus is encountered can trigger specific schema activation. Rachel Remen, for example, tells of the time in medical school when she was learning to draw blood, when she found herself scanning the veins of strangers, categorizing them in terms of the ease with which one could draw blood (Remen, 1996, XXV). Acute concerns can also trigger general activation of a broad set of schemata. Rejection, for example, tends to promote selective attention to social rather than nonsocial information (e.g., Gardner, Pickett, & Brewer, 2000). Third, schemas recently activated (or primed) in irrelevant contexts may remain active or easily accessible and hence affect processing of new information (see Wyer & Srull, 1989; Wyer, 2004, for reviews)-as, for example, when a woman sitting on a jury for a child sexual abuse trial suddenly views affectionate behavior between her child and the child's uncle in more suspicious terms. Finally, schemas may be activated by external influences-including prominently statements from others-that directly suggest a particular event category ("Look at that fight!"), stereotype ("Hello, I'm doctor Smith."), causal schema ("Could a missile attack have caused this crash?" "He had a clear motive to kill him for the money."), script ("Well, if you did kill him, how might you have gone about it?"), person schema ("I thought he was rather evasive and deceptive."), or narrative ("It all began when three guys met by chance at a bar... "). Although necessary to facilitate information processing, once activated, schemas exert a wide range of biasing effects on perception and interpretation, ranging from what a person believes he or she has observed to the perceived meaning or underlying causes of the event.

7.

SOURCES OF MISINFORMATION

199

Although we do not exhaustively review the literature on schematic influences on processing and memory here, we illustrate these varied effects with selected legally rele, vant empirical demonstrations. (For more exhaustive reviews, see Davis & Follette, 2001; Fiske & Taylor, 1991; Green, Strange, & Brock, 2002; Hastie, 1981; Hastie & Pennington, 2000; Kunda, 1999; Taylor & Crocker, 1981; Wyer, 2004; Wyer et al., 2002; Wyer & Srull, 1989).

Selective Attention A pervasive consequence of schematic processing is selective attention to, and hence memory for, schema,relevant over schema,irrelevant aspects of incoming information. Given the same background file for a 30,year,old man, for example, a clinical psycholo, gist expecting to see the young man as a patient is likely to notice and remember infor, mation relevant to potential psychological problems and diagnoses, whereas a potential employer is likely to notice and remember information relevant to anticipated job (see Taylor & Crocker, 1981; Wyer et al., 2002; Wyer & Srull, 1989; Wyer, 2004, for reviews of empirical demonstrations). Relevant schema,inconsistent information is sometimes remembered better than schema,consistent information. Schema,inconsistent information may be surprising and require greater processing to understand or explain and hence may trigger elaborative in, ferential processing and therefore be remembered better than schema,consistent informa, tion requiring less processing (e.g., see Bower, Black, & Turner, 1979; Grasser, Gordon, & Sawyer, 1979; Schank, 1982; Hastie, 1980, for illustrations regarding text processing, event memory, and social information processing). Upon hearing that their pastor was heard screaming obscenities at a local elderly widow, parishioners could be expected to find this unexpected behavior surprising (even shocking) and to speculate at length about possible causes. Unfortunately for the pastor, given such extensive processing, the story would not soon be forgotten.

Schema-Based Effects on Categorization Schemas fundamentally involve the processes of categorization and understanding of incoming stimuli. Without schematic processing, we would be unable to understand much of what we see or hear (e.g., Bransford &Johnson, 1972). However, although this categorization function of schemas is crucial to perception, it can also lead to predictable errors. When a schema is activated, incoming stimuli are more likely to be categorized in terms of the activated schema. Race and gender categories tend to be automatically activated when others are perceived (e.g., Ito & Urland, 2003). Once racial schemas are activated, information and behaviors relevant to such racial schemas are likely to be noticed and remembered, whereas if others such as occupational schemas are also made salient, persons are likely be thought of in occupational categories as well, and occupation,relevant information is more likely to be noticed and remembered. Hence, the likelihood that a witness will remember specific information will be strongly affected by which schemas were active when the original event occurred. Schema,irrelevant

200

DAVIS AND LOFTUS

information, which is less likely to be closely attended to during the target event, is less likely to be encoded and available for later retrieval.

Trickle-Down Categorization and the Misperception of Weapons Labeling sets up an expectation of life that is often so compelling we can no longer see things as they really are .... We are in a relationship with our expectations and not with life itself. -Rachel N. Remen, Kitchen Table Wisdom, 1996, XXV

Once categorized, one may suffer what might be called trickle-down categorization, whereby other features or behaviors, and associated targets such as belongings, are categorized or interpreted in a manner consistent with the activated person category. In the broadest sense, this occurs when persons are assigned to undesirable social categories, thereby activating a generally negative bias in categorization and interpretation of other traits and behaviors. Even unrelated stimuli are affected by the activation of social stereotypes. For example, white subjects were able to categorize good words faster (in a lexical decision task) when primed subliminally with the word white and negative words faster when primed with the word black (e.g., Wittenbrink, Judd, & Park, 1997), and this was true particularly for words stereotypically associated with the racial category. Similarly, white subjects asked to determine whether words were "good" or "bad" categorized favorable words as "good" faster when primed subliminally with a white face, but categorized unfavorable words as "bad" faster if primed with a black face; whereas the reverse was true for black subjects (Fazio, Jackson, Dunton, & Williams, 1995). Even black clothing biased perception of offenders and suspects who wore it, toward greater aggressiveness and threatening intentions than those who wore light clothes, presumably because of the stereotypical association of the color black with evil, aggression, and bad characteristics or objects (e.g., Vrij, 1997; see also Frank & Gilovich, 1988, for a similar demonstration of effects of uniform color on perceptions of football players' actions). Generally, priming a particular stereotype can affect activation in three ways. First, the valence of a primed social category tends to facilitate perception of equivalently valenced categorizations and evaluations of associated stimuli (including those strictly temporally associated, as in the study by Fazio et al.). That is, persons primed with negative stimuli or words are quicker to recognize and categorize other negative stimuli. Second, the prime may facilitate perception of all prime-relevant stimuli, irrespective of valence, such as when priming a black face would facilitate recognition of both weapons and athletic equipment (e.g., Judd, Blair, & Chapleau, 2004; Kawakami, Dion, & Dovidio, 1998). Finally, the prime may facilitate perception of stimuli that are both stereotypic of the primed category and of the same valence (e.g., Wittenbrink et al., 1997), such as when priming a black face facilitates recognition only of bad things associated with blacks (but not good things associated with blacks or bad things associated with whites). These processes may be responsible for the variety of situations in which police judge and respond to the behaviors of black citizens with greater suspicion and more negative reactions (e.g., Winkel, 1999).

7.

SOURCES OF MISINFORMATION

201

Several notorious research examples of this phenomenon were inspired by the 1999 case of Amidou Diallo, a black man killed by four New York City police officers who misperceived the wallet in Diallo's hand as a gun. Shortly after this incident, researchers began to investigate whether these misperceptions could have resulted from racial stereotypes associating violence with blacks. Years earlier, Allport (1947) had shown that as participants transmitted descriptions of a scene depicting a white man holding a razor arguing with an unarmed black man, at some point in the series of retellings the black man was depicted as holding the razor instead for more than half of the groups of storytellers. To show that stereotypical associations of violence with blacks can have more immediate effects on reactions, Payne (2001) asked participants to rapidly categorize objects as either weapons or tools after a brief prime with a picture of a white or black face. Those primed with a black face were quicker to categorize weapons than those primed with a white face and, when required to respond very quickly, were more prone to mistake tools for weapons, whereas the reverse was true of those primed with a white face. Furthermore, these biases tend to work in reverse as well, in that activation of abstract concepts such as "crime" or "basketball" tends to induce attentional biases toward black male faces (Eberhardt et al., 2004). Recent studies have used computer game simulations that require participants to shoot only targets holding weapons. Respondents were more likely to "shoot" at black than at white targets holding harmless objects (see Correll, Park, Judd, & Wittenbrink, 2002; Eberhardt, Goff, Purdie, & Davies, 2004; Judd, Blair, & Chapleau, 2004; Payne, Lambert, &Jacoby, 2002; Plant, Peruche, & Butz, 2005). Even police officers show this bias (Plant & Peruche, 2005). Greenwald, Oakes, and Hoffman (2003) recently provided evidence that these errors can result both from increased perceptual difficulty in distinguishing between weapons and harmless objects when held by blacks, and a response bias toward treating objects held by blacks as guns. Although these studies did not address biases in memory for weapons (as opposed to errors of perception) and the various additional schemas that might facilitate mistaken memory or perception for weapons (such as those for criminals, hunters, police, adversarial encounters, gangs, etc.), these issues offer both important and fertile ground for future exploration. The first author was recently retained for a case involving contested possession of a weapon by a man in a domestic dispute. Could knowledge of characteristics of the defendant, the history and nature of the relationship with the woman, stereotypes involving social class or those involved in domestic disputes, or other features of the event in question trigger schema-based expectations that would bias perception or memory toward inclusion of a weapon in the defendant's hand? The "weapons false alarm" research clearly suggests that they could.

Schematic Effects on Interpretation In addition to affecting perception and categorization, schemas bias the search and interpretation of incoming information-and hence later memory for it-toward consistency with activated applicable schemas. Broadly, such biases are reflected in heavily documented tendencies to interpret school work, essays, art, daily routines, or behaviors and

202

DAVIS AND LOFTUS

other personal products or behaviors in more favorable terms when described as coming from relatively favored social groups (e.g., men, capitalists, whites, intelligent persons, or smart vs. dumb rats) than when described as coming from stigmatized groups such as women, communists, minorities, or the less intelligent (see review by Kunda, 1999). Such confirmation biases have also been investigated in studies of clinical diagnosis and judgment (e.g., Rosenhan, 1973). When diagnostic categories are made salientthrough labeling of a patient, or even simply by asking the question of whether or not the person may suffer a specific disorder-the clinician tends to confirm the diagnosis. Given the same file, for example, clinicians asked to determine whether the person suffers from bipolar disorder will tend to diagnose the person as bipolar at a greater rate than if asked to determine whether the person might be histrionic (see reviews by Maddux, 1993; Salovey & Turk, 1991; Turk &Salovey, 1988). In part, such confirmatory diagnoses result from biased interpretation of incoming information. However, research on hypothesis testing has shown that information searches are biased by schematic influences, such that hypothesis testers tend to seek information that would confirm the hypothesis, but fail to seek disconfirming (and more informative) information (e.g., Wason, 1960). Interviewers, for example, select questions that would tend to elicit answers confirming rather than disconfirming the hypothesis (e.g., Copeland & Snyder, 1995; Snyder & Thomsen, 1988). Such diagnosis-driven interviewing choices and interpretations of client symptoms and information have been widely criticized, particularly by critics of "recovered memory therapy" (e.g., Loftus, 1993; Loftus & Ketcham, 1994; McNally, 2003; Ofshe & Watters, 1994; Piper, Pope, & Borowiecki, 2000; Pope, Oliva, & Hudson, 2002). A few studies have addressed schema-based distortions in perception and memory for physical appearance. Perception of facial features or expressions, for example, can be biased by schema-driven interpretative processes (see Carroll & Russell, 1996). Students told that a man was a Gestapo leader responsible for barbaric medical experiments on concentration camp victims were more likely to judge his facial expression as cruel, whereas those told he was a leader of the World War II anti-Nazi underground were more likely to judge it as warm and kind (Rothbart & Birrell, 1977). Of particular importance for eyewitness identification processes, memory for faces spontaneously categorized or labeled as belonging to particular racial/ethnic groups tends to shift over time toward typicality for that category (Corneille, Huart, Becquart, & Bredart, 2004; see also Eberhardt, Dasgupta, & Banaszynski, 2003), and faces labeled as angry or sad are later remembered as physically appearing more extremely angry or sad (e.g., Halberstadt, 2003). Temporary personal goals can also bias facial perception. Self-protection goals, for example, led white participants to perceive greater anger in black and Arab faces (outgroups associated with threat); whereas mate-search goals led male participants to perceive more sexual arousal in attractive targets (Maner, Kenrick, Becker, Robertson et al., 2005). Script-based expectations can also bias interpretation of facial expressions. This was first demonstrated by Russian film director Kulechov (and hence dubbed the "Kulechov effect"). Kulechov created three short films that first showed either a dead woman, a dish of soup, or a girl playing, and then presented the face of an actor with a neutral expression. Viewers judged the identical expression as relatively sad, thoughtful, or happy,

7.

SOURCES OF MISINFORMATION

203

depending upon which version of the film they saw. Similar effects have since been documented in more systematic research. For example, students told a person in a photo was being threatened by a vicious dog tended to view his expression as fearful, whereas those told he had just won money tended to see the expression as one of happiness (Trope, 1986). Causal schemas can also distort memory for facial expressions. People asked to explain why ambiguous faces are expressing anger perceived and remembered those faces as angrier than those asked to explain why they were expressing sadness, and these distortions were greater than those produced by simple application of the emotion label to the faces (Halberstadt, 2003, 2005; Haberstadt & Niedenthal, 2001). Finally, Davis and colleagues (Davis, Vanous, & Cucciare, 2005) have recently shown that motion schemas involving expectations of continuity in motion can contribute to failures to perceive differences between separate actors (i.e., "change blindness"; Simons & Levin, 1998). The authors demonstrated that participants were more likely to confuse two actors when they appeared in a sequence that promoted the illusion of continuity in motion, than to confuse two who appeared in a sequence involving equal temporal separation, but that did not promote the illusion of continuity. Furthermore, the actor who appeared in an apparently continuous sequence with a thief stealing a bottle of wine was later misidentified as the thief at a greater rate than the actor who appeared in the apparently discontinuous sequence involving the thief.

Constructive Processes and Distortion Schemas not only bias veridical memory toward schema-relevant features, they also form the basis of constructive errors in memory whereby one "remembers" additional schema-consistent information that was not part of the witnessed event. In its simplest form, such constructive processes have been demonstrated in what is known as the DRM paradigm (e.g., Roediger, 1996; Roediger &McDermott, 1995; Roediger &McDermott, 2000). Participants in the DRM research are asked to study lists of semantically related words (such as dream, tired, bed, snore) and then to try to recall or recognize the words they have heard. A central finding is that participants are prone to falsely remember nonpresented but schema-associated words (sleep). Ironically, these effects can be stronger when participants attempt to suppress memory for stereotypical information (Aray, Ekehammar, & Akrami, 2003). Similar effects have been found for social stimuli such as descriptions or observations of people and social events. In a typical experiment, participants study a description of a person (or watch an event involving a person or people). A script or schema may be activated either bottom up (because the information suggests a particular social or event category) or top down (by labeling the participants or event with a particular category). Later, participants' memories for the original information or event are tested. Participants tend to falsely recall or recognize schema-consistent, but unpresented, information (see reviews by Graesser & Nakamura, 1982; Kunda, 1999; Taylor & Crocker, 1981; Wyer & Srull, 1989). They infer characteristics, events, or behaviors based upon information they are exposed to and later falsely recall that these constructions were actually part of the original event or information. Furthermore, the likelihood that any

204

DAVIS AND LOFTUS

given specific schema or script~based information will intrude into memory is greater the more strongly it is associated with a particular schema or script (e.g., Graesser & Naka~ mura, 1982). Although such processes have been demonstrated across a wide range of stimuli and schemas, two illustrations are particularly relevant to witness event memory. Several studies have investigated the effects of crime~related schemas or scripts on memory. Witnesses to crimes tend to accurately remember script~relevant actions (whether consistent or inconsistent) more than script~irrelevant actions; but they also falsely remember crime~script consistent but unwitnessed actions-particularly those more central to the script-and such intrusions are more likely when questioned at longer retention intervals, or when the original event was shorter or more ambiguous (Garcia~Bajos & Migueles, 2003; Greenberg, Westcott, & Bailey, 1998; List, 1986; Tuckey & Brewer, 2003, in press; see also Brewer & Treyens, 1981; Lampinen, Copeland, & Neuschatz, 2001; Neuschatz, Lampinen, Preston, Hawkins, & Toglia, 2002; Pezdek, Whetstone, Reynolds, Askari, & Cougherty, 1989, for similar results regarding script~ based memory errors for teacher behaviors and schema~based memory intrusions for ob~ jects). Interestingly, one such intrusion involved the misreport of a bag as a gun (Tuckey & Brewer, 2003). Hence, like the previously discussed race~based expectations of vio~ lence, the script~based expectation that a robber would have a gun led subjects to mis~ perceive or misremember the bag. Such crime script~based expectations seem to perva~ sively influence the trial process, as they have been shown to also bias jurors' memory for witness testimony (Holst & Pezdek, 1992; Smith & Studebaker, 1996) and verdicts (Smith, 1991). Hannigan and Reinitz (2001) investigated the impact of causal schemas on mem~ ory intrusions. Participants viewed a scene involving a potential cause without an effect (a person removing an orange from the bottom of a stack) or an effect without a cause (a spilled stack of oranges). Those who saw the spilled stack of oranges later tended to falsely remember seeing the cause (i.e., someone removing an orange from the bottom of the stack). Hence, it appears that script~ based intrusions will include typical actions or component events, including those expected to play a causal role in the overall event. Schematic processes may also be responsible for source memory errors. Statements we believe, for example, tend to be misremembered as coming from a more credible source, presumably as a result of expectations that believable information comes from credible sources (e.g., see review in Fragale & Heath, 2004). These authors provided a particularly disturbing demonstration of this principle involving source attributions for incriminating evidence. That is, participants led to believe a suspect was guilty tended to misremember incriminating evidence as coming from a higher credibility source than those led to believe he was not guilty. Several studies have shown that we are prone to misremember who said what in pre~ dictable schema~driven ways. When asked to remember which of several speakers con~ tributed a particular statement to a discussion, for example, participants are more likely to misattribute the statements of one member of a social category to another person from the same category (i.e., those of the same gender, race, age, attractiveness, home town, clothing color, educational status, etc.) than to someone from a different categoryperhaps because they seem more similar, and source confusions between similar sources

7.

SOURCES OF MISINFORMATION

205

are more likely than between dissimilar source categories (see reviews in Davis & Friedman, this volume; Davis, Kemmelmeir, & Follette, 2005; Klauer & Wegener, 1998). For this reason, objects and episodes from one action sequence are more likely to be misattributed to an episode of the same general class (e.g., Restaurant Episode A to Restaurant Episode B) than to one of a different class (e.g., Hannigan & Reinitz, 2003). Moreover, schematic source errors can derive from trait schemata. For example, trait-related statements, such as "I work out almost every day" tend to be disproportionately misattributed to a person fitting that trait (e.g., an athlete) when they were actually made by a different speaker (Mather, Johnson, & De Leonardis, 1999; see also Sherman & Bessenoff, 1999; and Bayen, Nakamura, Dupuis, & Yang, 2000, for demonstrations of schema-driven source monitoring errors associating objects with rooms). Such schematic effects also appear to be stronger among those with poorer overall memory (Spaniol & Bayen, 2002), older persons (see Davis & Loftus, 2005, for a review), and those with stronger stereotyped associations (as measured by the Implicit Attitudes Test; Greenwald, McGhee, & Schwartz, 1998; Gawronski et al., 2003). Although source errors may sometimes be benign, they are raised and disputed surprisingly often in legal contexts. Which of a group of persons uttered a threat may be crucial, for example, and the target of an utterance can be equally important (such as who the threat addressed; see Davis & Friedman, this volume; Davis, Kemmelmeier, & Follette, 2005, for illustrations and research).

Reconstructive Processes Information acquired after exposure to target events can also activate schema-based reconstructive memory processes that bias memory in favor of information consistent with the newly activated schema-processes apparently central to the reports of witnesses to the TWA incident. Post-event schema activation can occur through several processes. Post-Event Information. Post-event information from external sources may activate a particular schema. For example, in a seminal study of such reconstructive processes, Snyder and Uranowitz (1979) had participants read about "Betty K" and after a delay told them that Betty was either heterosexual or gay. In a later memory test, participants reported memories of the original description of Betty that were consistent with the newly acquired category labels of "gay" versus "heterosexual." They were more likely to correctly remember previously presented schema-consistent information, but also incorrectly reported nonpresented schema-consistent information. A classic study by Cohen (1981) demonstrated that category schemas exerted essentially equivalent effects on memory, whether activated at the time an event was witnessed or later. Participants witnessed a video of a woman interacting with her husband. They were told either before or after witnessing the event that the woman was a waitress or a librarian. Those told she was a waitress tended to remember her appearance and behavior as more consistent with their stereotypes of waitresses (e.g., blond hair, drinks beer, affectionate with her husband, listens to rock and roll, etc.), whereas those told she was a librarian tended to remember librarian-consistent behaviors and attributes (such

DAVIS AND LOFTUS

206

as brown hair, drinks wine, listens to classical music, etc.; see Davis & Follette, 2001; Hirt, McDonald, & Markman, 1998; Schacter, 1996, for reviews of the tendency to reconstruct memories of past events toward consistency with current expectations).

Present Knowledge and Reconstructive Hindsight Biases Life is lived forwards, but understood backwards. -Sll)ren Kierkegaard

Knowledge of current events, emotions, or outcomes is well known to bias memory and judgments of processes leading up to them. Victims of the hindsight bias (Fishhoff, 1975) have been shown not only to overestimate, in hindsight, how likely a particular outcome was to occur, but also to overestimate what others could have, or should have, anticipated in foresight, thus giving them undue credit or blame for the outcome. Moreover, unaware of their own hindsight biases, they cannot accurately remember their own behavior or judgments before the outcome was known, recalling instead that they were wiser ("I knew it would happen this way."), and more confident ("I was sure of it!"), and that they had behaved more consistently with this knowledge ("I told you so!") before the event than was actually the case (see reviews by Christensen-Szalanski & Willham, 1991; Davis & Follette, 2001; Hawkins & Hastie, 1990; Hoffridge & Pohl, 2003). Baruch Fischhoff (1975) described this tendency to view past events as inevitable consequences of their predecessors "creeping determinism." In effect, outcome knowledge tends to activate causal schemas for how such outcomes are produced, which in turn engage schema-driven searches and reviews of past information and reconstructive memory processes that reorganize memory. These processes bias memory toward retrieval of information consistent with the now activated causal schema and promote schemaconsistent memory distortions and judgments of past events (see recent theoretical analyses of hindsight effects by Erdfelder & Buchner, 1998; Roese & Olson, 1996; Winman, Juslin, & Bjoerkman, 1998). A number of studies have examined implications of the hindsight bias in the legal system. Judgments in hindsight can be inappropriately harsh because people blame themselves or others for incorrect decisions, even though in foresight they were appropriate in light of available information. This has been shown with respect to medical and psychotherapeutic judgments and decisions, financial advisement, behaviors prior to rape, and judgments of probable cause, suggesting that both decisions to file suit and jury verdicts can be influenced by hindsight (Carli, 1999; Goggin & Range, 1985; see reviews in Christensen-Szalanski & Willham, 1991; Davis, 1991; Hawkins & Hastie, 1990; Hoffridge & Pohl, 2003). Moreover, the production of evidence, rather than reactions to it, can be influenced by hindsightful witnesses, as illustrated by Ulric Neisser's (1981) classic study of hindsight biases in John Dean's testimony before the Watergate hearings. John Dean testified that he warned the president that the cover-up would eventually be discovered and lead to Nixon's loss of the presidency, whereas no such warnings actually took place (see Davis & Friedman, this volume, for discussion of studies assessing biases in John Dean's memory, including hindsight and other schematic processing effects).

7.

SOURCES OF MISINFORMATION

207

Davis and colleagues (Davis, Lopez, Koyama, et al., 2005) recently found hindsight biases in memory for a conversation between a couple arguing over an interaction between the girl and a male classmate. Her boyfriend exhibited extreme jealousy and suspicion of her relationship with the classmate. Later half of the participants learned that the couple had broken up, whereas the other half were told the female was found dead, and her boyfriend was charged with murder. Those who learned he was charged with murder were both more likely to accurately remember threatening statements he made to his girlfriend during the argument, and more likely to falsely remember that he had hit his girlfriend during the vignette, and that he had also threatened her classmate. Hindsight and other retrospective biases are likely to exert pervasive influences on crucial witness reports, distorting memories of the past (including one's own) toward consistency with present knowledge (e.g., Dawes, 1991; Ross, 2000; Rubin, 1996; Schacter, 1996). Countless reports concern events, objects, or people witnessed well before the litigated event occurred, and often even longer before the witness was questioned about it. When asked about these issues after learning of the target event, retrospective or hindsight biases likely come into play. A witness who observed a vigorous conversation between a couple may remember it as "a fight" in light of new knowledge that the wife has been found murdered. Past behaviors of an employer newly accused of sexual harassment may suddenly be remembered as more sexually suggestive or coercive than before the accusation. Even memories of our own attitudes, emotions, or behaviors are colored by hindsight ("I always felt creepy when he was around"; see Dawes, 1991; Keuler & Safer, 1998; Levine, 1997; Ross, 2000; Safer, Levine, & Drapalski, 2002). Such processes may well have contributed to witness reports of inappropriate interactions with children in the case of superstar Michael Jackson. The memories of witnesses who learned of the allegations of sexual abuse well after they witnessed Jackson's interactions with the children could well have been distorted by current knowledge of the accusations. Hindsight biases may also contribute to the development of false "recovered memories" of sexual abuse, in that once the person has developed a self-schema as an abuse survivor, past events may be selectively searched and reinterpreted to fit the new selfidentity (see Loftus & Davis, 2006; Loftus & Ketcham, 1994; McNally, 2003). Indeed, researchers interested in autobiographical recollections, life stories, and narrative approaches to understanding experience have highlighted the way personal "identity" consists of evolving narrative constructions of the self that continually restructure the past to fit current conceptions of the self (see Fivush & Haden, 2003; Greenwald, 1980; McAdams, 2001; Ross, 2000; Rubin, 1996; Wilson & Ross, 2003 for reviews) or levels of self-esteem (e.g., Christensen, Wood, & Barrett, 2003; Tafarodi, Marshall, & Milne, 2003). In countless ways, old events, emotions, and behaviors take on new characteristics and meanings as new knowledge provides revised context for their interpretation.

SOURCES OF SPECIFIC MISINFORMATION

In addition to schema-based intrusions, misinformation may creep into memory from a variety of external and internal sources of specific incorrect information. Specific

DAVIS AND LOFTUS

208

misinformation may be mistakenly incorporated unchanged into memory, thereby simply adding false information to (for example, inserting the memory of a car) or replacing (substituting memory of a truck for that of a car) veridical memory. However, it may also distort memory (for example, changing the color or location of the car) and activate schematic processing, which causes far-reaching related inferential processing and associated distortions in memory. In this section, we consider sources of specific misinformation, provide examples of misinformation effects on event memory, and discuss proposed mechanisms via which misinformation affects memory.

External Sources of Misinformation

Media Media coverage of criminal and civil litigation is perhaps among the most common sources of misinformation in witness memory. Sometimes within hours of a target crime, the incident, and perhaps various witness reports and other evidence, are reported in television news accounts-reports that are soon followed by others in printed media. And these reports are typically repeated and elaborated upon in subsequent reports as the case unfolds or nears trial. Research has established the biasing effects of such reports and pretrial publicity among the community and prospective jurors, and motions for change of venue for the trial are commonly filed and granted on the basis of demonstrations of widespread case-relevant knowledge and biased opinions among the population of jurisdictions saturated with such publicity (e.g., Abbot & Batt, 1999; Davis & Follette, 2004). There is experimental as well as anecdotal evidence that media reports can also distort witness memory. For example, Loftus and Banaji (1989) directly tested the impact of case-related publicity on witness memory. The authors presented subjects with a 4-minute videotape of a robbery and shooting incident. It depicted two police officers on rounds at night who hear shots and confront two robbers. One officer and one robber were injured in the confrontation. Later, subjects were exposed to a television report of the incident, which they viewed under the premise that they were to decide whether the reporter was sufficiently talented to obtain a job in a major national news operation. The TV reporter inserted misinformation into the account shown to "misled" subjects but not into the version shown to "control" subjects. For example, in the misled version she suggested that there were three robbers rather than two, or that the victim liquor store was named "Midtown Liquor" rather than "Pete's Liquor." When tested later, control subjects almost never gave the suggested answer (1%), but misled subjects frequently did (28%). Loftus and Castelle (2000) offered a systematic analysis of the influence of media reports on witness memories of the widely publicized crash of TWA flight 800. Within days of the crash, the media began to present and discuss the theory that the plane had been downed by a missile attack-including graphic illustrations of how the shooting could have occurred and of witness accounts that could be interpreted as support for such a theory. The authors' account of the investigation of the case illustrated how the testimony of the primary witnesses was altered over time as publicity of the missile theory

7.

SOURCES OF MISINFORMATION

209

escalated, and how ultimately 183 other witnesses came forward with testimony supporting the theory. A more controlled study was conducted by Hans Crombag and his colleagues (Crombag, Wagenaar, & Van Koppen, 1996), who studied Dutch residents' memories of the crash of an El Al Boeing 747 into an 11-story Amsterdam apartment building. Media coverage never included the crash itself, but began within the first hour after the crash and included films of the ensuing fire and rescue operations. In two surveys, 10 months after the crash, the authors found that 55% and 66% of respondents reported having actually seen the crash on TY, and of those many gave vividly detailed descriptions of the crash they could not have actually seen. The authors attributed these false memories to the effects of imagination. That is, based on widespread media descriptions of the crash, participants would tend to visualize the crash and hence later falsely attribute the source of these images to TV videos of the crash rather than their own imaginations.

Memory Conformity: Effects of Co-witnesses The first author served as a memory expert for the case of Mr. Stroud, who was accused of robbing a Portland bank. Evidence included a police tape in which an employee of the bank reported the robbery. Having been asked to provide a description of the assailant, the employee responded, "Wait a minute, we're getting a consensus on that." What effect might such consensus-seeking discussions or other influences between witnesses have on memory accuracy? If witnesses seek consensus, disagreements must necessarily be resolved through alteration of conflicting accounts toward consistency. However, even in the absence of such absolute consensus seeking, witness accounts are influenced by those of others (for example, see accounts in Memon & Wright, 1999; Schacter, 2001, of influence between the three witnesses who reported that Timothy McVeigh (the Oklahoma City bomber) was accompanied by another man when he rented the truck used in the bombing). There is direct evidence from the eyewitness literature demonstrating that co-witnesses can influence the accuracy of one another's testimony as well as their confidence in those accounts. For example, Loftus and Greene (1980) found that participant-witnesses incorporated misleading details from other witnesses' written descriptions into their own descriptions of target faces. More recently, memory conformity has been demonstrated for diverse stimuli such as cars (Wright, Self, & Justice, 2000), details of written stories (Betz, Skowronski, & Ostrom, 1996), objects from various scenes (Meade & Roediger, 2003; Roediger, Meade, & Bergman, 2001), time estimation (Conway, 2004), and reports of witnessed criminal events (Gabbert, Memon, & Allan, 2003). Perhaps the clearest demonstration of these effects can be seen in a series of three studies by Shaw, Garven, and Wood (1997). In the first study, student witnesses were questioned in ways that were intended to be analogous to the experience of a witness who receives misinformation from an interviewer or questioner about what other witnesses have already said. A second and third study simulated the situation where witnesses received the information directly from co-witnesses. In all three studies, witness

210

DAVIS AND LOFTUS

reports were significantly affected by information about the reports of other witnesses, whether they received the information directly from the co-witnesses or indirectly from the interviewer. The most errors, however, were committed by witnesses who received incorrect information both directly from other witnesses and indirectly via the misleading questions of the interviewers. Similar results were obtained in a study of the induction of false memory reports by children of wrongdoing by a classroom visitor. False reports were most likely to come from children exposed to both reinforcement of such claims by the experimenter and co-witness information from other children (Garven, Wood, & Malpass, 2000). Consistent with the general literature on opinion and behavioral conformity, memory conformity appears to be greatest among those with less clear recollections of the original stimulus and among those exposed to more sources of bogus information (larger groups), and less when there are dissenters who break the unanimity of the bogus group reports or when the credibility of the other person is suspect (Hoffman, Granhag, See, & Loftus, 2001; Walther et al., 2002). Participants in the Walther et al. study were also less affected by bogus group judgments for stimuli that were sufficiently unusual as to cause them to believe they would have remembered them if they had seen them before, and under conditions where a clear lack of memory would reasonably be a valid indication that the event did not occur. In other words, participants appeared to use "metacognitive knowledge," that is, situational cues (such as original conditions of encoding), along with characteristics of the bogus group claims (such as whether it was plausible that the participant could forget the particular style of stimulus the group claims to have seen), group size, and degree of agreement among group members, to determine whether their own memories or the claims of the group members would be most likely to be accurate. These effects occurred for both correct and incorrect claims. The witness memory-conformity studies suggest that discussions between co-witnesses have great potential to influence the testimony of all witnesses, with far-reaching consequences. First, co-witnesses can shape one another's initial reports. When interviews by police or other investigators suggest information similar to that being conveyed by other witnesses, a target witness will be particularly likely to adopt the suggested account. Furthermore, a witness's first statements have been shown to shape reports during subsequent interviews or court appearances. In other words, once a witness has told his or her version of what happened, that witness is likely to stick with that account in the future. This phenomenon is sometimes referred to as the freezing effect (see Loftus, 1979). As time goes by, it becomes increasingly difficult for a witness to change or back down from an initial statement or identification. The first statements of even a single witness may also influence the subsequent investigation of a case, just as helicopter pilot Meyer's initial report of the TWA explosion triggered the "missile theory" of the crash and thousands of hours of investigation of that theory and influenced countless reports of witnesses who heard Meyer's account on television before their own interviews. Similarly, based upon one witness's initial account of a criminal incident, police may construct lineups, choose particular mug shots, interview specific witnesses, formulate detailed theories about the commission of the crime, and

7.

SOURCES OF MISINFORMATION

21 I

even make an arrest. Hence, even minor errors, misstatements, or omissions in one witness's initial memory report can have far-reaching effects on the investigation of the case, whether civil or criminal.

Inflation of Witness Confidence Through Collaboration/Corroboration. Discussions with other witnesses may not only contaminate memory, but also inflate confidence in information "confirmed" by others. Luus and Wells (1994}, for example, showed that witness confidence in perpetrator identifications was affected by confirming or disconfirming identifications by other witnesses. They staged a theft in front of 70 pairs of witnesses. Witnesses who were told their co-witness had made the same identification showed an increase in confidence in their identifications, as expressed to an ostensible police officer. Those who thought their co-witness disagreed showed deflation of confidence (see Semmler, Brewer, & Wells, 2004; Wells & Bradfield, 1998; Wells, Olson, & Charman, 2003, for similar demonstrations of the effects of police feedback on witness confidence; and Hafstad, Memon, & Logie, 2004; Meudell, Hitch, & Kirby, 1992; Stephenson & Wagner, 1989, for other demonstrations showing that collaboration inflates confidence in eyewitness identification; and Stephenson, Bandstatter, & Wagner, 1983, for effects of collaboration on validity and inflation of confidence in story recall; however, also see Allwood, Granhag, & Johansson, 2003, for conditions under which collaboration may increase the correspondence of witness confidence and witness accuracy). Inasmuch as witness confidence is perceived by jurors as an important indicator of accuracy, it is not surprising that feedback-inflated confidence can in turn affect the credibility of the witness's later testimony. Luus and Wells showed tapes of the 70 witnesses to other mock jurors who judged the credibility of their testimony. Witnesses who had experienced confirming feedback from their co-witnesses were judged as more credible by the mock jurors than those who had experienced disconfirming feedback, or who had not been given any feedback. In other words, witness confidence was affected by information about the opinions of other witnesses, and the witnesses' confidence, in turn, affected mock juror judgments of their credibility. Finally, it is important to point out that social influence between witnesses may occur because of influence of the information conveyed between them and the potential for motivational distortions in memory created by any desire they may have to avoid rejection, or to help or support other fact witnesses, investigators, or parties to the case (see Asch, 1955; see also Reysen, 2005, for evidence that memory conformity results from both processes}. To the extent this motivation exists, it would tend to both distort the witness's own memories in the direction consistent with the interests of the other person, and to render the witness more susceptible to the influence of any source who might be reporting the events in line with the other's interests. Expert testimony on issues of witness contamination can be valuable, because jurors will not understand the potential for witness collaboration to influence both memory and confidence. Rather they instead assume that confidence strongly reflects accuracy (see Davis & Follette, 2001, for a review, as well as Shaw, McClure, & Dykstra in this volume and Leippe & Eisenstadt in Volume 2 of this handbook).

DAVIS AND LOFTUS

212

Forensic and Clinical Interuiews and Procedures Among the most prolifically investigated issues in eyewitness psychology is the manner in which forensic interviewing procedures can lead to fabrication or distortion. This line of research began in the 1970s with the early studies of Loftus and her colleagues of the effects of question wording (see Loftus, 1979, for a summary). Participants were shown videos and other event depictions and later interviewed about their memories of the events. In some studies, for some participants the interviewers' questions more strongly implied particular information than for others (e.g., "Did you see 'the' [vs. "a"] broken headlight?"). The more strongly the interviewer's questions implied a particular answer, the more likely participants were to report memories consistent with the expectation. In other studies, wording was shown to affect estimates such as speed (e.g., "How fast was the car going when it 'hit' [vs. "bumped into" or "smashed"] into the other car), such that wording implying more forceful impact led to greater estimates of speed, as well as greater numbers of false reports of having seen broken glass. Although not conducted in the context of witness memory studies, studies among linguists and social psychologists have shown that both memory and inferences are affected by such additional features of questions as whether the target (e.g., an alleged rape victim) is described as the subject ("Did you dance with the defendant?") versus the object of the sentence ("Did the defendant dance with you?"); or whether the question inquires about action verbs ("Did you talk to the defendant?"), versus state verbs ("Did you like the defendant?). Respondents providing answers to questions involving actions (as opposed to states) of the target and those depicting the target as the subjects (rather than objects) of the sentence tend to attribute more causality to the target for the event in question (e.g., "I talked to the defendant because I... " versus "I talked to the defendant because he . .. "), and that these differences are associated with a variety of other inferences that affect memory, judgments regarding the event, as well as future predictions involving the parties andrecurrent or similar events. Furthermore, judgments by those who listen to the responses are reliably affected by the wording of the questions (see Semin & De Poot, 1997a,b). Loftus and her colleagues subsequently demonstrated that false information suggested during interviews can lead to the development of false memories for objects or events that were not included in the originally witnessed events-a phenomenon dubbed the misinformation effect. In these studies, participants were again shown depictions of scenes or events. However, the experiments then proceeded in two phases. In the first phase, witnesses were interviewed about what they had just witnessed. However, some were given misinformation during the interview. In some studies, they received false information about something that had been seen (e.g., that they had seen a stop sign when they had actually seen a yield sign), whereas in others the false information suggested events or objects not actually witnessed at all (such as broken glass, barns, or wounded animals). Participants were later interviewed again, after varying times of delay. In some early studies, Loftus showed that (mis)leading questions can cause witnesses to falsely remember seeing a yield rather than a stop sign, a conspicuous barn in a bucolic scene actually containing no buildings, broken glass, tape recorders or wounded animals that were not present, incorrect colors of objects, curly rather than straight hair,

7.

SOURCES OF MISINFORMATION

213

and Minnie Mouse when they actually saw Mickey. Thus, these studies showed that (mis)leading questions can lead people to both add things to their memories and alter memories of things they did see (Loftus, 2003; Loftus & Greene, 1980; Loftus & Palmer, 1974; Loftus, 1975, 1977, 1979). Unfortunately, interviewees confronted with interview, ers' misinformation are most likely to ignore it rather than to comment on or correct it (Hunt & Borgida, 2001), perhaps enhancing their vulnerability to later incorporating it into memory. Furthermore, a number of studies exposed participants to written or oral narratives after they had witnessed the original events. Like the biased questioning pro, cedures, these post,event accounts led participants to alter or add to their memories of the original event (e.g., Eakin, Schreiber, & Sergent,Marshall, 2003). Susceptibility to misinformation effects is generally related to the ability to detect discrepancies between event information and post,event misinformation (the "Principle of Discrepancy Detection"; Tousignant, Hall, & Loftus, 1986; see review in Gerrie, Garry, & Loftus, 2005). Hence, factors associated with the ability to detect discrepancy (such as characteristics of the witness or event, or of the similarity between the event and misinformation) or with the tendency to look for or notice discrepancies (such as the credibility of the source or the plausibility of the misinformation) predict the magni, tude of the effect. Misinformation effects tend to be stronger for some people (see Eisen, Winograd, & Qin, 2002; Gudjonsson, 2003, for reviews of individual differences in general suggestibil, ity). Misleading information creates more distortion among children than among adults (e.g., Ceci, Ross, & Toglia, 1987) and among elderly versus younger adults (see Davis & Loftus, 2005; Mueller, Johnson & Ceci, 2004; Polczyk et al., 2004, for reviews). Age dif, ferences may reflect the importance of cognitive resources, as misinformation effects are stronger when one encounters or retrieves information under conditions of stress, lim, ited attentional resources, or heavy cognitive load (Klein & Boals, 2001; Zaragoza & Lane, 1998), and among persons with chronically low cognitive resources, such as those with lower working memory capacity (Jaschinski & Wentura, 2002), poorer overall re, call for the original information (Liebmann et al., 2002; Tomes & Katz, 1997), lower IQ (see Gudjonnson, 2003), or poorer neuropsychological functioning (Lee, 2004). Surpris, ingly, acute anxiety is has been associated with reduced susceptibility (e.g., Ridley & Clifford, 2004; Ridley, Clifford, & Keogh, 2002), although this relationship can be re, versed for suggestibility in other contexts. Sometimes one's abilities confer increased susceptibility, as has been shown for those with enhanced ability or propensity for visual imagery (Dobson & Markham, 1993; Lindsay, 1990; Tomes & Katz, 1997; Winograd, Peluso, & Glover, 1998), although results for this have been somewhat mixed (see dis, cussion in Hekkanen & McEvoy, 2002). Additionally, several personality variables are associated with susceptibility to mis, information, including neuroticism (e.g., Liebman, McKinley,Pace, Leonard, Sheesley, et al., 2002; Zanni & Offermann, 1978), "se1£,monitoring" (the tendency to monitor one's own behavior for impressions made on others; Lassiter, Stone, & Weigold, 1988), empathy (Tomes & Katz, 1997), dispositional tendencies toward dissociation (Hekkanen & McEvoy, 2002; see review by Eisen & Lynn, 2001), and "absorption'' (as measured by the Tellegen Absorption Scale; Drivdahl & Zaragoza, 2001; Eisen & Carlson, 1998).

214

DAVIS AND LOFTUS

Temporary states such as hypnosis can also magnify misinformation effects (e.g., Scoboria, Mazzoni, Kirsch, & Milling, 2002), as can even the belief that one is under the influence of alcohol (e.g., Assefi & Garry, 2002) or that one is a "poor" rather than "good" eyewitness (Roper & Shewan, 2002). Nevertheless, the misinformation effect is widely generalizable-even to 3-month-old infants (Rovee-Collier, Borza, Adler, & Boller, 1993), pigeons, rats (Garry & Harper, 2005; Harper & Garry, 2000), and gorillas (Schwartz, Meissner, Hoffman, Evans, & Franzier, 2004)-and applies to other domains, such as experiences with products and public figures (e.g., the advertising misinformation effect; Braun, 1999; Braun, Ellis, & Loftus, 2002; Braun & Loftus, 1998; Cowley & Janus, 2004). Misleading questions tend to induce the greatest distortion when introduced after a delay, instead of immediately after the event (Loftus, 1979), when the wording of the misleading question is more definite (e.g., "Did you see the stop sign?" versus "Did you see a stop sign?" Loftus & Zanni, 197 5); when the source of the misleading information is credible, attractive, of high status or power, or apparently unbiased (Bregman & McAllister, 1982; Ceci, Ross, & Toglia, 1987; Dodd & Bradshaw, 1980; Loftus, 1980; Smith & Ellsworth, 1987; Vornick, Sharman, & Garry, 2003; see Blank, 1998; Vornick et al., 2003 for reviews); and when the person is asked about perceptual details of the suggested event (presumably inducing greater visual imagery of the suggested events promotes false memory; Drivdahl & Zaragoza, 2001). The emotionality of a witnessed scene also appears to affect susceptibility to misinformation such that misinformation regarding highly negative scenes is roughly twice as likely to induce false memories as that regarding either positive or neutral scenes (Porter, Spencer, & Birt, 2003). Misleading information also creates more distortion when it is more accessible (Eakin, Schreiber, & Sergent-Marshall, 2003), for peripheral rather than central detail (e.g., Loftus, 1979; Roebers & McConkey, 2003; Sutherland & Hayne, 2001), for more poorly remembered information (e.g., Marquis, Marshall, & Oskamp, 1972; Read & Bruce, 1984), when tested by recognition versus recall procedures (Sutherland & Hayne, 2001), and when it is subtle, rather than blatant (Loftus, 1979; Ornstein, Gordon, & Larus, 1992). On the other hand, less distortion is observed for negative information with high personal significance (e.g., Bruck, Ceci, & Hembrooke, 1998), for persons specifically asked to indicate the source of the misinformation at retrieval (e.g., Dodson & Johnson, 1993) or to ignore information from the source of the misinformation when reporting memories (e.g., Lindsay, 1990), for misinformation introduced by a suspect source (Dodd & Bradshaw, 1980), for secret versus nonsecret information and when the secret giver as opposed to the nonsecret giver interviews the person at retrieval (Wilson, Powell, Raju, & Romeo, 2004), and for those warned of misinformation effects (Eakin et al., 2003; Highhouse & Bottrill, 1995) or experimenter trickery (Chambers & Zaragoza, 200 1). The misinformation effect may also be reduced by explicit training in source monitoring (e.g., Theirry & Spence, 2002) and generally varies with the conditions of encoding or retrieval (see Ayers & Reder, 1998; Blank, 1998; Brainerd & Reyna, 2002; Eakin et al., 2003; Frost, 2000; Gerrie et al., 2005; McCloskey & Zaragoza, 1985; Roehers & McConkey, 2003; Schooler & Tanaka, 1991, for discussions of theoretical mechanisms of the misinformation effect and the conditions under which misinformation is

7.

SOURCES OF MISINFORMATION

215

most likely to induce false memories). Nevertheless, in the years since Loftus's original demonstrations of the misinformation effect, research has shown how (mis)leading in~ formation can cause a person to develop false memories of much more dramatic inci~ dents, including those involving the self or others (see Loftus & Davis, 2006). Faced in the late 1980s with rapidly escalating accounts of child abuse and adult "recovery" of repressed memories of such abuse, psychologists began to devote increas~ ing attention to the issue of how false memories for more dramatic events, and particu~ larly dramatic autobiographical events, might be implanted in memory. Because recov~ ered memory therapy commonly includes direct suggestions from the therapist that the client has been abused, many such studies examined the impact of direct suggestion that a particular event had occurred. For example, Mazzoni and her colleagues (Loftus & Mazzoni, 1998; Mazzoni, & Loftus, 1996; Mazzoni, Loftus, Seitz, & Lynn, 1999; Mazzoni, Lombardo, Malvagia, & Loftus, 1999) studied direct suggestion in the form of a clinical psychologist's bogus interpretation of dreams. For some participants, these bogus inter~ pretations (i.e., the same interpretation given to all subjects, regardless of the dream re~ ported, and with no reason to believe the interpretation applied to each subject) led to false memories for mildly traumatic suggested events. Dreams themselves may also be confused with actual experiences in some cases (Kemp, Burt, & Sheen, 2003). Perhaps the most widely used paradigm to study the effects of direct suggestion has become known as the "familial informant false~narrative procedure" (Lindsay, Hagen, Read, Wade, & Garry, 2004). Researchers first verify with family members that a child~ hood event did not take place, but later tell subjects that the event did take place, and that it was provided by the participant's family. Participants are asked to try to remember the events. Using this paradigm, researchers have led participants to develop such di~ verse false memories as having been lost in a mall for an extended time and later rescued by an elderly bystander, painful physical injuries, having been hospitalized overnight, having an accident at a family wedding, having been nearly drowned and rescued by a lifeguard, having been the victim of a vicious animal attack, hot air balloon rides, and many others (see Lindsay et al., 2004; Loftus, 2003, for reviews). Reviewing a set of such studies, Lindsay et al. (2004) found that on average 31% of participants develop false memories of the suggested event. The term rich false memories was coined to refer to instances in which the person feels confident, provides details, and even expresses strong emotions about made-up events that never happened (Loftus, 2003; Loftus & Bernstein, 2005). However, whereas some sorts of event memories appear to be easy to implant, other, less plausible events (having received a rectal enema) can be more difficult (Pezdek, Finger, & Hodge, 1997). Another frequently investigated method of implantation of false memories is the use of imagination or visual imagery, generally referred to as "imagination inflation" (Garry, Manning, Loftus, & Sherman, 1996). A host of studies have shown that active imagination/visualization of events, objects, or persons can lead to false memories of having actually seen, performed, or experienced them. These studies demonstrated that imagination/visual imagery can lead to false memories for common and bizarre personal behavior {both recent and historical), simple and complex events, pictures, sounds, written and spoken words/phrases, and inanimate and animate objects and persons, particularly

216

DAVIS AND LOFTUS

when a confederate also reports the imagined experience as real (e.g., Hoffman et al., 2001). Imagination has also induced false autobiographical beliefs and memories for a range of historical events (see reviews in Gerrie et al., 2005; Johnson, Hashtroudi, & Lindsay, 1993; Kassin, this volume; Loftus & Davis, 2006; Mazzoni & Memon, 2003; Schacter, 1996; 2001; Smith & Gleaves, this volume; Thomas & Loftus, 2002). Even paraphrasing event descriptions can produce inflation (e.g., Sharman, Garry, & Beuke, 2004). Although much of this work has primarily addressed the issue of autobiographical memory in an effort to understand the mechanisms underlying creation of false memories of sexual abuse, other work has examined the ability of imagination to create more mundane event predictions or memories. Imagining a future event, for example, has been shown to increase the subjective confidence that the event will actually occur (Carroll, 1978; Gregory, Cialdini, & Carpenter, 1982). Those who more easily imagined produced greater inflation in confidence (e.g., Sherman, Cialdini, Schwartzman, & Reynolds, 1985). Presumably, imagination and other implantation techniques work in a three-stage process whereby people first come to believe an event is plausible, next come to believe the event did actually occur, and finally reinterpret their narratives and images of the event as actual memories (Mazzoni, Loftus, & Kirsch, 2001; see the recent review by Henkel & Coffman, 2004). Imagery and imagination may contribute to all levels of this process. Imagery is crucial to plausibility and hence persuasion (see Green & Brock's [2002] review of tests of their "transportation-imagery model" of narrative persuasion, which provides evidence that narratives are persuasive to the extent they evoke imagery of their contents). Research from the source-monitoring tradition has shown that images can be confused with real memories, particularly when they have many of the subjective characteristics of real memories. Johnson, Foley, Suengas, and Raye (1988) found that there were almost no differences between rated characteristics of real and imagined memories for childhood events, although memories for real and imagined recent events differed in a number of respects. The authors reasoned that when real memories are vague and lacking in vivid detail, as when they are from the distant past, or were never encoded richly in the first place, it is easier to confuse imagined and real events. If true memories cannot be distinguished from false memories on the basis of vividness, richness of perceptual detail, and other contextual information, the two will seem equally real, and they will be easier to confuse. Thus, the imagination inflation research suggests that when testifying about events that were long past, were never encoded deeply, or are without rich imaging and contextual embedding, one will be more susceptible to memory distortion or creation through the use of imagination-based techniques This tendency can be quite dangerous in legal contexts. Memory recovery techniques emphasizing imagination (hypnosis, guided imagery, and age regression, for example) can lead to the creation of false memories in which the subject has great confidence (see Arburhnott, Arbuthnott, & Rossiter, 2001, regarding imagery; Mazzoni, Kirsch, & Lynn, this volume, on hypnosis; Spanos, Burgess, Burgess,Samuels, & Blois, 1999, on age regression), as illustrated in many cases of recovered memories (Loftus & Ketcham, 1994; Ofshe & Watters, 1994).

7.

SOURCES OF MISINFORMATION

217

Although false memories are often created in the context of therapeutic use of these techniques, law enforcement personnel sometimes attempt to induce suspects to engage in repeated acts of imagination of the criminal act as a means of inducing them to confess. In part, such imagination exercises were responsible for known cases of false confession, such Paul Ingram's confessions to abuse of his children (see Henkel & Coff, man, 2004; Kassin, this volume; Loftus & Ketcham, 1994). Another technique com, man to interrogation-explaining how a particular event might have happened-also tends to inflate confidence that the event actually did occur (Sharman, Manning, & Garry, 2005).

Props, Drawings, and Photographs When witnesses or victims encounter difficulty remembering persons or events, they may be asked to examine props, drawings, or photographs (or to revisit the scene of the event) in order to facilitate memory. Although it is clear that such physical context reinstatement procedures can indeed sometimes aid recall (e.g., Koutstaal, Schacter, & Johnson, 1999; Salmon & Pipe, 2000), it is equally clear that, like many memory re, trieval procedures, they can sometimes impair memory and/or trigger creation of false memories as well. For example, props or other aids used to remind the person of some target informa, tion may impair memory for noncued information (e.g., see Koutstall et al., 1999, for an illustration with use of photographs to facilitate event memory). As well, like forensic interviews, props, photos, and other retrieval aids may themselves incorporate misinfor, mation and thereby lead the interviewee to falsely remember the misinformation as true. Schacter and his colleagues (Schacter, Koutstall, Johnson, Gross, & Angell, 1997) dem, onstrated that (particularly older) persons who were shown misleading photos after having previously seen a videotape were later more likely to "remember" that events depicted in the photos had been in the original videotape. Photos can also bias memory for auditory information. Schacter (1996, p. 70) re, ported an experiment in which college students looked at photographs of people they had heard speak in either a pleasant or irritating tone of voice. Later, they were asked to recall the speaker's tone of voice. They were cued with photographs of the speaker pas, ing with either a slight smile or slight scowl. Regardless of the actual original tone of voice of the speaker, subjects who were cued with a smiling face were more likely to "remember" a pleasant tone of voice, and those cued with a scowling face were more likely to "remember" an irritating tone of voice. In other words, memory was greatly in, fluenced by the retrieval cue, drifting toward consistency with the cue. Unfortunately, autobiographical memory can be similarly distorted through expo, sure to photographs. Lindsay and his colleagues (Lindsay, Hagen, Read, Wade, & Garry, 2004) asked participants to work at remembering three childhood events (two true and one false, as established by their parents). Half of the participants were also given their school class photos from the years of the target events. The use of the photos signifi, candy increased the rate of false memories of the target pseudo,events, even though they did not directly depict or imply them.

218

DAVIS AND LOFTUS

SELF-GENERATED MISINFORMATION Updating and Reevaluating As noted earlier, memories and beliefs are continually revised in response to new information and beliefs. This process is reflected in previously reviewed phenomena such as hindsight and misinformation effects, and other schematic reconstructive memory processes. However, studies concerned with the retrospective bias have shown that reports of our own past attitudes or behaviors are biased by current attitudes or recently acquired information. Recollections of past political attitudes, for example, tend to be distorted significantly by current political beliefs (e.g., Dawes, 1991; Levine, 1997). Similarly, recollections of one's own behavior tend to change to conform to newly acquired information about how one should behave. That is, we reconstruct memory of the past so that we believe we behaved in a more consistent, sensible, or desirable way than we actually did. Generally, current self-views, goals, and beliefs bias memories of past selves and behavior (and vice versa; Wilson & Ross, 2003).

Mood and the Affect Infusion Model It is important to note that mood and other affective states can exert both direct and indirect effects on judgments. For example, the phenomenon called mood congruent retrieval has been shown to affect retrieval through contextual cuing processes (whereby mood serves as a biasing retrieval cue for mood consistent content), as well as motivated retrieval processes, or the desire to avoid mood incongruent memories (e.g., Bower, 1992; Mineka &Nugent, 1995). More generally, Forgas and East (2003) define "affect infusion" as "the process whereby affectively loaded information exerts an influence on, and becomes incorporated into, a person's cognitive and behavioral processes, entering into his or her constructive deliberations and eventually coloring the outcome in a mood-congruent direction." They provide evidence that affect can indirectly and implicitly influence judgments by affecting access to (i.e., memory of) and use of knowledge and incoming information-and can exert explicit influence when judges simply infer a response based on their current affective state (see the entire January 2002 issue of Psychological Inquiry, 13(1), devoted to Forgas's affect infusion model and related data). In effect, affective states can result in all the same biases in memory and judgment characteristic of schematic processing (see van den Bos, 2003, for specific applications to justice judgments).

Retrieving and Retelling A particularly pervasive source of misinformation is one's own repeated retrievals and recountings. Unfortunately, although failure to rehearse information tends to impair memory, rehearsal can itself impair or distort memory, especially when the rehearsal is selective or biased, or involves retrieval in a form different from encoding. Some neuroscientists have provided evidence they believe demonstrates that memories become un-

7.

SOURCES OF MISINFORMATION

219

stable during the act of retrieval (see also Estes' (1997) "perturbation" account of mem~ ory distortions), thereby requiring new consolidation when the memory is returned to storage. In turn, each reconsolidation can incorporate accurate information as well as misinformation acquired during retrieval and can delete information previously in mem~ ory (Nader, 2003; Nader, Schafe, & le Doux, 2000). Although the propositions of Nader and his colleagues are controversial, it is clear that retrieval is not benign. The phenomenon of "retrieval~induced forgetting," for example, refers to circum~ stances in which selective retrieval of some aspects of an event can lead to selective for~ getting of others (Barnier, Hung, & Conway, 2004). When retrieval occurs in a mode that is different from that in which the event was experienced, memory can be distorted. Witness attempts to verbally describe perpetrators have been shown to impair later at~ tempts to identify the perpetrator in a lineup-a phenomenon labeled "verbal overshad~ owing" (see Schooler & Miessner, this handbook, Vol. II, for a detailed discussion). This can also occur when retrieval requires the person to selectively extract information (e.g., How many times did ... ?) or to organize or judge the event in terms different from those that occurred when it was originally witnessed (see Davis & Friedman, this volume, and the next section on recountings). The wording one uses to describe some aspects of an event can affect memory for other features, such as duration (Pedersen & Wright, 2002). Furthermore, biased retrievals induced by forensic, therapeutic, or social interactions can induce selective memory and distortion. Even attempts to suppress memory can success~ fully impair it (Anderson et al., 2004; Anderson & Green, 2001).

Telling and Retelling: Distortion through Conversational Recountings Although attempts to retrieve information are often themselves sufficient to impair or distort memory, perhaps even greater problems occur as a result of conversational re~ countings. Even the simple act of repeating a statement can increase the strength of one's belief in its truth. Memory researchers have dubbed this effect the "illusory truth effect" (e.g., Arkes, Boehm, & Xu, 1991; Arkes, Hackett, & Boehm, 1989; Begg, Anas, & Farinacci, 1992). The effect occurs from the mere act of repetition, even when state~ ments are simply repeated in the laboratory, with no implication the statement is true, and out of context, without the personal involvement and very real consequences in~ herent in case~relevant testimony. Furthermore, even in the absence of any input from a conversational partner (such as alternative accounts, leading or misleading questions, feedback, therapeutic suggestions and procedures, etc.), persons asked to describe wit~ nessed materials or events (a) bias their descriptions to reflect their own goals (for ex~ ample, entertainment, deception, or informing accurately) or characteristics of the lis~ tener (for example, toward consistency with the listener's known attitudes or toward selective inclusion of aspects of interest to the listener), (b) subsequently report believ~ ing these biased descriptions, and (c) falsely recall the original stimuli as consistent with their biased descriptions (e.g., Higgins & Rholes, 1978; Higgins & McCann, 1984).

220

DAVIS AND LOFTUS

Pasupathi (2001) proposed two general principles governing conversational recounting of past events: co-construction (the joint influences of speakers and interpersonal contexts on conversational reconstructions) and consistency (the influence of initial conversational reconstructions on subsequent memory). Considerable evidence indicates that all recountings are shaped by (1) the mere expectation that there will be listeners; (2) the speaker's own goals with respect to the listener(s); (3) characteristics of listeners that affect expectations concerning their likely interest, understanding, or approval of specific content; and (4) listener behaviors and reactions during the interaction. In addition, aspects of the context may affect what is viewed as appropriate, possible, or desirable to discuss. Hence, reports may be adjusted with respect to the reported gist of the event, the nature of opinions expressed about the event, which of many aspects or specific details of an event are discussed, or the interpretation and evaluations of the events; and through additions and distortions designed to please, entertain, or protect oneself from negative reactions by the listener (see reviews by Clark, 1996; Davis & Friedman; Davis, Kemmelmeier, & Follette, 2005; Dudukovic, Marsh, & Tversky, 2004; Holtgraves, 2002, this volume; Hyman, 1994; Krauss & Chiu, 1998; Marsh & Tversky, 2004; Pasupathi, 2001; Sedikides, 1990; Todorov, 2002; Tversky & Marsh, 2000). Each such choice affects subsequent memory-both the amount recalled and the specific types of errors in recall (e.g., Tversky & Marsh, 2000). Such biased reports of witnessed events take place in a wide variety of social interactions, such as parent-child interactions, professional interviews with victims and suspects, therapeutic interactions, and the many conversations that might take place between co-witnesses or between a victim and supportive friends and family prior to reports to police. Indeed, the more informal conversations, which no doubt take place in greater numbers and variety, may shape accounts more powerfully than the professional interactions we have tended to study. Dudukovic et al. (2004) reviewed evidence indicating that strict accuracy is rarely the goal of informal interactions and that people widely report lying and otherwise distorting autobiographical and other information in everyday conversations. Such distorted accounts can exert long-term effects on memory. Indeed, accounts given to entertain (likely more typical of everyday conversations than those given with the strict goal of accuracy) are qualitatively different and result in poorer overall memory for the original information than those given with the goal of accurate recounting (e.g., Dudukovic et al., 2004; Wade & Clark, 1993). If the distorted recountings do elicit reinforcers, one would expect them to be repeated. Moreover, a variety of consistency and other motivations serve to maintain these reconstructed versions. Once these distorting conversations have taken place, subsequent memory is biased in the direction of the distorted reports-i.e., a "saying is believing" effect (see review in Holtgraves, 2002). Pasupathi argued that conversational recountings influence subsequent memory because they "function as a rehearsal that is selective, is schematic, and can lead to source memory confusions," (p. 658). Dudukovic et al. (2004) also suggest that retellings affect memory via creation of schemas that are later used to guide reconstruction of events: "Schemas are used during retelling to re-organize events, to select

7.

SOURCES OF MISINFORMATION

221

some details and omit others. The schema provides a top-down structure for events, and thus allows for elaborations in the retelling to bridge across and interpret events" (p. 127). Over time, a witness who discusses witnessed events with others can be expected to provide biased reports to those others and ultimately to come to believe the distorted accounts and report these false recollections to still others. Even the anticipation of recounting to a specific audience (e.g., children versus adults) tends to shape memory for the narrative to be reported (e.g., Adams, Smith, Pasupathi, & Vitolo, 2002; Schuurmans & Vandierendonck, 1985). And, given the many and varied conversations and interviews any given witness can be expected to experience, multiple conversation-related distortions can be expected to accumulate. Dudukovic et al. (2004) also noted that the difference in organizational structures during retellings (such as informal conversations involving witnesses) and the target recall (such as during a forensic interview or trial testimony) can impair accuracy. When the story must suddenly be restructured at recall, retrieval pathways do not match the way the information is stored. Indeed, Dudukovic et al. found that when storytellers had to switch perspectives at recall, the new chronological structure aided recall of information not previously recounted in accounts given to entertain, whereas it impaired memory for information that was previously recounted (and hence tightly tied to the story schema), presumably because the organization of that information during the previous storytelling could no longer be used. Adaval and Wyer (2004) examined the impact of making intervening judgments and reports on later memory for a conversation. Participants witnessed a conversation with the sole objective of comprehension. Subsequently, they either wrote their impressions of the characters involved or described the sequence of interaction, and later completed recognition tasks for aspects of the original conversation. Communicating impressions of the protagonists decreased recognition of verbal aspects of the conversation, but not nonverbal behaviors; whereas reporting on the sequence of events decreased recognition of both verbal and nonverbal behaviors. The authors likened their findings to the "verbal overshadowing" (e.g., Schooler & Engstler-Schooler, 1990) findings in the eyewitness identification literature (see also Krauss & Chiu, 1998, for a similar interpretation). Consistent with the "bin" model of social memory (Wyer & Srull, 1989), the authors described the processes promoting distortion as follows: At the time we witness the interaction, our only objective may be to understand what the protagonists said and did. Later, however, we may be called upon to use the information we have acquired for a reason we did not anticipate .... To attain this objective, we presumably retrieve a mental representation of the interaction that we had formed at the time we first observed it and construe its implications for the communication we intend to deliver. In the process of generating the communication, however, we are likely to construct a new representation of the experience whose content and implications differ in several respects from the one we had formed earlier. ... Once this new representation is formed, it can be used as a basis for reconstructing the initial events we had observed. Our reliance on the new representation can sometimes distort our memory for the original events and in some cases, can lead to memory errors. (Adaval & Wyer, 2004, p. 450)

222

DAVIS AND LOFTUS

Deliberate Fabrication: Lying and Feigning Amnesia Recent research has begun to investigate the role of providing deliberately false accounts of the provider's own subsequent memory. Theoretically, deliberate fabrications-either through knowingly providing false information or through feigning failure to remember-might affect veridical memory through several distinct theoretical processes (e.g., Christianson & Bylin, 1999). First, either focusing on false information or diverting attention from true information in order to feign amnesia can interfere with rehearsal of true information. Second, as suggested by the previously reviewed research on "retrievalinduced forgetting," selective retrieval of some information might selectively impair memory for the nonreported true information that had been suppressed in order to lie or feign amnesia (e.g., Anderson, Bjork, & Bjork, 1994; Christianson & Bylin, 1999; Ciranni &Shimamura, 1999; Macrae &MacLeod, 1999; Shaw, Bjork, &Handal, 1995; Wright, Loftus, & Hall, 2001). Third, the person may have constructed new information or distorted aspects of witnessed information in order to effectively lie or feign amnesia. Particularly if one developed elaborated scenarios or visualizations of the new or distorted version of events, one might later experience a source-misattribution of the new version to the original event instead of one's own constructions (e.g., Johnson, Hashtroudi, & Lindsay, 1993). Finally, although the first three processes would tend to impair veridical memory, a fourth, suggested by Polage (2004), would tend to facilitate veridical memory. That is, in order to lie convincingly, the person may have to remember (and hence rehearse) the true information in order to ensure that false information is reported. In addition, this retrieval would occur under conditions involving a clear tag indicating that the new information is false. Indeed, there is evidence that each of these processes occurs under some circumstances. Polage (2004) tested the fourth process, using the familiar imagination inflation method (Garry et al., 1996). However, instead of being asked to imagine a particular event, participants were asked to lie, falsely claiming that some of the events actually happened to them. The authors found that the most common effect of this procedure was to deflate ratings (given 1 week later) of the likelihood that the falsely reported event had actually occurred. However, a minority of participants (10-16%) in two studies increased their certainty to the maximum rating. Hence, while lying tended to result in "fabrication deflation" (Polage, 2004) for most subjects, some developed great confidence that the lied-about event had actually occurred. Several studies (Bylin & Christianson, 2002; Christianson & Bylin, 1999; Oorsouw & Merckelbach, 2004) have found evidence of memory impairment among those asked to feign amnesia. These have differed in important methodological details from the Pollage studies, however. Participants in the Bylin and Christianson (2002) and Christianson and Bylin (1999) studies read crime narratives while imagining themselves in the role of perpetrator. They were then asked either to report the narrative veridically or to report it as if trying to evade responsibility for the crime, including by feigning memory loss. Oorsouw and Merckelback (2004) had participants actually engage in a mock crime and then offer veridical or simulations of amnestic reports. In all three studies, participants who reported falsely to evade responsibility and/or simulate memory failure were

7.

SOURCES OF MISINFORMATION

223

less accurate when subsequently asked to report the truth, either immediately or 1 week later. Furthermore, the simulators were equally confident in their memories as previously untested truth-telling controls (Oorsouw & Merkelbach, 2004). Unlike Polage's (2004) participants, who were specifically asked to lie about whether a specific event had taken place, participants in these studies were allowed to fabricate, distort, or feign amnesia for a variety of details to accomplish their deceptive goals. Hence, they may have constructed more elaborate general scenarios, distorted true events, and added false events. These processes may have facilitated source monitoring errors and provided less clear tags for which things were true and which false than Polage's procedure oflying about specific events that did not happen. Pickel (2004) tested the effects of fabrication in witnesses to a robbery. Witnesses correctly described the robber, fabricated a description of him or of someone else, or did not describe anyone. One week later, those who had previously fabricated a description of the robber and those who did not previously describe him remembered fewer correct details than did truthful witnesses or those who fabricated about another person. Those who fabricated about the robber also reported more incorrect details than truthful or noninterviewed witnesses. Evidence that deliberate fabrications can impair veridical memory is of considerable practical significance, given the prevalence of both lying and feigning amnesia among criminal defendants. Of particular interest is the fact that police interrogators are trained to elicit false accounts from offenders (for example, admitting to inaccurate facts or scenarios that would render their crimes less serious) as a stepping-stone approach to eventually getting the full truth (e.g., lnbau, Reid, Buckley, &Jayne, 2001; see Davis & O'Donohue, 2004, for a review). Police are generally insensitive to the potential of their procedures to elicit false accounts and even less aware of the potential of their procedures to impair memory such that the truth is lost altogether.

Personal Routines and Repeated Events Repetitive events become easily confused, such that one can lose track of which specific occurrences took place in which event. In line with the tendency for source similarity to increase source confusion (see Wyer, 2004, for a review of effects of similarity between events on event-specific memory), the more additional similar events intervene between the target event and attempt to recall, the greater the likelihood of confusion between them or failure to remember the target at all. Hence, schematic inferences concerning what one was likely to have done, or what was likely to have occurred in the instance in question, tend to replace veridical memory (see review in Haber & Haber, 2000). This is of particular concern for parties to cases involving routinized professional interactions, such as doctor-patient, attorney-client, and others. CONCLUSIONS

While often witness memory is not only valuable, but crucial, it must be appreciated that even the most scrupulously honest and sincere witness cannot be presumed to be entirely accurate. Human information processors are simply overwhelmed by the magnitude of

224

DAVIS AND LOFTUS

incoming information and efforts to process and remember it accurately, leaving us vulnerable to error. Some information is necessarily lost, and other information may be confused or distorted. Misinformation is introduced through countless internal and external mechanisms. We have shown that misinformation can create errors in memory for events that were actually witnessed, and that it can also lead to the planting of entirely false beliefs or memories. It is crucial for the various participants in the legal system to gain greater understanding of the magnitude of the potential for distortion in witness memory and of the many and varied sources of memory distortion for events relevant to a wide variety of litigated events. Unfortunately, expert testimony has been largely restricted to criminal trials and issues of eyewitness identification, child witnesses, and recovered memories (which is also offered in civil trials). This represents only the tip of the iceberg of memory issues with potential relevance to the courts. Davis & Friedman (this volume), for example, point to pervasive testimony involving memory for statements in both civil and criminal trials. Yet memory experts have yet to penetrate barriers to introduction of expert testimony on conversational memory. They have come a long way with their contributions to the legal system. Still, given the varied sources of distortion and the many events for which witness memory is an issue, we can only expect the range of testimony from witness memory researchers to expand in the future.

REFERENCES Abbot, W. F., & Batt, J. (1999). A handbook of jury research. Philadelphia: ALA-ABA. Adams, C., Smith, M. C., Pasupathi, M., & Vitolo, L. (2002). Social context effects on story recall in older and younger women: Does the listener make a difference? journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 57B, P28-P40. Allwood, C. M., Granhag, P. A., & Johansson, M. (2003). Increased realism in eyewitness confidence judgements: The effect of dyadic collaboration. Applied Cognitive Psychology, 17, 545-561. Anderson, M. C., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Psychonomic Bulletin and Review, 20, 1063-1087. Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by executive control. Nature, 410, 366-369. Anderson, M. C., Ochsner, K. N., Kuhl, B., Cooper,]., Robertson, S. W., Gabrieli, G. H., et al. (2004). Neural systems underlying the suppression of unwanted memories. Science, 303, 232-235. Anderson, R. C., & Prichert, ]. W. (1978). Recall of previously unrecallable information following a shift in perspective. Journal of Verbal Learning and Verbal Behavior, 17, 1-12. Araya, T., Ekehammar, B., & Akrami, N. (2003). Remembering things that never occurred: The effects of to-be-forgotten stereotypical information. Experimental Psychology, 50, 27-32. Arbuthnott, K. D., Arbuthnott, D. W., & Rossiter, L. (2001). Guided imagery and memory: Implications for psychotherapists. Journal of Counseling Psychology, 48, 123-132. Arkes, H. R., Boehm, L. E., & Xu, G. (1991). Determinants of judged validity. journal of Experimental Social Psychology, 27, 576-605. Arkes, H. R., Hackett, C., & Boehm, L. (1989). The generality of the relation between familiarity and judged validity. journal of Behavioral Decision Making, 2, 81-94. Asch, S. E. (1955, November). Opinions and social pressure. Scientific American, 31-35. Assefi, S. L., & Garry, M. (2002). Absolut memory distortions: Alcohol placebos influence the misinformation effect. Psychological Science, 14, 77-80. Ayers, M.S., & Reder, L. M. (1998). A theoretical review of the misinformation effect: Predictions from an activation-based memory model. Psychonomic Bulletin and Review, 5, 1-21.

7.

SOURCES OF MISINFORMATION

225

Bargh, J. A., Raymond, P., & Pryor, J. B. (1995). Attractiveness of the underling: An automatic power ~ sex association and its consequences for sexual harassment and aggression. Journal

of Personality and Social Psychology, 68, 768-781.

Barnier, A. J., Hung, L., & Conway, M.A. (2004). Retrieval-induced forgetting of emotional and unemotional autobiographical memories. Cognition & Emotion, 18, 457-477. Bayen, U. J., Nakamura, G. V., Dupuis, S. E., & Yang, C. (2000). The use of schematic knowledge about sources in source monitoring. Memory and Cognition, 28, 480-500. Begg, I. M., Anas, A., & Farinacci, S. (1992). Dissociation of processes in belief: Source recollection, statement familiarity, and the illusion of truth. Journal of Experimental Psychology:

General, 121,446-458.

Belezza, F. S., & Bower, G. H. (1981). Person stereotypes and memory for people. Journal of Per-

sonality and Social Psychology, 41, 856-865. Betz, S., Skowronski, J. J., & Ostrom, T. (1996). Shared realities: Social influence and stimulus memory. Social Cognition, 14, 113-140. Blank, H. (1998). Memory states and memory tasks: An integrative framework for eyewitness memory and suggestibilty. Memory, 6, 481-529. Bower, G. H. (1992). How might emotions affect learning? InS. Christianson (Ed.), The handbook of emotion and memory: Research and theory (pp. 3-31). Hillsdale, NJ: Lawrence Erlbaum Associates. Bower, G. H., Black, J. B., & Turner, T. J. (1979). Scripts in memory for text. Cognitive Psychol-

ogy, 11, 177-220.

Brainerd, C. J., & Reyna, V. F. (2002). Fuzzy-trace theory and false memory. Current Directions in

Psychological Science, 11, 164-169.

Bransford, J.D., &Johnson, M. K. (1972). Contextual prerequisites for understanding: Some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behav-

ior, 11, 717-726.

Braun, K. A. (1999). Postexperience advertising effects on consumer memory. Journal of Con-

sumer Research, 25,319-334.

Braun, K. A., Ellis, R., & Loftus, E. F. (2002). Make my memory: How advertising can change our memories of the past. Psychology & Marketing, 19, 1-23. Braun, K. A., & Loftus, E. F. (1998). Advertising's misinformation effect. Applied Cognitive Psy-

chology, 12, 569-591.

Bregman, N.J., & McAllister, H. A. (1982). Eyewitness testimony: The role of commitment in increasing reliability. Social Psychology Quarterly, 45, 466-472. Brewer, W. F., & Treyens, J. C. (1981). Role of schemata in memory for places. Cognitive Psy-

chology, 13, 207-230.

Bruce, V., Ness, H., Hancock, P. J. B., Newman, C., & Rarity, J. (2002). Four heads are better than one: Combining face composites yields improvements in face likeness. Journal of Applied

Psychology, 87, 894-902.

Bruck, M., Ceci, S. J., & Hembrooke, H. (1998). Reliability and credibility of young children's reports: From research to policy and practice. American Psychologist, 53, 136-151. Bylin, S., & Christianson, S.-A. (2002). Characteristics of malingered amnesia: Consequences of withholding vs. distorting information on later memory of a crime event. Legal and Crimino-

logical Psychology, 7, 45-61.

Carli, L. L. (1999). Cognitive reconstruction, hindsight, and reactions to victims and perpetrators. Personality and Social Psychology Bulletin, 25, 966-979. Carroll, J. M., & Russell, J. A. (1996). Do facial expressions signal specific emotions? Judging emotion from face in context. Journal of Personality and Social Psychology, 70, 205-218. Carroll, J. S. (1978). The effect of imagining an event on expectations for the event: An interpretation in terms of the availability heuristic. Journal of Personality and Social Psychology, 36,

1501-1511.

Ceci, S. J., Ross, D. F., & Toglia, M.P. (1987). Suggestibility of children's memory: Psycholegal implications. Journal of Experimental Psychology: Applied, 116, 38-49.

226

DAVIS AND LOFTUS

Chambers, K. L., & Zaragoza, M.S. (2001). Intended and unintended effects of explicit warnings on eyewitness suggestibility: Evidence from source identification tests. Memory & Cognition, 29, 1120-1129. Christensen, T. C., Wood, J. V., & Barrett, L. F. (2003). Remembering everyday experience through the prism of self-esteem. Personality and Social Psychology Bulletin, 29, 51-62. Christensen-Szalanski, J. J. J., & Willham, C. F. (1991). The hindsight bias: A meta-analysis. Organizational Behavior and Human Decision Processes, 48, 147-168. Christianson, S.-A., & Bylin, S. (1999). Does simulating amnesia mediate genuine forgetting for a crime event? Applied Cognitive Psychology, 13, 495-511. Ciranni, M. A., & Shimamura, A. P. (1999). Retrieval induced forgetting in episodic memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25, 1403-1414. Clark, H. H. (1996). Using language. Cambridge: Cambridge University Press. Cohen, C. E. (1981). Person categories and social perception: Testing some boundaries of the processing effect of prior knowledge. Journal of Personality and Social Psychology, 40, 441-452. Conway, L. G. (2004). Social contagion of time perception. Journal of Experimental Social Psychology, 40, 113-120. Copeland, J., & Snyder, M. (1995). When counselors confirm: A functional analysis. Personality and Social Psychology Bulletin, 21, 1210-1221. Corneille, 0., Huart, J., Becquart, E., & Bredart, S. (2004). When memory shifts toward more typical category exemplars: Accentuation effects in the recollection of ethnically ambiguous faces. Journal of Personality and Social Psychology, 86, 236-250. Correll, J., Park, B., Judd, C. M., & Wittenbrink, B. (2002). The police officer's dilemma: Using ethnicity to disambiguate potentially threatening individuals. Journal of Personality and Social Psychology, 83, 1314-1329. Cowley, E., &Janus, E. (2004). Not necessarily better, but certainly different: A limit to the advertising misinformation effect on memory. Journal of Consumer Research, 31, 229-235. Crombag, H. F. M., Wagenaar, W. A., & Van Koppen, P. J. (1996). Crashing memories and the problem of source monitoring. Applied Cognitive Psychology, 10, 95-104. Davis, D. (1991). What would I think ifl didn't know the patient died? In D. Davis (Ed.), From the mind's eye (Vol. 1), pp. 1-4. Davis, D., & Follette, W. C. (2001). Foibles of witness memory for traumatic/high profile events. Journal of Air Law and Commerce, 66, 1421-1549. Davis, D., & Follette, W. C. (2003). Toward an empirical approach to evidentiary ruling. Law and Human Behavior, 27, 661-684. Davis, D., & Follette, W. C. (2004). Jurors CAN be selected: Non information, misinformation and their strategic use for jury selection. In W. T. O'Donohue & E. Levensky (Eds.), Handbook of forensic psychology (pp. 782-805). New York: Elsevier Academic Press. Davis, D., Kemmelmeier, M., & Follette, W. C. (2005). Conversational memory on trial. In Y. I. Noy & W. Karwowski (Eds.), Handbook of human factors in litigation (pp. 12-11 to 12-29). New York: CRC Press. Davis, D., & Loftus, E. F. (2005). Age and functioning in the legal system: Victims, witnesses and jurors. In Y. I. Noy & W. Karwowski (Eds.), Handbook of human factors in litigation (pp. 1111 to 11-53). New York: CRC Press. Davis, D., Loftus, E. F., & Follette, W. C. (2001). How, when, and whether to use informed consent for recovered memory therapy. Journal of the American Academy of Psychiatry and the Law, 29, 148-159. Davis, D., Lopez, P., Koyama, M., White, B., Knaack, D., Kusal, T., et al. (2005). Memory for threats in conversation enhanced by later knowledge of violence between participants. Paper presented at the American Psychological Society, Los Angeles, CA. Davis, D., & O'Donohue, W. T. (2004). The road to perdition: Extreme influence tactics in the interrogation room. In W. T. O'Donohue & E. Levensky (Eds.), Handbook of forensic psychology (pp. 897-996). New York: Elsevier, Academic Press.

7.

SOURCES OF MISINFORMATION

227

Davis, D., Vanous, S., & Cucciare, M. (2002). "Unconscious transference" as an instance of change blindness. Unpublished manuscript, Reno, Nevada. Davis, D., Vanous, S., & Cucciare, M. (2005). Unconscious transference as an instance of "change blindness." Paper presented at the American Psychological Society, Los Angeles, CA. http:// www.sierratrialandopinion.com/papers/ChangeBlindnessPaper.rtf. Dawes, R. M. (1991). Biases of retrospection. Issues in Child Abuse Accusations, 1, 25-28. Dobson, M., & Markham, R. (1993). Imagery ability and source monitoring: Implications for eyewitness memory. British]ournal of Psychology, 32, 11-118. Dodd, D. H., & Bradshaw, J. M. (1980). Leading questions and memory: Pragmatic constraints. Journal of Verbal Learning and Verbal Behavior, 19, 695-704. Dodson, C. S., &Johnson, M. K. (1993). Rate offalse source attributions depends on how questions are asked. American]ournal of Psychology, 106, 541-557. Drivdahl, S. B., & Zaragoza, M.S. (2001). The role of perceptual elaboration and individual differences in the creation of false memories for suggested events. Applied Cognitive Psychology, 15, 265-281. Dudukovic, N. M., Marsh, E. J., & Tversky, B. (2004). Telling a story or telling it straight: The effects of entertaining versus accurate retellings on memory. Applied Cognitive Psychology, 18, 125-143. Dunning, D., & Sherman, D. A (1997). Stereotypes and tacit inference. journal of Personality and Social Psychology, 73,459-471. Eakin, D. K., Schreiber, T. A., & Sergent-Marshall, S. (2003). Misinformation effects in eyewitness memory: The presence and absence of memory impairment as a function of warning and misinformation accessibility. journal of Experimental Psychology: Learning, Memory, and Cognition, 29,813-825. Eberhardt, J. L., Dasgupta, N., & Banaszynski, T. L. (2003). Believing is seeing: The effect of racial labels and implicit beliefs on face perception. Personality and Social Psychology Bulletin, 29, 360-370. Eberhardt, J. L., Goff, P. A., Purdie, V. J., & Davies, P. G. (2004). Seeing black: Race, crime, and visual processing. Journal of Personality and Social Psychology, 87, 876-893. Eisen, M. L., & Carlson, E. B. (1998). Individual differences in suggestibility: Examining the influence of dissociation, absorption, and a history of childhood abuse. Applied Cognitive Psychology, 12, S4 7-S61. Eisen, M. L., & Lynn, S. J. (2001). Dissociation, memory and suggestibility in adults and children. Applied Cognitive Psychology, 13, S49-S73. Eisen, M. L., Winograd, E., & Qin, J. (2002). Individual differences in adults' suggestibility and memory performance. In M. L. Eisen, J. AQuas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 205-233). Mahwah, NJ: Lawrence Erlbaum Associates. Erdfelder, E., & Buchner, A (1998). Decomposing the hindsight bias: A multinomial processing tree model for separating recollection and reconstruction in hindsight. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24,387-414. Estes, W. K. (1997). Processes of memory loss, recovery and distortion. Psychological Review, 104, 148-169. Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology, 69, 1013-1027. Fishhoff, B. (1975). Hindsight does not equal foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1, 288-299. Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York: McGraw-Hill. Fivush, R., & Haden, C. A (Eds.). (2003). Autobiographical memory and the construction of a narrative self. Mahwah, NJ: Lawrence Erlbaum Associates.

228

DAVIS AND LOFTUS

Forgas, 1. P., & East, R. (2003). Affective influences on social judgments and decisions: Implicit and explicit processes. In J.P. Forgas & K. D. Williams (Eds.), Social judgments: Implicit and explicit processes (pp. 198-226). New York: Cambridge University Press. Fragale, A. R., & Heath, C. (2004). Evolving informational credentials: The (mis)attribution of believable facts to credible sources. Personality and Social Psycholog;y Bulletin, 30, 225236. Frank, M. G., & Gilovich, T. (1988). The dark side of self- and social perception: Black uniforms and aggression in professional sports. Journal of Personality and Social Psycholog;y, 37, 1532-1538. Frost, P. (2000). The quality of false memoty over time: Is memory for misinformation "remembered" or "known"? Psychonomic Bulletin and Review, 7, 531-536. Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can eyewitnesses influence each other's memories for an event? Applied Cognitive Psycholog;y, 17, 533-543. Garcia-Bajos, E., & Migueles, M. (2003). False memories for script actions in a mugging account. European Journal of Cognitive Psycholog;y, 15, 195-208. Gardner, W. L., Pickett, C. L., & Brewer, M. B. (2000). Social exclusion and selective memory: How the need to belong influences memory for social events. Personality and Social Psycholog;y Bulletin, 26, 486--496. Garry, M., & Harper, D. N. (2005). Pigeons, rats and humans show analogous misinformation effects. Garry, M., Manning, C. G., & Loftus, E. F. (1996). Imagination inflation: Imagining a childhood event inflates confidence that it occurred. Psychonomic Bulletin & Review, 3, 208214. Garven, S., Wood, 1. M., & Malpass, R. S. (2000). Allegations of wrongdoing: The effects of reinforcement on children's mundane and fantastic claims. Journal of Applied Psycholog;y, 85,38-49. Gawronski, B., Ehrenberg, K., Banse, R., Zuhova, 1., & Klauer, K. C. (2003). It's in the mind of the beholder: The impact of stereotypic associations on category-based and individuating impression formation. Journal of Experimental Social Psycholog;y, 39, 16-30. Gerrie, M., Garry, M., & Loftus, E. F. (2005). False memories. InN. Brewer & K. Williams (Eds.), Psycholog;y and law: An empirical perspective (pp. 222-253). New York: Guilford. Goggin, W. C., & Range, L. M. (1985). The disadvantages of hindsight in the perception of suicide. Journal of Social and Clinical Psycholog;y, 3, 232-237. Graesser, A. C., Gordon, S. E., & Sawyer, 1. D. (1979). Recognition memory for typical and atypical actions in scripted activities: Tests of a script pointer + tag hypothesis. Journal of Verbal Learning and Verbal Behavior, 18,319-332. Graesser, A. C., & Nakamura, G. V. (1982). The impact of schemas on comprehension and memory. In G. H. Bower (Ed.), The psycholog;y of learning and motivation (Vol. 16, pp. 59109). New York: Academic Press. Graesser, A. C., Olde, B. R., & Klettke, B. (2002). How does the mind construct and represent stories? In M. C. Green, 1. 1. Strange & T. C. Brock (Eds.), Narrative impact: Social and cognitive foundations (pp. 229-262). Mahwah, N1: Lawrence ErlbaumAssociates. Green, M. C., & Brock, T. C. (2002). In the mind's eye: Transportation-imagery model of narrative persuasion. In M. C. Green, 1· 1. Strange, & T. C. Brock (Eds.), Narrative impact: Social and cognitive foundations (pp. 315-341). Mahwah, N1: Lawrence Erlbaum Associates. Green, M. C., Strange, 1. 1., & Brock, T. C. (Eds.). (2002). Narrative impact: Social and cognitive foundations. Mahwah, N1: Lawrence Erlbaum Associates. Greenberg, M.S., Westcott, D. R., & Bailey, S. E. (1998). When believing is seeing: The effect of scripts on eyewitness memory. Law and Human Behavior, 22, 685-694. Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of personal history. American Psychologist, 35, 603-618.

7.

SOURCES OF MISINFORMATION

229

Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psyclwlogy, 74, 1464-1480. Greenwald, A. G., Oakes, M.A., & Hoffman, H. G. (2003). Targets of discrimination: Effects of race on responses to weapons holders. Journal of Experimental Social Psyclwlogy, 39, 399-405. Gregory, W. L., Cialdini, R. B., & Carpenter, K. M. (1982). Self-relevant scenarios as mediators of likelihood estimates and compliance: Does imagining make it so? Journal of Personality and Social Psychology, 88-99. Gudjonsson, G. H. (2003). Psychology brings justice: The science of forensic psychology. Criminal Behaviour and Mental Health, 13, 159-167. Haber, R.N., & Haber, L. (2000). Experiencing, remembering and reporting events. Psychology, Public Policy, and Law, 6, 1057-1097. Hafstad, G. S., Memon, A., & Logie, R. (2004). Post-identification feedback, confidence and recollections of witnessing conditions in child witnesses. Applied Cognitive Psyclwlogy, 18, 901-912. Halberstadt, J. (2003). The paradox of emotion attribution: Explanation biases perceptual memory for emotional expressions. Current Directions in Psychological Science, 12, 197-200. Halberstadt, J. (2005). Featural shift in explanation-biased memory for emotional faces. Journal of Personality and Social Psyclwlogy, 88, 38-49. Halberstadt, J. B., & Niedenthal, P.M. (2001). Effects of emotion concepts on perceptual memory for emotional expressions. Journal of Personality and Social Psychology, 81, 587-598. Hannigan, S. L., & Reinitz, M. T. (2001). A demonstration and comparison of two types of inference based memory errors. Journal of Experimental Social Psychology: Learning, Memory, and Cognition, 27, 931-940. Hannigan, S. L., & Reinitz, M. T. (2003). Migration of objects and inferences across episodes. Memory and Cognition, 31,434-444. Harper, D. N., & Garry, M. (2000). Postevent cues bias recognition performance in pigeons. Animal Learning and Behavior, 28, 59-67. Hastie, R. (1980). Memory for behavioral information that confirms or contradicts a personality impression. In R. Hastie, T. M. Ostrom, E. Ebbesen, R. S. Wyer, Jr., D. Hamilton & D. Carlston (Eds.), Person memory: The cognitive basis of social perception (pp. 141-172). Hillsdale, NJ: Lawrence Erlbaum Associates. Hastie, R. (1981). Schematic principles in human memory. In E. T. Higgins, C. P. Herman & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 39-88). Hillsdale, NJ: Lawrence Erlbaum Associates. Hastie, R., & Pennington, N. (1995). The big picture: Is it a story? In R. S. Wyer (Ed.), Advances in social cognition (Vol. 8, pp. 133-138). Mahwah, NJ: Lawrence Erlbaum Associates. Hastie, R., & Pennington, N. (2000). Explanation based decision making. InT. Connolly & H. R. Arkes (Eds.),Judgment and decision making: An interdisciplinary reader (2nd ed., pp. 212228). New York: Cambridge University Press. Hawkins, S. A., & Hastie, R. (1990). Hindsight: Biased judgments of past events after the outcomes are known. Psychological Bulletin, 107, 311-327. Hekkanen, S. T., & McEvoy, C. (2002). False memories and source-monitoring problems: Criterion differences. Applied Cognitive Psychology, 16, 73-85. Henkel, L. A., & Coffman, K. J. (2004). Memory distortions in coerced false confessions: A source monitoring framework analysis. Applied Cognitive Psychology, 18, 567-588. Hicks, J. L., & Cockman, D. W. (2003). The effect of general knowledge on source memory and decision processes. Journal of Memory and Language, 48, 489-501. Higgins, E. T., & McCann, C. D. (1984). Social encoding and subsequent attitudes, impresions and memory. Journal of Personality and Social Psychology, 47, 26-39. Higgins, E. T., & Rholes, W. S. (1978). Saying is believing: Effects of message modification on memory and liking for the person described. Journal of Experimental Social Psychology, 14, 141-154.

DAVIS AND LOFTUS

230

Highhouse, S., & Bottrill, K. V. (1995). The influence of social (mis)information on memory for behavior in an employment interview. Organizational Behavior and Human Decision Processes, 62, 200-229. Hirt, E., McDonald, H. E., & Markman, K. D. (1998). Expectancy effects in reconstructive memory: When the past is just what we expected. InS. Lynn & K. M. McConkey (Eds.), Truth and memory (pp. 62-89). New York: Guilford. Hodgins, H. S., & Zuckerman, M. (1993). Beyond selecting information: Biases in spontaneous questions and resultant conclusions. Journal of Experimental Social Psychology, 29, 387-407. Hoffman, D. D. (1998). Visual intelligence. New York: W. W. Norton & Co. Hoffman, H. G., Granhag, P. A., See, S. T. K., & Loftus, E. F. (2001). Social influences on reality monitoring decisions. Memory & Cognition, 29, 394-404. Hoffrage, U., & Pohl, R. F. (Eds.). (2003). Special issue of the journal memory: Hindsight bias. Hove, UK: Taylor & Francis. Holst, B. F., & Pezdek, K. (1992). Scripts for typical crimes and their effects on memory for eyewitness testimony. Applied Cognitive Psychology, 6, 573-587. Holtgraves, T. M. (2002). Language as social action. Mahwah, NJ: Lawrence Erlbaum Associates. Horgan, T. G., Mast, M.S., Hall, J. A., & Carter, J.D. (2004). Gender differences in memory for the appearance of others. Personality and Social Psychology Bulletin, 30, 185-196. Hubbard, T. L. (1996). Representational momentum, centripetal force, and curvilinear impetus. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22, 1049-1060. Hunt, J. S., & Borgida, E. (2001). Is that what I said?: Witnesses' responses to interviewer modifications. Law and Human Behavior, 25, 583-603. Hyman, I. E. (1994). Conversation remembering: Story recall with a peer versus for an experimenter. Applied Cognitive Psychology, 8, 49-66. Hyman, I. E., & Pentland, J. (1996). The role of mental imagery in the creation offalse childhood memories. Journal of Memory and Language, 35, 101-117. Inbau, F. E., Reid, J. E., Buckley, J.P., & Jayne, B. C. (2001). Criminal interrogations and confessions (4th ed.). Gaithersburg, MD: Aspen. Ito, T. A., & Urland, G. R. (2003). Race and gender on the brain: Electrocortical measures of attention to the race and gender of multiply categorizable individuals. Journal of Personality and Social Psychology, 85, 616-626. Johnson, M. K., Foley, M.A., Suengas, A. G., & Raye, C. L. (1988). Phenomenal characteristics of memories for perceived and imagined autobiographical events. Journal of Experimental Psychology: General, 117, 371-376. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin,

114,3-28.

Judd, C. M., Blair, I. V., & Chapleau, K. M. (2004). Automatic stereotypes vs. automatic prejudice: Sorting out the possibilities in the Payne (2001) weapon paradigm. Journal of Experimental Social Psychology, 40, 75-81. Kawakami, K., Dion, K. L., & Dovidio, J. F. (1998). Racial prejudice and stereotype activation. Personality and Social Psychology Bulletin, 24,407-416. Kemp, S., Burt, C. D. B., & Sheen, M. (2003). Remembering dreamt and actual experiences. Applied Cognitive Psychology, 17, 577-591. Keuler, D. J., & Safer, M.A. (1998). Memory bias in the assessment and recall of pre-exam anxiety: How anxious was I? Applied Cognitive Psychology, 12, 127-137. Klauer, K. C., & Wegener, I. (1998). Unraveling social categorization in the "Who said what?" paradigm. Journal of Personality and Social Psychology, 73, 1155-1178. Klein, K., & Boals, A. (2001). The relationship of life event stress and working memory capacity. Applied Cognitive Psychology, 15, 565-579. Koutstaal, W., Schacter, D. L., Johnson, M. K., & Galluccio, L. (1999). Facilitation and impairment of event memory produced by photograhp review. Memory & Cognition, 27,478-493.

7.

SOURCES OF MISINFORMATION

231

Krauss, R. M., & Chiu, C. Y. (1998). Language and social behavior. In D. T. Gilbert, S. T. Fiske & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 2, pp. 41-88). New York: McGraw-Hill. Kunda, Z. (1999). Social cognition: Making sense of people. Cambridge, MA: MIT Press. Lampinen, J. M., Copeland, S.M., & Neuschatz, J. S. (2001). Recollections of things schematic: Room schemas revisited. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 1211-1222. Lassiter, G. D., Stone, J. I., & Weigold, M. F. (1988). Effect of leading questions on the selfmonitoring-memory correlation. Personality and Social Psychology BUlletin, 13, 537-545. Lee, K. (2004). Age, neuropsychological, and social cognitive measures as predictors of individual differences in susceptibility to the misinformation effect. Applied Cognitive Psychology, 18, 997-1019. Levine, L. J. (1997). Reconstructing memory for emotions. Journal of Experimental Psychology: General, 126, 165-177. Liebmann, J. I., McKinley-Pace, M. J., Leonard, A.M., Sheesley, L.A., Gallant, C. L., Renkey, M. E., et al. (2002). Cognitive and psychosocial correlates of adults' eyewitness accuracy and suggestibility. Personality and Individual Differences, 33, 49-66. Lindsay, D. S. (1990). Misleading suggestions can impair eyewitnesses' ability to remember event details. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 10771083. Lindsay, D. S., Hagen, L., Read, J.D., Wade, K. A., & Garry, M. (2004). True photographs and false memories. Psychological Science, 15, 149-154. Lindsay, D. S., & Johnson, M. K. (1989). The eyewitness suggestibility effect and memory for source. Memory & Cognition, 17, 349-358. List, J. A. (1986). Age and schematic differences in the reliability of eyewitness testimony. Developmental Psychology, 22,50-57. Loftus, E. F. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7, 560-572. Loftus, E. F. (1977). Shifting human color memory. Memory and Cognition, 5, 696-699. Loftus, E. F. (1979). Eyewitness testimony. Cambridge, MA: Harvard University Press. Loftus, E. F. (1979). Reactions to blatantly contradictory information. Memory and Cognition, 7, 368-374. Loftus, E. F. (1980). Impact of expert psychological testimony on the unreliability of eyewitness identification. Journal of Applied Psychology, 65, 9-15. Loftus, E. F. (1996). Eyewitness testimony. Cambridge, MA: Harvard University Press (original work published in 1979). Loftus, E. F. (2003). Make-believe memories. American Psychologist, 867-873. Loftus, E. F., & Banaji, M. (1989). Memory modification and the role of the media. In V. A. Gheorghiu, P. Netter, H. J. Eysenck, & R. Rosenthal (Eds.), Suggestibility: Theory and research (pp. 279-294). Berlin: Springer-Verlag. Loftus, E. F., & Bernstein, D. M. (2005). Rich false memories: The royal road to sucess. In A. Healy (Ed.), Experimental cognitive psychology and its applications: Festschrift in honor of Lyle Bourne, Waler Kintsch, and Thomas Landauer (pp. 101-113). Washington, DC: American Psychological Association. Loftus, E. F., & Castelle, G. (2000). Crashing memories in legal cases. In P. J. VanKoppen & N. H. M. Roos (Eds.), Rationality, information and progress in law and psychology (pp. 115-12 7). Maastricht: Maastricht University Press. Loftus, E. F., & Davis, D. (2006). Recovered memories. Annual Review of Clinical Psychology, 2, 469-498. Loftus, E. F., & Greene, E. (1980). Warning: Even memory for faces may be contagious. Law and Human Behavior, 4, 323-334. Loftus, E. F., & Ketcham, K. (1994). The myth of repressed memory. New York: St. Martin's Press.

232

DAVIS AND LOFTUS

Loftus, E. F., & Mazzoni, G. A. L. (1998). Using imagination and personalized suggestion to change people. Behavior Therapy, 29, 691-706. Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589. Loftus, E. F., & Zanni, G. (1975). Eyewitness testimony: The influence of the wording of a question. Bulletin of the Psychonomic Society, 5, 86-88. Luus, C. A. E., & Wells, G. L. (1994). The malleability of eyewitness confidence: Co-witness and perseverance effects. Journal of Applied Psychology, 79, 714-723. Macrae, C. N., & MacLeod, M.D. (1999). On recollections lost: When practice makes imperfect. Journal of Personality and Social Psychology, 77,463-473. Maner, J. K., Kenrick, D. T., Becker, D. V., Robertson, T. E., Hofer, B., Neuberg, S. L., et al. (2005). Functional projection: How fundamental social motives can bias interpersonal perception. Journal of Personality and Social Psychology, 88, 63-78. Marquis, K. H., Marshall, J., & Oskamp, S. (1972). Testimony validity as a function of question form, atmosphere, and item difficulty. Journal of Applied Social Psychology, 2, 167-186. Marsh, E. J., & Tversky, B. (2004). Spinning the stories of our lives. Applied Cognitive Psychology, 18, 491-503. Mather, M., Johnson, M. K., & De Leonardis, D. M. (1999). Stereotype reliance in source monitoring: Age differences and neuropsychological test correlates. Cognitive Neuropsychology, 16,437-458. Mazzoni, G., Kirsch, I., & Lynn, S. J. (Eds.). (in press). Using hypnosis in eyewitness memory: Past and current issues. Mahwah, NJ: Lawrence Erlbaum Associates. Mazzoni, G. A. L., & Loftus, E. F. (1996). When dreams become reality. Consciousness and Cognition, 5, 442-462. Mazzoni, G. A. L., Loftus, E. F., & Kirsch, I. (2001). Changing beliefs about implausible autobiographical events: A little plausibility goes a long way. Journal of Experimental Psychology: Applied, 7, 51-59. Mazzoni, G. A. L., Loftus, E. F., Seitz, A., & Lynn, S. J. (1999). Creating a new childhood: Changing beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13, 125-144. Mazzoni, G. A. L., Lombardo, P., Malvagia, S., & Loftus, E. F. (1999). Dream interpretation and false beliefs. Professional Psychology: Research and Practice, 30, 45-50. Mazzoni, G., & Memon, A. (2003). Imagination can create false autobiographical memories. Psychological Science, 14, 186-188. McAdams, D.P. (2001). The psychology of life stories. Review of General Psychology, 5, 100-1122. McCloskey, M., & Zaragoza, M.S. (1985). Misleading post event information and memory for events: Arguments and evidence against memory impairment hypothesis. Journal of Experimental Psychology: General, 14, 1-16. McNally, R. J. (2003). Remembering trauma. Cambridge, MA: Harvard University Press. Meade, M. L., & Roediger, H. L., III. (2002). Explorations in the social contagion of memory. Memory and Cognition, 13, 140-144. Memon, A., & Wright, D. B. (1999). Eyewitness testimony and the Oklahoma bombing. The Psychologist, 12, 292-295. Meudell, P.R., Hitch, G. J., & Kirby, P. (1992). Are two heads better than one? Experimental investigations of the social facilitation of memory. Applied Cognitive Psychology, 6, 525-543. Mineka, S., & Nugent, K. (1995). Mood-congruent memory biases in anxiety and depression. In Memory distortion: How minds, brains and societies reconstruct the past (pp. 173-196). Cambridge, MA: Harvard University Press. Mueller-Johnson, K., & Ceci, S. J. (2004). Memory and suggestibility in older adults: Live event participation and repeated interview. Applied Cognitive Psychology, 18, 1109-1127. Nader, K. (2003). Memory traces unbound. Trends in Neuroscience, 26,65-72.

7.

SOURCES OF MISINFORMATION

233

Nader, K., Schafe, G. E., & le Doux, J. E. (2000). Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature, 406, 722-726. Neisser, U. (1981). John Dean's memory: A case study. Cognition, 9, 1-22. Neuschatz, J. S., Lampinen, J. M., Preston, E. L., Hawkins, E. R., & Toglia, M. P. (2002). The effect of memory schemata on memory and the phenomenological experience of naturalistic situations. Applied Cognitive Psychology, 16, 687-708. Ofshe, R. J., & Watters, E. (1994). Making monsters: False memories, psychotherapy, and sexual hysteria. New York: Scribner. Oorsouw, K. V., & Merckelbach, H. (2004). Feigning amnesia undermines memory for a mock crime. Applied Cognitive Psychology, 18, 505-518. Ornstein, P. A., Gordon, B. N., & Larus, D. M. (1992). Children's memory for a personally experienced event: Implications for testimony. Applied Cognitive Psychology, 6, 49-60. Pasupathi, M. (2001). The social construction of the personal past and its implications for adult development. Psychological Bulletin, 127, 651-672. Payne, B. K. (2001). Prejudice and perception: The role of automatic and controlled processes in misperceiving a weapon. Journal of Personality and Social Psychology, 81, 181-192. Payne, B. K., Lambert, A. J., & Jacoby, L. L. (2002). Best laid plans: Effects of goals on accessibility bias and cognitive control in race-based misperceptions of weapons. Journal of Experi-

mental Social Psychology, 38, 384-396.

Pedersen, A. C. I., & Wright, D. B. (2002). Do differences in event descriptions cause differences in duration estimates? Applied Cognitive Psychology, 16, 769-783. Pennington, N ., & Hastie, R. ( 1986). Evidence evaluation in complex decision making. Journal of

Personality and Social Psychology, 51, 242-258.

Pennington, N., & Hastie, R. (1992). Explaining the evidence: Tests of the story model for juror decision making. Journal of Personality and Social Psychology, 62, 189-206. Pezdek, K., Finger, K., & Hodge, D. (1997). Planting false childhood memories: The role of event plausibility. Psychological Science, 8, 437-441. Pezdek, K., Whetstone, T., Reynolds, K., & Askari, N. E. A. (1989). Memory for real world scenes: The role of consistency with schema expectation. Journal of Experimental Psychology:

Learning, Memory, and Cognition, 15, 587-595.

Pickel, K. L. (2004). When a lie becomes the truth: The effects of self-generated misinformation on eyewitness memory. Memory, 12, 14-26. Piper, A., Jr., Pope, H. G., & Borowiecki, J. J., III. (2000). Custer's last stand: Brown, Scheflin, and Whitfield's latest attempt to salvage "dissociative amnesia." Journal of Psychiatry & Law,

28, 149-213.

Plant, E. A., & Peruche, B. M. (2005). The consequences of race for police officers' responses to criminal suspects. Psychological Science, 16, 180-183. Plant, E. A., Peruche, B. M., & Butz, D. A. (2005). Eliminating automatic racial bias: Making race non-diagnostic for responses to criminal suspects. Journal of Experimental Social Psychol-

ogy, 41, 141-156.

Polage, D. C. (2004). Fabrication deflation? The mixed effects oflying on memory. Applied Cog-

nitive Psychology, 18, 455-465.

Polczyk, R., Wesolowska, B., Gabarczyk, A., Minakowska, I., Supska, M., & Bomba, E. (2004). Age differences in interrogative suggestibility: A comparison between young and older adults. Applied Cognitive Psychology, 18, 1097-1107. Pope, H. G., Jr., Oliva, P. S., & Hudson, J. I. (2002). Repressed memories. The scientific status of research on repressed memories. In D. L. Faigman, D. H. Kage, M. J. Saks, &J. Sanders (Eds.), Science in the law: social and behavioral science issues (pp. 487-526). St. Paul, MN: West Group. Porter, S., Spencer, L., & Birt, A. R. (2003). Blinded by emotion? Effect of the emotionality of a scene on susceptibility to false memories. Canadian Journal of Behavioral Science, 35, 165-175. Read, J.D., & Bruce, D. (1984). On the external validity of questioning effects in eyewitness testimony. International Review of Applied Psychology, 33, 33-49.

234

DAVIS AND LOFTUS

Remen, R. (1996). Kitchen table wisdom: Stories that heal. New York: Riverhead Books. Reysen, M. B. (2005). The effect of conformity on recognition judgments. Memory, 13, 87-94. Richeson, J. A., &Ambady, N. (2001). Who's in charge: Effects of situational roles on automatic gender bias. Sex Roles, 44,493-512. Richeson, J. A., &Ambady, N. (2003). Effects of situational power on automatic racial prejudice. Journal of Experimental Social Psychology, 39, 177-183. Ridley, A.M., & Clifford, B. R. (2004). The effects of anxious mood induction on suggestiblity to misleading post-event information. Applied Cognitive Psychology, 18, 233-244. Ridley, A.M., Clifford, B. R., & Keogh, E. (2002). The effects of state anxiety on the suggestibility and accuracy of child eye-witnesses. Applied Cognitive Psychology, 16,547-558. Roebers, C. M., & McConkey, K. M. (2003). Mental reinstatement of the misinformation context and the misinformation effect in children and adults. Applied Cognitive Psychology, 17, 477-493. Roediger, H. L., III. (1996). Memory illusions. Journal of Memory and Language, 35, 76-100. Roediger, H. L., III, Jacoby, J.D., & McDermott, K. B. (1996). Misinformation effects in recall: Creating false memories through repeated retrieval. Memory and Cognition, 35, 300-318. Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. Roediger, H. L., III, Meade, M. L., & Bergman, E. T. (2001). The social contagion of memory. Psychonomic Bulletin and Review, 8, 365-371. Roese, N. J., & Olson, J. M. (1996). Counterfactuals, causal attributions, and the hindsight bias: A conceptual integration. Journal of Experimental Social Psychology, 32, 197-227. Roper, R., & Shewan, D. (2002). Compliance and eyewitness testimony: Do eyewitnesses comply with misleading "expert pressure" during investigative interviewing? Legal and Criminological Psychology, 7, 155-163. Rosenhan, D. L. (1973). On being sane in insane places. Science, 179, 250-258. Ross, M., & Wilson, A. E. (2000). Constructing and appraising past selves. In D. L. Schacter & E. Scarry (Eds.), Memory, brain, and belief (pp. 231-258). Cambridge: Harvard University Press. Rothbart, M., & Birrell, P. (1977). Attitude and perception of faces. Journal of Research in Personality, 11, 209. Rovee-Collier, C., Borza, M.A., Adler, S. A., & Boller, K. (1993). Infants' eyewitness testimony: Effects of postevent information on a prior memory representation. Memory and Cognition, 21, 267-279. Rubin, D. C. (Ed.). (1996). Remembering our past: Studies in autobiographical memory. Cambridge: Cambridge University Press. Safer, M.A., Christianson, S.-A., & Autry, M. W. (1998). Tunnel memory for traumatic events. Applied Cognitive Psychology, 12, 99-117. Safer, M.A., Levine, L. J., & Drapalski, A. L. (2002). Distortion in memory for emotions: The contributions of personality and post-event knowledge. Personality & Social Psychology Bulletin, 28, 1495-1507. Salmon, K., & Pipe, M.-E. (2000). Recalling an event one year later: The impact of props, drawing and a prior interview. Applied Cognitive Psychology, 14, 99-120. Salovey, P., & Turk, D. C. (1991). Clinical judgment and decision-making. In C. R. Snyder & D. R. Forsyth (Eds.), Handbook of social and clinical psychology: The health perspective (pp. 416437). Elmsford, NY: Pergamon Press. Schacter, D. L. (1996). Searching for memory: The brain, the mind, and the past. New York: Basic Books. Schacter, D. L. (2001). The seven sins of memory: How the mind forgets and remembers. New York: Houghton Mifflin Company. Schacter, D. L., Koutstaal, W., Johnson, M. K., Gross, M.S., & Angell, K. E. (1997). False recollection induced by photographs: A comparison of older and younger adults. Psychology and Aging, 12, 203-215.

7.

SOURCES OF MISINFORMATION

235

Schank, R. C. (1982). Dynamic memory: A theory of reminding and learning in computers and people. Cambridge: Cambridge University Press. Schooler, J. W., & Tanaka, J. W. (1991). Composites, compromises, and CHARM: What is the evidence for blend memory representations? Journal of Experimental Psychology: General, 120, 96-100. Schuurmans, E., & Vandierendonck, A. (1985). Recall as communication: Effects of frame anticipation. Psychological Research, 47, 199-124. Schwartz, B. L., Meisssner, C. A., Hoffman, M., Evans, S., & Frazier, L. D. (2004). Event memory and misinformation effects in a gorilla. Animal Cognition, 7, 93-100. Scoboria, A., Mazzoni, G., Kirsch, I., & Milling, L. S. (2002). Immediate and persisting effects of misleading questions and hypnosis on memory reports. Journal of Experimental Psychology: Applied, 8, 26-32. Sedikides, C. (1990). Effects of fortuitously activated constructs versus activated communication goals on person impressions. Journal of Personality and Social Psychology, 58, 397-408. Semin, G. R., & De Poot, C. J. (1997a). Bringing partiality to light: Question wording and choice as indicators of bias. Social Cognition, 15, 91-106. Semin, G. R., & De Poot, C. J. (1997b). The question-answer paradigm: You might regret not noticing how a question is worded. Journal of Personality and Social Psychology, 73, 4 72-480. Semmler, C., Brewer, N., & Wells, G. L. (2004). Effects of postidentification feedback on eyewitness identification and nonidentification confidence. Journal of Applied Psychology, 89, 334-346. Sharman, S. J., Garry, M., & Beuke, C. J. (2004). Imagination or exposure causes imagination inflation. American Journal of Psychology, 117, 157-169. Sharman, S. J., Manning, C. G., & Garry, M. (2005). Explain this: Explaining childhood events inflates confidence for those events. Applied Cognitive Psychology, 19, 67-74. Shaw, J. S., III. (1997). Co-witness information can have immediate effects on eyewitness memory reports. Law and Human Behavior, 21, 503-524. Shaw, J. S., III, Bjork, R. A., & Handal, A. (1995). Retrieval induced forgetting in an eyewitness memory paradigm. Psychonomic Bulletin and Review, 2, 249-253. Sherman, J. W., & Bessenoff, G. R. (1999). Stereotypes as source-monitoring cues: On the interaction between episodic and semantic memory. Psychological Science, 10, 106-110. Sherman, S. J., Cialdini, R. B., Schwartzman, D. F., & Reynolds, K. D. (1985). Imagining can heighten or lower the perceived likelihood of contracting a disease: The mediating effect of ease of imagery. Personality and Social Psychology Bulletin, 11, 118-12 7. Simons, D. J., & Levin, D. T. (1998). Failure to detect changes to people during a real-world interaction. Psychonomic Bulletin and Review, 5, 644-649. Smith, V. L. (1991). Prototypes in the courtroom: Lay representations oflegal concepts. Journal of Personality and Social Psychology, 61, 857-872. Smith, V. L., & Ellsworth, P. C. (1987). The social psychology of eyewitness accuracy: misleading questions and communicator expertise. Journal of Applied Psychology, 72, 294-300. Smith, V. L., & Studebaker, C. A. (1996). What do you expect?: The influence of people's prior knowledge of crime categories on fact-finding. Law and Human Behavior, 20, 517-532. Snyder, M., & Thomsen, C. J. (1988). Interactions between therapists and clients: Hypothesis testing and behavioral confirmation. In D. C. Turk & P. Salovey (Eds.), Reasoning, inference, and judgment in clinical psychology. New York: Free Press. Snyder, M., & Uranowitz, S. W. (1979). Reconstructing the past: Some cognitive consequences of person perception. Journal of Personality and Social Psychology, 36, 941-950. Spaniol, J., & Bayen, U. J. (2002). When is schematic knowledge used in source monitoring? Journal of Experimental Psychology: Learning, Memory, and Cognition, 28, 631-651. Spanos, N. P., Burgess, C. A., Burgess, M. F., Samuels, C., & Blois, W. 0. (1999). Creating false memories of infancy with hypnotic and non-hypnotic procedures. Applied Cognitive Psychology, 13, 201-218.

236

DAVIS AND LOFTUS

Spiecker, S.C., & Worthington, D. L. (2003). The influence of opening statement/closing argument organizational strategy on juror verdict and damage awards. Law and Human Behavior, 27, 437-456. Stephenson, G. M., Brandstatter, H., & Wagner, W. (1983). An experimental study of social performance and delay on the testimonial validity of story recall. European Journal of Social Psychology, 13, 175-191. Stephenson, G. M., & Wagner, W. (1989). Origins of misplaced confidence effect in collaborative recall. Applied Cognitive Psychology, 3, 227-236. Sutherland, R., & Hayne, H. (2001). The effect of postevent information on adults' eyewitness reports. Applied Cognitive Psychology, I5, 249-263. Tafarodi, R. W., Marshall, T. C., & Milne, A. B. (2003). Self-esteem and memory. Journal of Personality and Social Psychology, 84, 29-45. Taylor, S. E., & Crocker, J. (1981). Schematic bases of social information processing. In E. T. Higgins, C. P. Herman, & M.P. Zanna (Eds.), Social cognition: The Ontario Symposium (Vol. 1, pp. 89-134). Hillsdale, NJ: Lawrence Erlbaum Associates. Thierry, K. L., & Spence, M. J. (2002). Source-monitoring training facilitates preschoolers' eyewitness memory performance. Developmental Psychology, 38, 428-237. Thomas, A. K., & Loftus, E. F. (2002). Creating bizzare false memories through imagination. Memory & Cognition, 30, 423-431. Thomassin, L., & Alain, M. (1990). Performance of witnesses when giving evidence and making eyewitness identification. Canadian Police College Journal, I4, 233-246. Todorov, A. (2002). Communication effects on memory and judgment. European Journal of Social Psychology, 32, 531-546. Tomes, J. L., & Katz, A. N. (1997). Habitual susceptibility to misinformation and individual differences in eyewitness memory. Applied Cognitive Psychology, II, 233-251. Tousignant, J. P., Hall, D., & Loftus, E. F. (1986). Discrepancy detection and vulnerability to misleading postevent information. Memory & Cognition, I4, 329-338. Tuckey, M. R., & Brewer, N. (2003). The influence of schemas, stimulus ambiguity, and interview schedule on eyewitness memory over time. Journal of Experimental Psychology: Applied, 9, 101-118. Tuckey, M. R., & Brewer, N. (in press). How schemas affect eyewitness memory over repeated retrieval attempts. Applied Cognitive Psychology. Turk, D. C., & Salovey, P. (Eds.). (1988). Reasoning, inference, and judgment in clinical psychology. New York: Free Press. Turtle, J. W., & Yuille, J. C. (1994). Lost but not forgotten details: Repeated eyewitness recall leads to reminiscence but not hypermnesia. Journal of Applied Psychology, 79, 260-271. Tversky, B., & Marsh, E. J. (2000). Biased retellings of events yield biased memories. Cognitive Psychology, 40, 1-38. van den Bos, K. (2003). On the subjective quality of social justice: The role of affect as information in the psychology of justice judgments. Journal of Personality and Social Psychology, 85, 482-498. Vornick, L.A., Sharman, S. A., & Garry, M. (2003). The power of the spoken word: Sociolinguistic cues influence the misinformation effect. Memory, II, 101-109. Vrij, A. (1997). Wearing black clothes: The impact of offenders' and suspects' clothing on impression formation. Applied Cognitive Psychology, II, 4 7-53. Wade, E., & Clark, H. H. (1993). Reproduction and demonstration in quotations. Journal of Memory and Language, 32, 805-819. Walther, E., Bless, H., Strack, F., Rackstraw, P., Wagner, D., & Werth, L. (2002). Conformity effects in memory as a function of group size, dissenters and uncertainty. Applied Cognitive Psychology, I6, 793-810. Wason, P. C. (1960). On the failure to eliminate hypotheses in a conceptual task. Quarterly Journal of Experimental Psychology, I2, 129-140.

7.

SOURCES OF MISINFORMATION

237

Wells, G. L., & Bradfield, A. L. (1998). "Good, you identified the suspect": Feedback to eyewitness reports distorts their reports of the witnessing experience. Journal of Applied Social Psychology, 83, 360-376. Wells, G. L., Olson, E. A., & Charman, S.D. (2003). Distorted retrospective eyewitness reports as functions of feedback and delay.Journal of Experimental Psychology: Applied, 9, 42-52. Wilson, A. E., & Ross, M. (2003). The identity function of autobiographical memory: Time is on ourside.Memory, 11,137-149. Wilson, J. C., Powell, M. B., Raju, S., & Romeo, R. (2004). Secrecy and suggestibility: Are children's memories for secrets less suggestible than other memories. Applied Cognitive Psychology, 18, 251-261. Winkel, F. W. (1999). A frustration/negative cue model of unfavorable police treatment of black citizens. International Journal of Law and Psychiatry, 22, 273-286. Winman, A., Juslin, P., & Bjoerkman, M. (1998). Journal of Experimental Psychology: Learning, Memory, and Cognition, 24, 415-431. Winograd, E., Peluso, J. P., & Glover, T. A. (1998). Individual differences in susceptibility to memory illusions. Applied Cognitive Psychology, 12, S5-S27. Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at the implicit level and its relationship with questionnaire measures. Journal of Personality and Social Psychology, 72, 262-274. Wright, D. B., Loftus, E. F., & Hall, M. (2001). Now you see it; now you don't: Inhibiting recall and recognition of scenes. Applied Cognitive Psychology, 15, 471-482. Wright, D. B., Self, G., & Justice, C. (2000). Memory conformity: Exploring misinformation effects when presented by another person. British Journal of Psychology, 91, 189-202. Wyer, R. S., Jr. (2004). Social comprehension and judgment: The role of situation models, narratives, and implicit theories. Mahwah, NJ: Lawrence Erlbaum Associates. Wyer, R. S., Jr., Adaval, R., & Colcombe, S. J. (2002). Narrative-based representations of social knowledge: Their construction and use in comprehension, memory and judgment. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 133-197). San Diego: Academic Press. Wyer, R. S., &Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates. Zanni, G. R., & Offermann, J. T. (1978). Eyewitness testimony: An exploration of question wording upon recall as a function of neuroticism. Perception and Motor Skills, 46, 163-166. Zaragoza, M.S., & Lane, S. M. (1998). Processing resources and eyewitness suggestibility. Legal and Criminological Psychology, 3, 305-320.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

8 False Memory Research: History, Theory, and Applied Implications Jeffreys. Neuschatz University of Alabama, Huntsville

James M. Lampinen University of Arkansas

Michael P. Toglia State University of New York, Cortland

David G. Payne Educational Testing Service

Elizabeth Preston Cisneros Lama Linda University

Clarence Elkins was convicted of the 1998 rape and murder of his mother~in~law, Judith Johnson, and the rape and beating of his 6~year~old niece, Brooke Sutton. His niece wit~ nessed the murder and was left to die by the murderer~rapist. With no physical evidence linking Elkins to the attack, his niece's testimony was the compelling crucial evidence that persuaded the jury to convict her uncle. Only hours after the attack, the girl told police that the intruder was her uncle. While in jail Elkins was able to obtain a cigarette butt of another inmate, Earl Mann, who was a convicted sex offender and was staying next door to the Elkinses at the time of the attack. DNA testing revealed that the DNA from the cigarette was a match with the samples found at the murder scene. After spending 7 years in prison, Elkins was exonerated in December of 2005. This example illustrates that illusory or false memories can have damaging effects. This issue is par~ ticularly important to practitioners in the legal field because the consequence of false 239

240

NEUSCHATZ ET AL.

memories can be devastating, as in the case of Clarence Elkins. Furthermore, this is just one of many cases in which people have been wrongly convicted and incarcerated based on a false memory. The term false memories has been used broadly to refer to memorial experiences that are remembered differently from the way the experience actually occurred, as well as accounts of recollected events that never actually happened. False memories can include, but are not limited to, a word misremembered as being in a list when it was not presented, remembered items from a picture that were not presented in the scene, and misrembered events from childhood. As we document, these inaccurate memories are often associated with the same vivid experiential details that accompany veridical memories. Thus, vivid false recollections may be difficult to discriminate from true memories. The topic of illusory memory is one that has recently received a great deal of attention in the popular media because of a multitude of high-profile criminal cases in which the major incriminating evidence has been memories exhumed, sometimes with questionable techniques, during the course of psychotherapy; however, memory theorists have long been concerned with false memories. Their concern, in part, is based on the notion that understanding illusory recollections can aid in the understanding of human memory in general. Interest in memory illusions surfaced at least as early as Bartlett's (1932) seminal work on reconstructive memory, with which we begin to trace the history of false memory research. This historical sketch also includes studies focusing on verbal learning (e.g., implicit associative responses), schemas, the misleading post-event information paradigm, the Deese/Roediger and McDermott (DRM) paradigm, and imagination inflation. Within the context of this overview, we discuss three important factors. First, we outline the evolution of false memory research from simple word lists to more complex and lifelike materials such as stories, dreams, staged crimes, and lectures. Second, we illustrate the ease with which false memories can be created, the confidence with which they are held, and the difficulty in guarding against them. Particular attention is given to the phenomenological experiences associated with false memories. Third, we discuss several methodologies (e.g., warnings, varied testing methods) by which researchers have attempted to eliminate or reduce false recollections. In the concluding sections of this chapter we (a) describe ways in which false memories occur in the real world, such as in therapeutic and legal environments, and (b) provide a series of recommendations and suggestions as to how therapists and legal practitioners may approach the difficulty, faced by experimentalists as well, of distinguishing real from false memories.

HISTORICAL OVERVIEW The genesis of false memory research can be traced to Sir Fredric Bartlett's (1932) writings on reconstructive memory. As noted by Koriat, Goldsmith, and Pansky (2000), although Bartlett is generally given credit as being the founder of the reconstructive memory approach, some of this work was well under way prior to the appearance of Bartlett's influential book. More specifically, in the early 1900s, Binet (1900), Stern (1910) and

8.

FALSE MEMORY RESEARCH

241

others demonstrated that children amended their memories to include misleading infor, mation suggested by an interviewer. For example, Munsterberg (1923) induced young children to remember that their teacher had a mustache, when, in fact, he was clean shaven (see Ceci & Bruck, 1993; Ceci, Ross, & Toglia, 1987, for reviews). At about the same time, Gestalt psychologists became disenchanted with Ebbinghaus's view of mem, ory as reproductive (e.g., Ebbinghaus & Ruger, 1913) and began investigations that ap, proached memory as much richer than its verbatim components. Importantly, their in, terests also extended beyond verbal materials and included the study of remembering nonverbal materials such as objects. In spite of this body of early work on the reconstructive nature of memory, Bartlett's (1932) book is generally cited as the beginning of the reconstructive memory move, ment. In his best,known work, Cambridge University participants were asked to read and remember the Native American folktale "The War of the Ghosts." Upon repeated testing of participants' memory for the folktale, Bartlett found that their recall protocols contained many errors of commission or remembered information that was not actually conveyed in the folktale. The added information generally made the story more similar to what the participants had come to expect in folktales with which they were familiar. In other words, the additions made the stories more consistent with their schemas, a term Bartlett borrowed from Piaget and others to refer to patterns of knowledge or a mental model defining what typically occurs in a particular situation. Schemas include contextual items, actions, and thoughts that typically occur in a particular situation. Thus, for Bartlett's Cambridge, England subjects, their folktale schema probably included a hero, an antihero, a crisis, and a resolution. Similarly, a schema for an office would probably contain a desk and a computer. Research regarding false memories was relatively dormant until the late 1950s and early 1960s (Deese, 1959; Underwood, 1965). Underwood investigated illusory memo, ries within the context of a continuous recognition testing paradigm, wherein partici, pants study a long list of words and their task is to determine whether each presented word is new or if the word had previously been presented in the list. Underwood found that when new words (i.e., spaghetti) were associatively related to previously studied words (i.e., meatball), participants were more likely to indicate that they remembered these new items than they were new items that were not associatively related to previously studied words (i.e., dog). Underwood described the effect in terms of implicit associative re, sponses (lARs). Consider, for example, that studying the word black makes one think of the word white. If at testing the word white is encountered, a participant might think that it was on the study list, because it had been implicitly activated in memory by the earlier association. In the early 1970s, several researchers began studying the memory errors in there, tention of prose. For example, Bransford and Franks (1971) presented participants with simple sentences that could be combined to form longer, more complex sentences con, taining four propositions (idea units). They used several complex sentences, each form, ing a short story or theme that was never presented. For example, the thematic sentence The girl who lived next door broke the large window on the porch could be broken down into several simple sentences: The girl lived next door, The girl broke the large window, The girl

242

NEUSCHATZ ET AL.

who lived next door broke the large window, and so forth. On a subsequent recognition memory test, Bransford and Franks found that participants were more likely to say that the full four-idea sentence was actually studied than sentences containing fewer propositions. Furthermore, participants were more confident that the complex sentences were presented at study than they were about the actual studied sentences. Also in the 1970s, Elizabeth Loftus and her colleagues (Loftus, 1979; Loftus, Miller, & Burns, 1978; Loftus & Greene, 1980) developed the misleading post-event information (MPI) paradigm as way of studying suggested memories. In the basic paradigm, participants are exposed to some event. For example, in the seminal studies they watched a slide show depicting what appeared to be a car-pedestrian accident. Early in the sequence participants saw a red Datsun moving along a street, approach, and stop at a stop sign. Next, the participants answered a questionnaire about the event. For some participants the questions contained only consistent information regarding the event (e.g., "Did another car pass the red Datsun while it was stopped at the stop sign?"). Other participants, however, received misleading information via the questionnaire (e.g., "Did another car pass the red Datsun while it was stopped at the yield sign?"). On the final recognition memory test, participants who received misleading information were much more likely to indicate that the original traffic sign was a yield sign than were control participants. With this paradigm, Loftus and colleagues have repeatedly shown the influence of misleading information on the recollection of specific details; for example, that participants will indicate that a stop sign was a yield sign, a barn was presented in a bucolic scene in which there was no barn, or that there was broken glass after an accident in which there was no broken glass. In the 1980s, psychologists adapted the schema concept to memory for naturalistic events and found evidence for false memories for schema-consistent items. For instance, Nakamura, Graesser, Zimmerman, and Riha (1985) had students enrolled in an experimental psychology course view a lecture by a speaker who performed actions that were relevant (e.g., underlining a word on the blackboard) or irrelevant (e.g., sipping coffee) to the lecture. They found that participants' recognition memory was better for schemairrelevant actions because participants rarely remembered the lecturer engaging in irrelevant actions that he did not actually perform. Conversely, falsely remembering relevant actions that were never actually executed by the lecturer occurred often; for example, some participants remembered the teacher erasing the board, when in fact he did not do so. A few years earlier, Brewer and Treyens (1981) had demonstrated that participants who were seated briefly in what they thought was a graduate student's office later falsely recalled and recognized objects (e.g., a stapler) that were consistent with (relevant to) an office schema. In recent years, there has been an explosion of research on false memories. This dramatic proliferation has been fueled in part by recent media exposure of repressed/ recovered memories in many high-profile court cases (e.g., George Franklin, Paul Ingram, and others; Franklin v. Duncan, 1995; State of Washington v. Ingram, 1988; see Smith & Gleaves, this volume). During this surge of interest, many researchers have adopted a paradigm that was originally developed by Deese (1959), later popularized and extended by Roediger and McDermott (1995) and has since been referred to as the DRM (Deese,

8.

FALSE MEMORY RESEARCH

243

Roediger, & McDermott) paradigm. In the DRM paradigm, participants are presented with lists of words (e.g., bed, rest, nap, etc.) that are all associatively related to a single nonpresented critical lure (e.g., sleep). This procedure results in very high rates of false recognition and recall of the critical nonpresented (CNP) lure (Payne, Elie, Blackwell, & Neuschatz, 1996; Read, 1996; Roediger & McDermott, 1995). It is important to note that much of the research discussed thus far has employed recognition testing, but the DRM paradigm is exceptional in that it produces high rates of illusory memories with both recognition and recall testing (e.g., Toglia, Neuschatz, & Goodwin, 1999). The DRM paradigm is popular, in part, because the ease with which the paradigm can be carried out makes it an economical technique for examining basic questions about the nature of human memory. Researchers have produced a number of important findings of practical significance concerning the nature of the false memory effect with this paradigm. For instance, since the inception of DRM research it has been repeatedly shown that participants endorse critical nonpresented items with high levels of sub, jective confidence (Payne et al., 1996; Roediger & McDermott, 1995; Toglia et al., 1999), an endorsement that also typifies false recollections of events (Loftus, 2004; Lof, tus & Bernstein, 2005; Moritz & Woodward, 2002). They not only claim to have seen or heard the critical nonpresented word, but participants continue to endorse critical non, presented items, even when given warnings about the nature of the task (although warn, ings can sometimes reduce the magnitude of the effect, as shown by Gallo, Roberts, & Seamon, 1997; McDermott & Roediger, 1998, and with practice combined with warnings, even substantial reductions are possible, as was demonstrated by Watson, McDermott, & Balota, 2004). The effect is not limited to memory for words, because false memories also occur, though at somewhat lower levels, for similarly organized presentations of pic, tures (Israel & Schacter, 1997; Schacter, Israel, & Racine, 1999) and stories (Lampinen, Faries, Neuschatz, & Toglia, 2000). Such materials offer a theme that, once extracted, induces associatively related false memories. Although illusory recollections are proba, bly not entirely stimulus,driven in DRM,type studies, this seems to be the major factor. Because of the minimal role played by the experimenter in these studies, false memory creation has been viewed as a spontaneous process. Despite this so,called spontaneity, recent thinking has argued for the role of both encoding and retrieval factors in the false memory effect via activation,monitoring theory (McDermott & Watson, 2001; Roediger, Watson, McDermott, & Gallo, 2001). Critical items activated at study are not necessar, ily reported, because meta,memorial editing (monitoring) may successfully prevent them from being produced on recognition or recall tests. In contrast with situations that promote spontaneous false memories, a number of researchers have studied memory inaccuracies in what has been termed the implanta, tion paradigm (Hyman, Husband, & Billings, 1995; Hyman & Pentland, 1996; Loftus & Pickrell, 1995). Such inaccuracies are often referred to as implanted false memories (see Brainerd, Reyna, & Poole, 2000, for an extensive discussion of spontaneous vs. implanted false memories). In a typical study of this type, participants are asked to recall several events from their childhood. Some of the events actually occurred, and some have been concocted by the experimenter. (The experimenter consults the participants' relatives to confirm whether the events did or did not occur.) After the initial interview, participants

244

NEUSCHATZ ET AL.

are asked to continue to attempt to recall more details about the true and false events, either through repeated interviews or by writing in a journal each night. On the final interview, participants are asked to recall all of the events. Typically as many as 30% of the participants come toremember specific details about the false events (Hyman et al., 1995; Hyman & Pentland, 1996; Loftus & Pickrell, 1995; Pezdek, Finger, & Hodge, 1997; Pezdek & Hodge, 1999; Porter, Yuille, & Lehman, 1999). As we complete this historical overview, we turn to an examination of work by Garry, Manning, Loftus, and Sherman (1996), who developed the imagination inflation paradigm. In their original experiment, participants completed the Life Events Inventory (LEI), with which participants rated the likelihood that several plausible life events actually occurred. For example, questions included how likely it was that he or she had found a $10 bill in a parking lot or had broken a window with his or her hand. The LEI rating scale ranges from 1 (definitely did not happen) to 8 (definitely did happen). After completing the LEI, participants were asked to return to the laboratory at some later point (2 weeks in the original study, but varying intervals have since been employed), and they were asked to imagine some of the events from the LEI. Nter the imagination task, participants completed the LEI a second time. In the original study and several follow~ups, comparison of the scores for the two administrations of the LEI generally reveals that items that were imagined are rated as being more likely to have happened than when initially rated. Researchers have interpreted this finding to mean that the mere act of imagining an event increases participants' belief that it occurred (Garry et al., 1996; Garry, Sharman, Wade, Hunt, & Smith, 2001; Heaps & Nash, 1999; Maz~ zoni & Memon, 2003; Paddock et al., 1998). Goff and Roediger (1998) employed a variant of the imagination inflation paradigm to demonstrate that imagination inflation actually could lead to the creation of false memories. In the original paradigm (Garry et al., 1996), the experimenter had no con~ trol over whether the imagined events actually occurred. Goff and Roediger created a laboratory analogue in which they manipulated which events occurred and which did not. During the first session in this study, participants heard, performed, or imagined simple actions (e.g., picking up a ruler). In the second session, participants imagined actions one, three, or five times. The actions in this session were items that were heard, performed, or imagined in the first session. In the final session, participants completed a recognition test. Following the recognition test, participants were asked to indicate whether the items that they recognized were heard, performed, or imagined in the first session. The primary result was that increasing the number of imaginings increased the likelihood that participants would indicate that imagined actions had been performed. The importance of this experiment is that it demonstrates imagination inflation while maintaining experimental control of the stimulus events. As the research on false memories has evolved it has become much more complex and has more closely simulated real-life situations. As is more evident in the next section, research has also become increasingly concerned with the phenomenological experience associated with false memories. Finally, it will also be made apparent that re~ searchers have become progressively more concerned with generalizability to more real~ life situations, as experiments have been designed to more closely mimic what can hap-

8.

FALSE MEMORY RESEARCH

pen sometimes in therapeutic environments, and hence the employment of the tation and imagination inflation paradigms.

245

implan~

PREVALENCE, PHENOMENOLOGY, AND WARNINGS

In this section we discuss the pervasiveness of false memory creation in the laboratory and how this relates to more applied settings. The experience of illusory memories and how they might be prevented or attenuated are also described in this section. We start by reexamining the DRM paradigm. Close attention is paid to the question of whether participants really believe that false memories are based on actual experience. Thus, a central question in the false memory literature concerns whether participants in fact be~ lieve they have recollections of these purported false memories, or if they are just sue~ cumbing to experimental demand. To answer this question, we draw on research that deals with the phenomenological experience associated with these items, relative to those of veridical memories. We then discuss some of the problems that have been raised in the literature with regard to generalization of the results from the DRM para~ digm to applied settings. The last part of this section is devoted to a discussion of other memory illusion paradigms that are relevant to the ecological validity of false memory research. The subjective experience that accompanies a false memory is an important aspect that distinguishes an illusory recollection from other memory errors. That is, partici~ pants in the DRM paradigm not only recall and recognize items that were never pre~ sented, but they also recall details of the presentation of those never~presented items. In their seminal study, Roediger and McDermott (1995) used Tulving's (1985) remember/know distinction in conjunction with the DRM paradigm. In the remember/ know task, participants are asked to indicate whether the items they recall or recognize are remembered or known items. "Remember" items are defined as those about which the participant can independently recollect some detailed aspect of the presentation of the item (e.g., the speaker's voice, where the item was presented in the list). "Know" items are defined as items about which the participant cannot recollect any specific details, but because of a strong sense of familiarity, the participant is sure that the items were presented. Roediger and McDermott found participants were quite willing to give critical nonpresented items remember responses. In fact, they indicated that they remem~ bered true and false memories at about the same very high rate. This finding regard~ ing phenomenology has been replicated several times (Gallo & Roediger, 2003; Mather, Henkel & Johnson, 1997; Payne et al., 1996; Schacter, Verfarllie, & Paradere, 1996). These results are somewhat surprising, because participants are actually indicating that they remember details of items that were never actually presented. Not only will participants indicate that that they remember something about the presentation of these items, they will even go so far as to confabulate details about the presentation of the items, as demonstrated by Payne et al. (1996). After viewing two speakers alternately reading several DRM lists, participants were asked to recall the list three times. After the final recall, participants were asked to go back over their final

246

NEUSCHATZ ET AL.

recall protocols and indicate which of the two speakers had spoken the recalled words. Participants were warned to attribute the word to a speaker only if they were certain that the person had indeed said the word. Even with this cautionary instruction, participants indicated almost 90% of the time that they could remember the person who said the critical nonpresented items that were never spoken. In comparison, the modal number of attributions for items that were recalled but were not related to a critical theme was zero. Also in a DRM context, similar results involving the misattribution of voices were reported by Roediger, McDermott, Pisani, and Gallo (2004). In sum, these findings suggest that participants believe they really remember experiencing the presentation of these false items. The possibility still remains that participants are yielding (giving in to experimental demand) rather than actually having a memory of someone saying the word. Once a participant has committed to remembering a test word, they may reason that an experimenter must have said the word. This line of thinking results in selecting a speaker. Employing such a strategy would make some sense in that most people would agree that they typically do not falsely remember items and events that did not occur, so that the experimenter must have said the item. There are several lines of evidence that argue against the demand characteristic interpretation, and this evidence also suggests that participants actually think that items were presented. One form of evidence comes from a study conducted by Lampinen, Neuschatz, and Payne (1999) in which they directly tested whether participants actually believe they remember the source of the critical nonpresented items or if they are simply yielding to experimental demand. In order to do this, participants listened to a list of words read aloud by two experimenters. After completing the memory test, participants indicated which of the speakers said the words that were recognized. Following this source attribution task, participants were instructed to rate how confident they were about their source attributions (Experiment 1), or they were told that some of their source attributions were incorrect and they could change their responses if they chose to do so (Experiment 2). The authors hypothesized that if source attributions were due to experimental demand, then participants should be less confident in their attributions and should be more willing change the source attribution for critical nonpresented items, as opposed to true memories. In contrast, if participants think that the critical nonpresented items were actually experienced and their memories felt veridical, then they should be as confident and as likely to change their source attributions for true and illusory memories. The results revealed that participants were slightly but not significantly more confident in their attributions for true than false memories. The more important finding was that they were not more likely to change their attributions for false compared with true memories, an outcome indicating that participants were not giving in to experimental demand. Lampinen et al. concluded that under these circumstances people come to believe that they really remember one of the experimenters speaking these nonpresented items. Another indication of the compelling nature of these false memories comes from studies in which participants are warned about the nature of the DRM lists. Once again, if participants were simply yielding to demand when they claimed to falsely remember

8.

FALSE MEMORY RESEARCH

247

the critical nonpresented items, then they should inhibit reporting these items when they are warned, because the demand characteristic is in opposition to reporting these items. The first experiment to employ warnings in the DRM paradigm was conducted by Gallo et al. (1997). In the condition with strongest warning, referred to as the fore, warned condition, participants were explicitly cautioned about the nature of the DRM lists. To bolster the potential effectiveness of the forewarning, participants were provided with a sample DRM list and were made aware of the critical item that was missing from this list. The results revealed that the forewarned group falsely recognized fewer critical nonpresented items, relative to a No Warning control condition. It is important to note that even though there was a reduction in the false memory rate in the forewarned group, false memories were not eliminated in this condition. In a more stringent test, McDermott and Roediger (1998) also reported only a modest reduction in false recognition due to warnings. In their second experiment, McDermott and Roediger employed a one,item recognition test after each study list, with the one item on the test always being the critical item to which all of the study list items were related, in conjunction with an explicit warning that included an example of the DRM lists. Even though McDermott and Roediger's procedures were intended to maximize the chances of finding an effect of warning (e.g., a single,item recognition test immediately following the study list), they still found a moderately large false memory effect. One reason why warnings might not completely eliminate the false memory effect is that people may be generally unaware of how to differentiate true and false memories. The Source Monitoring Framework predicts that the experiential content of true and false memories should differ systematically (Johnson, Hashtroudi, & Lindsay, 1993). That is, the memory records of true memories usually contain more contextual, perceptual, and emotional details than the memory records of false memories. If participants can identify phenomenological differences between true and false memories, they may be able use this information to avoid reporting false memories. To test this notion, Neuschatz, Payne, Lampinen, and Toglia (2001) employed warnings that included information about the nature of the lists, a sample list, and information regarding the phenomenological differences between true and false memories. For instance, participants were told that memory traces of externally presented items (e.g., words that appeared in a list) are likely to include more perceptual details and fewer details about cognitive operations than are memories of internally generated, or imagined, items. Recognition performance revealed that warning participants about the phenomenological differences did not reduce the false memory effect either immediately or after a 48,hour retention interval. Neuschatz et al. replicated these null results with a recall test in another experiment reported in the same paper. Although the results from the DRM paradigm are important in their own right, there are some questions about the generalizability of this line of research, especially to legal situations when the victim may be the only eyewitness (but see Reyna, Mills, Estrada, & Brained, this volume for a counter argument). For example, Freyd and Gleaves (1996) argued that the research from the DRM paradigm lacks the traumatic and emotional content associated with real,world events such as child sexual abuse. More specifically, they argued that the false memory of critical nonpresented (CNP) items utilized in DRM

248

NEUSCHATZ ET AL.

lists was not for surprising, sexually suggestive, or emotionally charged words. Therefore, instead of presenting participants with a list of words such as walk, leg, heel, ankle, toe, etc. and then examining how often the participant recollected the word foot, the task should be to determine how often the participant would misidentify the word penis as having been presented on the list. Freyd and Gleaves also suggested that having false memories for semantically related words is hardly proof that false memories for events as complex as childhood sexual abuse can and do occur. They argued that people are much more likely to make memory errors about items on a list of words than about an overall episode. Pesta, Murphy, and Sanders (2001) empirically tested Freyd and Gleaves' (1996) criticism and presented lists of words containing nonemotional words that were orthographically similar (e.g., cape, tape, ripe; part, perk, dark) to either emotional (e.g., rape) or unemotional (e.g., park) critical lures (CNPs). They found that participants were likely to falsely remember emotional lures, even though no emotional words appeared in the study. When there were three emotional presented words in the study list, false recognition of emotional CNP items more than doubled. Although participants were more likely to falsely recollect nonemotional CNP items, they were more likely to assign remember judgments to falsely recollected emotional CNP items. In response to Freyd and Gleaves' (1996) claim that remembering a single item is not the same as falsely remembering an entire episode, which would be the case with child sexual abuse, researchers have targeted more complex and naturalistic settings in designing experiments and continue to find high levels of false memories. Following earlier work by Brewer and Treyens (1981), briefly described previously, and by Pezdek and colleagues (1989), Lampinen, Copeland, and Neuschatz (2001) had participants sit in a room set up to look like a graduate student's office to examine memory for objects in this naturalistic environment. They found that participants were quite likely to falsely recognize schema-consistent office items, and such false recognitions increased over a 48-hour retention interval. Similar results were obtained by Neuschatz, Lampinen, Preston, Hawkins, and Toglia (2002), who had participants watch a short, videotaped lecture in which the instructor enacted several schema-consistent actions (e.g., writing on a whiteboard) and schema-inconsistent actions (e.g., smoking a cigarette). Participant frequently reported schema-consistent actions that, in fact, never occurred during the lecture. Thus, once again the conclusion is that false memories can easily be created with materials that are more complex than simple word lists (Brewer & Treyens, 1981; Lampinen et al., 2000, 2001; Neuschatz et al., 2002; Pezdek et al., 1989). Not only have researchers extended the study of illusory memories to more complex situations, they have also extended the study of false recollections to groups. Basden, Basden, Thomas, and Souphasith (1998) examined the question of whether the collaboration of groups of three people would increase false recall of CNP items in the DRM paradigm. They found that collaborative groups (groups who worked together on the task) were more likely to produce false memories than were nominal groups (groups of people who worked independently on different parts of the task). In addition, collaboration actually inhibited the recall of presented words.

8.

FALSE MEMORY RESEARCH

249

Although the research of Basden et al. (1998) explored the effects of collaborative recall on memory, the DRM materials utilized in their study raise the same cautions as studies involving individual memory for CNP items in the DRM paradigm. That is, in legal and clinical situations, practitioners are not concerned directly with people's memory for lists of words. They are primarily interested in witnesses' and clients' memory of past events. Roediger, Meade, and Bergman (2001) and Meade and Roediger (2002) attempted to create a more naturalistic environment in which to study the possibility that false memories can be contagious when participants are recalling together. In their study, participants came to the lab with another "participant" who was actually an experimental confederate. Both were shown pictures of household scenes such as a kitchen, and then they engaged in a collaborative recall task. In the Contagion condition, the confederate's recall protocol purposely included some items that did not actually appear in the scenes. After hearing the confederate falsely recall items from the pictures, participants frequently incorporated the confederates' recollection into their own memories and later came to recall items from the scene that were, in fact, not present. It is important to note that this situation may have similarities to one in which one sibling tells another sibling about the discovery of a memory, real or fabricated, of childhood abuse or group therapy sessions where many people try to collectively recall their past experiences. Thus, false memories in the context of collaborative recollection bolster the prevalence argument we have been making, while also demonstrating that they appear to be "contagious" (Meade & Roediger, 2002). Gabbert, Memon, Allan, and Wright (2004) and Wright, Self, and Justice (2000) have found that eyewitness memory can be altered when another witness introduces inaccurate information about the event vis-a-vis faceto-face discussion. That is, merely talking to another witness who has a different account of the event can alter one's memory of the event. In addition to group recall, researchers have also studied false memories with the use of a methodology, referred to earlier as the implantation paradigm, developed by Loftus and Pickrell (1995) and exemplified by the "Lost in the Mall" experiment. This line of research is particularly relevant to legal practitioners, because it demonstrates that people can fabricate entire episodes under the proper social conditions. In the Loftus and Pickrell experiment, participants were asked to remember a plausible childhood event (e.g., getting lost in a mall). Although the participants are led to believe that they are recalling an event that actually occurred, in reality the event never happened. After several attempts to recall the information at varying delays, around 30% of participants came to recall some part of the event in great detail, even though the event never occurred. Since the original study, many researches have replicated and extended these findings with a similar procedure (Hyman et al., 1995; Hyman & Pentland, 1996; Loftus, Coan, & Pickrell, 1996; Pezdek et al., 1997; Porter, et al., 1999). In a similar vein, Hyman et al. (1995) asked participants to recall several childhood events. In the first experimental session, participants were to recall and elaborate on two childhood events and two false events that were invented by the experimenters (i.e., knocking over a punch bowl at a party}. Participants were presented with brief descriptions of the events and were to recall the details. The experimenters obtained descriptions of the true childhood events by contacting the participant's parents.

250

NEUSCHATZ ET AL.

After the first session, participants were asked to try hard to remember more details of the four events before the next experimental session (regardless of whether or not they remembered the events), which occurred 2 weeks after the first session. In the first experiment, 20% of the participants came to remember details about the fictitious events. In Experiment 2, in which an extra experimental session was added, 25% of the participants confabulated details of the false events. Using a variation on the implantation paradigm, Wade, Garry, Read, and Lindsay (2002) demonstrated that memories could be implanted by showing digitally altered pictures to participants. Wade et al. (2002) gave participants a photograph of themselves as a young child and asked each of them to recall the event that was depicted in the photograph. One photograph was digitally created; according to the participants' relatives, it never occurred. It was a picture of the participant in the basket of a hot air balloon. Amazingly, 50% of the participants reported remembering something about the fictitious event. Overall the research from the implantation studies seems to suggest that additional recall attempts will lead to more memories of false events, but continued research is needed before any strong conclusion can be made. In fact, Lindsay, Hagen, and Read (2004) reviewed the body of work that employed this methodology and reported that, across all studies, an average of 31% of the participants falsely remembered an event from their childhood. These implantation studies certainly indicate that participants can indeed produce false memories of entire episodes or at least partial episodes. Although many studies have demonstrated that false memories can be reliably produced with a variety of methodologies, materials, and events, these studies typically fail to capture the emotional and traumatic nature that sometimes accompanies false memories, which limits the generalizablity of these results to more naturalistic situations (Berliner & McDougall, 1997; Freyd & Gleaves, 1996). This is especially true in cases of recovered memories, which is, in fact, a situation to which some authors have compared this research (Freyd & Gleaves, 1996; Pezdek et al., 1997; Pezdek & Roe, 1997). For instance, Pezdek and colleagues have specifically argued that the results from implantation studies, like Loftus et al.'s (1996) "Lost in the Mall" study, must be viewed with care because they may not be generalizable to the legal and clinical arena, because the events in question are plausible and relatively benign. In legal or therapeutic settings, they submit, memories are generally more traumatic and are related to atypical events (e.g., childhood sexual abuse). It is important that experimentalists are raising these concerns, ones that not surprisingly have frequently been advanced by clinicians (e.g., Madill & Holch, 2004). Pezdek and Hodge (1999) argued that false memories can be implanted to the extent that rememberers think that the event is plausible and that they possess scriptbased knowledge of the event. It follows logically that Loftus and Pickrell (1995) and Hyman et al. (1995) successfully implanted memories of being lost in a mall and knocking over a punch bowl at a wedding, because these are relatively typical events, and most people probably have some script-based knowledge of these events. To test their hypothesis, Pezdek and Hodge (1999) attempted to implant false memories of a typical and therefore plausible event (getting lost in a mall) and an atypical event (receiving a

8.

FALSE MEMORY RESEARCH

251

rectal enema). Presumably, most children have no script-based knowledge of having a rectal enema, such that it should be viewed as implausible. Pezdek and Hodge had parents read brief descriptions of four events, the two false episodes just noted and two true episodes. The children were either 5-7 -year-olds or 9-12-year-olds, and they were told that the events had taken place when they were age 4. After hearing the descriptions, the children were asked to recall any details that they could about the events. Fourteen of the children remembered one false event, the plausible one, and one child recalled only the enema event. The younger children were more likely to recall a false event, and in fact three of them recalled both the mall and enema episodes; none of the older group remembered details of both illusory events. Pezdek et al. ( 1997) reported similar results with high school students. More specifically, they asked Jewish high school students to remember attending a Catholic ritual, Catholic students to remember attending a Jewish ritual, and both Catholic and Jewish students to remember attending rituals of their own faith when they were 8 years old. She found that participants were much more likely to remember events that were consistent with their own faith, as opposed to those inconsistent with their faith. That is, Catholics were much more likely than Jewish participants to remember the Catholic events, and Jewish participants were much more likely to remember Jewish events. These results support Pezdek and Hodge's (1999) assertion that false memories can be suggestively implanted, if they are plausible and if script-relevant knowledge is available. However, the work of Bartlett, while not using an implantation procedure, revealed long ago that false memories were plentiful when participants did not have script-appropriate knowledge. Bartlett's findings are particularly informative because they demonstrate that participants, when confronted with an event that contains many atypical episodes that are low in plausibility, may attempt to understand and later recall the event from a plausible perspective. The requirement of plausibility has also been questioned, and we now examine that challenge. Plausibility likely plays an important role in moderating false memory effects. However, plausibility itself may be a pliable construct. For instance, Pezdek and Hodge (1999) argued that children have scripts for getting lost in a mall. They argue that children develop scripts based on reading of classic children's fairy tales. If one accepts that children could develop schema-based knowledge from sources such as books, then it is reasonable to assume that adults could develop script-based knowledge from reading and watching television as well. Mazzoni, Loftus, and Kirsch (2001) provided a theoretical explanation of how false memories of atypical implausible events might be created. They argued in favor of a three-step model. First, the event must be perceived as plausible. Second, the person must believe that it is reasonably likely that the events could have happened to him or her. Third, the person must interpret his or her fantasies and thoughts about the events as being memories. In other words, there must be some source confusion about where the items or events in memory originated. The major assumption of this model is that plausibility and likelihood of occurrence are malleable and not fixed. That is, something that seems implausible can become plausible, depending on the context. In order to test their model, Mazzoni et al. (200 1) asked people to complete plausibility ratings and the LEI (Life Events Inventory) for a series of plausible and implausible

252

NEUSCHATZ ET AL.

events. One of the critical events that participants rated for plausibility and likelihood of occurring was "witnessing a person being possessed." Those participants who indicated that that event was implausible and unlikely were selected for inclusion in the rest of the study. In the second phase of the experiment, all participants read 12 articles. Three of these articles dealt with the topic of possession. These articles indicated that satanic possession was quite common and that it is not uncommon for many children to witness a possession and described what typically happens during a possession. In the third phase, participants completed a questionnaire about their fears and were given a suggestion that their answers to the questionnaire were very similar to the profiles of someone who had witnessed a possession as a child. Finally, participants were asked to complete the plausibility scale and the LEI. The results revealed that participants who read the articles and received the false feedback rated satanic possession as more plausible and more likely to have happened to them than did control participants. In Experiment 2, Mazzoni et al. (2001) found similar increases in plausibility and likelihood for an event that was initially rated as both implausible and unlikely without giving the participants false feedback. Therefore, it is possible for people to believe that events that never occurred did happen, even if they are improbable and implausible If it is true that plausibility and likelihood are malleable and not fixed, then it should be possible for people in positions of authority (e.g., doctors, police officers, judges, etc.) to influence both factors. This assertion was confirmed by Mazzoni, Lombardo, Malvagia, and Loftus (1999). In this study, Mazzoni and colleagues had participants attend a therapy session in which a certified therapist interpreted a recent dream of the participants. The therapist suggested to all of the participants that the dream was consistent with dreams of persons who had had a traumatic experience before the age 3 (e.g., being lost for an extended period of time). As predicted by the three-step model, participants who received false feedback regarding the dream were more confident that they had experienced the event, even though they had previously denied it. Thus, by simply setting a context that makes memories seem more plausible and likely (i.e., supplying false feedback about a dream), it is possible to induce people to accept that the implausible events in question actually occurred.

APPLICATIONS

Over the last 10 years, research interest in illusory memory has increased dramatically. Initially, experimentalists focused on simply demonstrating the false memory effect and revealing its robustness in a variety of paradigms across a wide array of stimulus materials. Their focus soon shifted, as we have seen, to examining the feel of false recollections and to how might one prevent, perchance not succumb to, such compelling "memories." In all of these types of investigations, memory researchers have not lost sight of the question, What does this research mean for lawyers, juries, and others working in the criminal justice system? The accuracy of the memory of eyewitnesses is crucial in the criminal justice system. In many cases, the most important piece of evidence may be the memory of an eye-

8.

FALSE MEMORY RESEARCH

253

witness. This is especially true in cases that involve repressed/recovered memories of child sexual abuse, where the victim and the accused were likely the only people present during the alleged crime. Because in these situations there is often no corroborating physical evidence, it is the jurors' task to decide if the memories are real or if they are fabricated. Furthermore, it may be important for therapists to be able to assess the validity of memories in order to chart a course of treatment. The first problem that jurors, lawyers, and therapists with little familiarity with the workings of memory must resolve is whether or not it is possible to have vivid memories of events that never occurred. The literature that has been reviewed in this chapter provides convincing evidence that people can, in fact, have detailed memories of events that they have only imagined. As noted earlier, it is relatively easy for people to construct memories of whole events, such as getting lost in a mall (Loftus & Pickrell, 1995) or knocking over a punch bowl at a party as a small child (Hyman et al., 1995). In addition, it is clear that people can have false memories of atypical events that are highly traumatic and emotional. For example, Pynoos and Nader (1989) demonstrated that children produced distorted memories of a sniper attack that took place at a school. These distortions included memories of the attack by children who were on vacation at the time the event took place. Furthermore, Spanos, Burgess, Burgess, Samuels, and Blois (1999) demonstrated that people could have memories of sexual abuse in past lives. These studies, as well as others (Hyman et al., 1995; Loftus & Pickrell, 1995), also demonstrated that participants, under the right conditions, can falsely remember entire events, rather than just isolated items from a particular scene. It is important to note that not all memories of past traumatic events are false. In fact, Wagenaar and Groeneweg (1990) demonstrated that survivors of Nazi concentration camps have vivid detailed memories of events that happened to them some 40 years before they were interviewed. Furthermore, Schooler (1996, 1997, 2001) documented a number of cases in which there was physical evidence corroborating recovered memories of childhood sexual abuse. The point is that not all, perhaps not even most, recovered memories are false memories. The fact that people can have true and false memories that are vivid and perceptually rich complicates the issue of how to discriminate the two types of memories. This leads to two equally important questions. First, are there ways to accurately discriminate true and false memories? Second, are there therapeutic and interviewing procedures that are more risky in terms of constructing false memories? With respect to the first question, it is necessary to investigate ways to accurately distinguish between true and false memories. At the level of reporting it is important to make a distinction between a false memory and a willful fabrication. In the former case, the rememberer believes in the authenticity of his or her memory; in the latter case, the rememberer knows that what he or she is recounting is untrue. There is a large literature on how to distinguish a lie from the truth (see Ekman & O'Sullivan, 1991). However, most research indicates that trained lie detectors exhibit accuracy rates near chance (Vrij, 2000). So setting aside lying, the predicament that one faces outside the laboratory is that in most cases when people are recollecting a false memory, they are convinced that the memories are true. Thus, confidence cannot serve as a surrogate for accuracy (see Read

254

NEUSCHATZ ET AL.

& Connolly, this volume), a point that is reinforced by a large literature showing that a weak confidence-accuracy relationship is typical (see Shaw et al., this volume). Because there is often an artificial sense of confidence, false memories are not similar to fabrications on a variety of dimensions (Porter et al., 1999). Thus, we are skeptical that models of discriminating true memories from lies will be helpful in the discrimination of true and false memories. However, we believe this would be an instructive and fascinating line of inquiry. One possible way to discriminate true and false memories may be contained in the memory records themselves. As alluded to earlier, in the Source Monitoring Framework (e.g., Johnson, et al., 1993; Johnson & Raye, 1981), memory traces are viewed as a record of the cognitive processes and operations that have been performed when an event is experienced. This record is thought to include, in varying degrees, information about the surface structure of the presented item, emotional responses to the item, cognitive operations performed in generation or interpretation of the item, as well as contextual and semantic details. The Source Monitoring Framework assumes that memories from different sources (e.g., imagined vs. externally presented) differ in their average qualitative characteristics. For instance, memory traces of words that appeared in a list are likely to include more perceptual details and fewer details about cognitive operations than are memories of internally generated, or imagined, items. Source errors are thought to occur when participants confuse internally generated items, such as critical nonpresented items, with externally presented items, such as list items. To the extent that internally generated items share features in memories with list items, the result is that participants would experience true and false memories as being quite similar. How, then, do internally generated items take on features that are similar to externally generated memories? If internally generated items come to mind during encoding, it is likely that many of the contextual, perceptual, and emotional details that were occurring at the time will also be included in the memory records of the internally generated items. These records may therefore be treated more like verbatim traces than gist traces. Mather et al. (1997) demonstrated that participants' self-rated phenomenological ratings of true and false memories differ in the ways that are predicted by the Source Monitoring Framework. That is, participants indicated that the memory records of true memories contain more contextual, perceptual, and emotional details than false memories. Others have independently replicated this work (Neuschatz et al., 2001; Norman & Schacter, 1997). These findings might suggest that these characteristics could be used by a trained psychologist to evaluate memory content and decide which memories are real and which are false. In a similar vein, Porter et al. (1999) examined whether features of real, fabricated, and implanted memories could be distinguished. Participants were asked to recall a memory for a real event and an event that never occurred (false event) over three experimental sessions. During the third session, participants were asked to try to remember the real and implanted, or false, memory, as well as to confabulate a memory. More specifically, participants were asked to fabricate a believable traumatic childhood event. The authors conducted a content analysis of all three types of events (real memory, false

8.

FALSE MEMORY RESEARCH

255

memory, and fabricated event) and importantly found that false memories differed in reliable ways from true memories and fabricated reports. Implanted memories were less vivid and clear than real or fabricated memories and were rated as less coherent than the other two types of memories. Furthermore, false memories tended to include fewer details than real memories. These results suggest that true and false memories often differ in discernible ways. However, can these characteristics be employed to discriminate between true and false memories, particularly when the evaluator does not know the veracity of the memory in advance? In other words, could a content analysis expert use the characteristics of real, false, and fabricated memories to distinguish the veracity of the memory? This notion has not been tested thoroughly, but the results of such an examination would be very informative to practitioners, in both the clinical and legal arenas. We are somewhat skeptical, however, that content analysis will provide an effective means of distinguishing memories. This skepticism is based on the fact that participants who have been provided with the characteristics predicted, based on empirical work, to differentiate true and false memories, have been found to be unable to use these attributes diagnostically to discriminate between true and false memories, even whenthey are explicitly warned (Neuschatz et al., 2001). Thus having access to potentially analytical information does not appear to be sufficient for distinguishing between veridical and illusory memories. Before dismissing a diagnostic approach, as a caveat we admit that our research involved participants who were untrained. In addressing this cautionary note, Blandon-Gitlin, Pezdek, Rogers, and Brodie (2005) found that training did not increase participants' ability to discriminate a true from a fabricated event. More specifically, they tested how effectively Criterion-Based Content Analysis (CBCA) could discriminate between true and fabricated reports of familiar and unfamiliar events. The CBCA is the most widely used technique for assessing deception in the verbal protocols (statements) of children (Vrij, Akehurst, Soukara, & Bull, 2002). CBCA is based on the assumption that true and false statements will differ qualitatively, and this qualitative difference can be used to discriminate true and false statements. In the study of Blandon-Gitlin et al. (2005), trained participants could not distinguish between true and false statements, but they did discriminate familiar from unfamiliar events regardless of their veracity. We have spent considerable time discussing the methodology and results of numerous studies that employed the DRM false memory paradigm involving lists of words or pictures. As we turn our attention briefly to questioning practices that legal and clinical practitioners should, and often do, reject, it is legitimate for readers to ask about the relevance of the DRM procedure that in the laboratory effortlessly leads to the creation of illusory recollections. Interviewers and therapists ask questions about core events that are believed to have occurred and may suggest via questions or assertions during questioning that other events may have occurred. The goal is to find out what happened by probing in a manner that is minimally intrusive. This objective is compromised if questions are repeated and/or if the questioner presses, with an array of queries, for details about aspects of an event that may not have occurred. Such a strategy is in line with

256

NEUSCHATZ ET AL.

Fuzzy Trace Theory (FTT), a position articulated by Brainerd and Reyna (1995), in which they argued that when events are encoded, two independent representations are formed in parallel, a verbatim trace and a gist-based trace. Verbatim traces involve specific details of the event captured by a copy of what witnesses saw or what was presented in the case of an experiment. These veridical traces tend to decay rather quickly. On the other hand, memory traces of gist are more general and include a main idea or theme of an episode. Such thematic information is typically quite durable, lasting weeks or longer. In FTT terms, intrusive interviewing is problematic because the questioner is targeting by fiat generalities (as presumably they are almost never privy to any specific/verbatim information). Intrusive, suggestive questioning, when coupled with repetition of evocative questions, can produce suggested memories that are supported by durable memory traces. The strengthening of gist in this manner is tantamount to the repeated cuing of gist, in much the same way that occurs in the DRM paradigm, that, as we have seen, leads to vivid false memories (see Reyna et al., this volume; Toglia, 1996). Critically, these vivid "recollections" may start out as rather vague memories (just the gist) but later evolve into elaborately recalled narratives supported by "verbatim" details. Thus, such questioning tactics should be avoided. Training and continuing education are important in this regard because circumventing suggestive questions is not as easy as it might seem (e.g., Brady, Poole, & Warren, 1999), as interviews may unwittingly phrase questions in suggestive terms. Similarly, asking a witness, victim-witness, or patient to imagine what might have happened as we have seen is apt to create false recollections, and in the case of witnesses, imagination exercises may lead to inaccurate testimony, as noted by Reyna et al. (this volume). Encouraging the interviewee to construct inferences concerning what may have happened is just as damaging according to Reyna et al., because such constructions could be fabrications. Once engendered, however, the constructions take on an aura of reality as they become firmly and confidently held viewpoints. Thus believability, much like plausibility discussed earlier, is another basis on which a person can arrive at the conclusion that a false memory is a true memory. The "risky" techniques we have just counseled against are exacerbated by the timing of the questioning. Unlike most DRM studies (but see Seamon, Luo, & Kopecky, 2002; Toglia et al., 1999), investigators and therapists often do not speak to a witness or client until days, weeks, or even longer periods of time have passed. It is well known that such delays have particularly insidious effects on memory, producing substantial levels of forgetting as well as gist-based alterations in memory (see Read & Connolly, this volume). Also unlike most DRM experiments, therapists and forensic interviewers are confronted frequently with both emotional memories and the emotional tone that tend to accompany such memories. It is natural for patients to recount memories whose fabric contains an affective tinge, and so one must guard against any tendency that one or both of these are diagnostic of true memories (see Reisberg & Heuer, this volume, for an in-depth treatment of emotional recollections). In sum, evaluating the authenticity of remembered experiences is an extremely challenging task. This task becomes more difficult if evaluators unintentionally increase the likelihood that a witness may provide a false

8.

FALSE MEMORY RESEARCH

257

report and then embrace it as a memory. Clearly, additional realistic research is needed to disentangle illusory from true recollections.

REFERENCES Bartlett, F. C. (1932). Remembering: A study in experimental psychology. Cambridge, England: Cambridge: University Press. Basden, B. H., Basden, D. R., Thomas, R. L., & Souphasith, S. (1998). Memory distortion in group recall. Current Psychology: Developmental, Learning, Personality, Social, 16, 225-246. Berliner, L., & McDougall, J. (1997). Agenda for research: Clinical approaches to recollections of trauma. In J.D. Read & D. S. Lindsay (Eds.), Recollections of trauma: Scientific evidence and clinical practice. NATO ASI series: Series A: Life sciences, 291, 523-529. New York: Plenum Press. Binet, A. (1900). Attention et adaptation. L'annee psychologique, 6, 248-404. Blandon-Gitlin, I., Pezdek, K., Rogers, M., & Brodie, L. (2005). Detecting deception in children: An experimental study on the effect of event familiarity on CBCA ratings. Law and Human Behavior, 29(2), 187-197. Brady, M.S., Poole, D. A., & Warren, A. R. (1999). Young children's responses to yes-no questions: Patterns and problems. Applied Developmental Science, 3, 47-57. Brainerd, C. J., Reyna, V. F., & Poole, D. A. (2000). Fuzzy-trace theory and false memory: Memory theory in the courtroom. In D. F. Bjorkland (Ed.), False memory creation in children and adults: Theory, research, and implications (pp. 93-127). Mahwah, NJ: Lawrence Erlbaum Associates. Bransford, J.D., & Franks, J. J. (1971). The abstraction of linguistic ideas. Cognitive Psychology 2, 331-350. Brewer, W. F., & Treyens, J. C. (1981). Role of schemata in memory for places. Cognitive Psychology, 13, 207-230. Ceci, S. J., & Bruck, M. (1993). The suggestibility of children's recollections: An historical review and synthesis. Psychological Bulletin, 113, 403-439. Ceci, S. J., Ross, D., & Toglia, M. (1987). Suggestibility of children's memory: Psycholegal implications. Journal of Experimental Psychology: General, 116, 38-49. Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58, 17-22. Ebbinghaus, H., & Ruger, H. A. (1913). Memory: A contribution to experimental psychology. New York: Teachers College Press. Ekman, P., & O'Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46, 913-920. Franklin v. Duncan, 884 F.Supp. 1435 (N.D. Cal. 1995), affd, 70 F.3d 75 (9th Cir. 1995). Freyd, J. J., & Gleaves, D. H. (1996). "Remembering" words not presented in lists: Relevance to the current recovered/false memory controversy. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22, 811-813. Gabbert, F., Memon, A., Allan, K., & Wright. D. B. (2004). Say it to my face: Examining the effect of socially encountered misinformation. Legal and Criminal Psychology, 9, 215-227. Gallo, D. A., Roberts, M.A., & Seamon, J. G. (1997). Remembering words not presented in lists: Can we avoid creating false memories? Psychonomic Bulletin & Review, 4, 271-276. Gallo, D. A., & Roediger, H. L. (2003). The effects of associations and aging on illusory recollection. Memory & Cognition, 31, 1036-1044. Garry, M., Manning, C. G., Loftus, E. F., & Sherman, J. (1996). Imagination inflation: Imagining a childhood event inflates confidence that it occurred. Psychonomic Bulletin & Review, 3, 208-214.

NEUSCHATZ ET AL.

258

Garry, M., Sharman, S. J., Wade, K. A., Hunt, M. J., & Smith, P. J. (2001). Imagination inflation is a fact, not an artifact: A reply to Pezdek and Eddy. Memory and Cognition, 29, 719-729. Goff, L. M., & Roediger, H. L., III. (1998). Imagination inflation for action events: Repeated imaginings lead to illusory recollections. Memory and Cognition, 26, 20-33. Heaps, C., & Nash, M. (1999). Individual differences in imagination inflation. Psychonomic Bulletin and Review, 6, 313-318. Hyman, I. E., Husband, T. H., & Billings, J. F. (1995). False memories of childhood experiences. Applied Cognitive Psychology, 90, 181-197. Hyman, I. E., & Pentland, J. (1996). The role of mental imagery in the creation offalse childhood memories. journal of Memory and Language, 35, 101-117. Israel, L., & Schacter, D. L. (1997). Pictorial encoding reduces false recognition of semantic associates. Psychonomic Bulletin and Review, 4, 577-581. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28. Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Review, 88, 67-83. Koriat, A., Goldsmith, M., & Pansky, A. (2000). Toward a psychology of memory accuracy. Annual Review of Psychology, 51, 481-537. Lampinen, J. M., Copeland, S.M., & Neuschatz, J. S. (2001). Recollections of things schematic: Room schemas revisited. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 1211-1222. Lampinen, J. M., Faries, J. M., Neuschatz, J. S., & Toglia, M.P. (2000). Recollections of things schematic: The influence of scripts on recollective experience. Applied Cognitive Psychology, 14, 543-554. Lampinen, J. M., Neuschatz, J. S., & Payne, D. G. (1999). Source attributions and false memories: A test of the demand characteristic account. Psychonomic Bulletin & Review, 6, 130-135. Lindsay, D. S., Hagen, L., & Read, J.D. (2004). True photographs and false memories. Psychological Science, 15, 149-154. Loftus, E. F. (1979). The malleability of human memory. American Scientist, 67, 312-20. Loftus, E. F. (2004). Memories of things unseen. Current Directions in Psychological Science, 13, 145-147. Loftus, E. F., & Bernstein, D. M. (2005). Rich false memories: The royal road to success. In A. F. Healy (Ed.), Experimental cognitive psychology and its applications. Washington, DC: American

Psychological Association.

Loftus, E. F., Coan, J. A., & Pickrell, J. E. (1996). Manufacturing false memories using bits of reality. In L. M. Reder (Ed.), Implicit memory and metacognition. (195-220). Mahwah, NJ: Lawrence Erlbaum Associates. Loftus, E. F., & Greene, E. (1980). Warning: Even memory for faces may be contagious. Law and Human Behavior, 4, 323-334. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4, 19-31. Loftus, E. F., & Pickrell, J. E. (1995). The formation offalse memories. Psychiatric Annals, 25, 720-725. Madill, A., & Holch, P. (2004). A range of memory possibilities: The challenge of the false memory debate for clinicians and researchers. Clinical Psychology & Psychotherapy, 11, 299-310. Mather, M., Henkel, L. A., & Johnson, M. K. (1997). Evaluating the characteristics of false memories: Remember/know judgments and memory characteristics questionnaire compared. Memory & Cognition, 25, 826-837. Mazzoni, G. A. L., Loftus, E. F., & Kirsch, I. (2001). Changing beliefs about implausible autobiographical events. Journal of Experimental Psychology: Applied, 7, 51-59. Mazzoni, G. A. L., Lombardo, P., Malvagia, S., & Loftus, E. F. (1999). Dream interpretation and false beliefs. Professional Psychology: Research and Practice, 30, 45-50. Mazzoni, G., & Memon, A. (2003). Imagination can create false autobiographical memories. Psychological Science, 14, 186-188.

8. FALSE MEMORY RESEARCH

259

McDermott, K. B., & Roediger, H. L. (1998). Attempting to avoid illusory memory: Robust false recognition of associates persists under conditions of explicit warnings and immediate tests. Journal of Memory and Language, 39, 508-520. McDermott, K. B., & Watson, J. M. (2001). The rise and fall of false recall: The impact of presentation duration. Journal of Memory and Language, 39, 160-176. Meade, M. L., & Roediger, H. L., III. (2002). Explorations in the social contagion of memory. Memory and Cognition, 30, 995-1009. Moritz, S., & Woodward T. S. (2002). Memory confidence and false memories in schizophrenia. Journal of Nervous and Mental Disease, 190, 641-643. Munsterberg, H. (1923). On the witness stand. New York: Clark Boardman. Nakamura, G. V., Graesser, A. C., Zimmerman, J. A., & Riha, J. (1985). Script processing in a natural situation. Memory and Cognition, 13, 140-144. Neuschatz, J. S., Lampinen, J. M., Preston, E. L., Hawkins, E. R., & Toglia, M.P. (2002). The effect of memory schemata on memory and the phenomenological experience of naturalistic situations. Applied Cognitive Psychology, 16,687-708. Neuschatz, J. S., Payne, D. G., Lampinen, J. M., & Toglia, M.P. (2001). Assessing the effectiveness of warnings and the phenomenological characteristics of false memories. Memory, 9, 53-71. Norman, K. A., & Schacter, D.L. (1997). False recognition in younger and older adults: Exploring the characteristics of illusory memories. Memory & Cognition, 25, 838-848. Paddock, J. R., Joseph, A. L., Chan, F. M., Terranova, S., Manning, C., & Loftus, E. F. (1998). When guided visualization procedures may backfire: Imagination inflation and predicting individual differences in suggestibility. Applied Cognitive Psychology, 12, S63-S75. Payne, D. G., & Blackwell, J. M. (1998). Truth in memory: Caveat emptor. InS. J. Lynn & K. M. McConkey (Eds), Truth in Memory (pp. 32-61). New York: Guilford Press. Payne, D. G., Elie, C. J., Blackwell, J. M., & Neuschatz, J. S. (1996). Memory illusions: Recalling, recollecting, and recognizing events that never occurred. Journal of Memory and Language, 35, 261-285. Pesta, B. J., Murphy, M.D., & Sanders, R. E. (2001). Are emotionally charged lures immune to false memory? Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 328-338. Pezdek, K., Finger, K., & Hodge, D. (1997). Planting false childhood memories: The role of event plausibility. Psychological Science, 8, 437-441. Pezdek, K., & Hodge, D. (1999). Planting false childhood memories in children: The role of event plausibility. Child Development, 70, 887-895. Pezdek, K., & Roe, C. (1997). The suggestibility of children's memory for being touched: Planting, erasing, and changing memories. Law and Human Behavior, 21,95-106. Pezdek, K., Whetstone, T., & Reynolds, K. (1989). Memory for real-world scenes: The role of consistency with schema expectation. Journal of Experimental Psychology: Learning, Memory, & Cognition, 15, 587-595. Porter, S., Yuille, J. C., & Lehman, D. R. (1999). The nature of real, implanted, and fabricated memories for emotional childhood events: Implications for the recovered memory debate. Law and Human Behavior, 23, 517-537. Pynoos, R. S., & Nader, K. (1989). Children's memory and proximity to violence. Journal of the American Academy of Child & Adolescent Psychiatry, 28, 236-241. Read, J. D. (1996). From passing thoughts to false memories in 2 minutes: Confusing real and illusory events. Psychonomic Bulletin & Review, 3, 105-111. Roediger, H. L. (1996). Memory illusions. Journal of Memory and Language, 35, 76-100. Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. Roediger, H. L., McDermott, K. B., Pisoni, D. B., & Gallo, D. A. (2004). Illusory recollection of voices. Memory, 12, 586-602.

260

NEUSCHATZ ET AL.

Roediger, H. L., III, Meade, M. L., & Bergman, E. T. (2001). The social contagion of memory. Psychonomic Bulletin and Review, 8, 365-3 71. Roediger, H. L., Watson, J. M., McDermott, K. B., & Gallo, D. A. (2001). Factors that determine false recall. A multiple regression analysis. Psychonomic Bulletin & Review, 8, 385-407. Schacter, D. S., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger and older adults: The distinctiveness heuristic. Journal of Memory and Language, 40, 1-24. Schacter, D. S., Verfaellie, M., & Pradere, D. (1996). The neuropsychology of memory illusions: false recall and recognition in amnesiac patients. Journal of Memory and Language, 35, 319-334. Schooler, J. W. (1996). Seeking the core: The issues and evidence surrounded recovered accounts of sexual trauma. In K. Pezdek & W. P. Banks (Eds), The recovered memory/false memory debate (pp. 279-296). Thousand Oaks, CA: Academic Press. Schooler, J. W. (1997). Taking the middle line: Can we accommodate both fabricated andrecovered memories of sexual abuse? In M. A. Conway (Ed.), Recovered memories and false memories (pp. 251-292). London: London University Press. Schooler, J. W. (2001). Discovering memories of abuse in light of meta-awareness. Journal of Aggression, Maltreatment & Trauma, 42, 105-136. Seamon, J. G., Luo, C. R., & Kopecky, J. J. (2002). Are false memories more difficult to forget than accurate memories? The effect of retention interval on recall and recognition. Memory & Cognition, 30, 1054-1064. Spanos, N. P., Burgess, C. A., Burgess, M. F., Samuels, C., & Blois, W. 0. (1999). Creating false memories of infancy with hypnotic and non-hypnotic procedures. Applied Cognitive Psychology, 13, 201-218. State of Washington v. Ingram, No. 88-100752-1, Superior Ct., Olympia, Washington. Stern, W. (1910). Abstracts oflectures on the psychology of testimony and on the study of individuality. American Journal of Psychology, 21, 270-282. Toglia, M.P. (1996). Repressed memories: Lost and found? InK. Pezdek and W. P. Banks (Eds.), The recovered memory/false memory debate (pp. 313-323). San Diego: Academic Press. Toglia, M.P., Neuschatz, J. S., & Goodwin, K. A. (1999). Recall accuracy and illusory memories: When more is less. Memory, 7, 233-256. Tulving, E. (1985). Memory and consciousness. Canadian Psychologist, 26, 1-12. Underwood, B. J. (1965). False recognition produced by implicit verbal responses. Journal of Experimental Psychology, 70, 122-129. Vrij, A. (2000). Detecting lies and deceit: The psychology of lying and the implications for professional practice. London: Wiley Vrij, A., Akehurst, L., Soukara, S., & Bull, R. (2002). Will the truth come out? The effect of deception, age, status, coaching, and social skills on CBCA scores. Law and Human Behavior, 26(3), 261-283. Wade, K. A., Garry, M., Read, J.D., & Lindsay, D. S. (2002). A picture is worth a thousand lies: Using false photographs to create false childhood memories. Psychonomic Bulletin and Review, 9, 597-603. Wagenaar, W. A., & Groeneweg, J. (1990). The memory of concentration camp survivors. Applied Cognitive Psychology, 4, 77-87. Watson, J. M., McDermott, K. B., & Balota, D. A. (2004). Attempting to avoid false memories in the Deese/Roediger-McDermott paradigm: Assessing the combined influence of practice and warnings in young and old adults. Memory & Cognition, 32, 135-141. Wright, B. D., Self, G., & Justice, C. (2000). Memory conformity: Exploring misinformation effects when presented by another person. British Journal of Psychology, 91, 189-202.

g Psychological Impairment, Eyewitness Testimony, and False Memories: Individual Differences Sal A. Soraci Tufts University

Michael T. carlin University of Massachusetts Medical School

J. Don Read Simon Fraser University, British Columbia

Terri Krangel Pogoda, Yvonne Wakeford, Sarah Cavanagh, and Lisa Shin Tufts University

This chapter focuses on memory strengths and deficiencies in individuals with intellectual disabilities (i.e., mental retardation, autism) or various psychological impairments (i.e., schizophrenia, Alzheimer's disease, dissociative disorder, substance abuse, trauma). For each population, we review empirical work on memory and false memory, and discuss implications for eyewitness testimony. Because many of these disorders have been understudied, we also emphasize methodological limitations and suggestions for future research. These disorders generally are marked by complicated patterns of memory strengths and weaknesses that render decisions regarding reliability and validity of memory reports quite difficult. Understanding the ability of these individuals to serve as witnesses is important, because many of these disorders are associated with higher rates of criminal contact, perpetration, and victimization. Although we assume that cognitive impairments resulting from the short-term abuse of substances like alcohol and other 261

262

SORACI ET AL.

drugs are temporary, unlike the chronic nature of other disorders described herein, an understanding of these impairments is nonetheless relevant to legal procedure. Equal justice in legal contexts for all of these populations requires greater understanding of their cognitive abilities and limitations. Though most studies of eyewitness testimony and accuracy have used individuals without cognitive impairments, focus on such groups is important for theory and application. From a theoretical perspective, results from studies involving individuals with cognitive impairments can provide important information regarding the neurological bases for normal memory functioning and particular patterns of memory deficit. This provides a richer understanding of memory, the particular disorder being studied, and potential avenues for remediation of the observed memory deficiencies. From an applied perspective, knowledge of patterns of memory strengths and weaknesses associated with particular cognitive impairments will allow for an informed decision about the individual's ability to provide reliable and valid recollections of past events. For many of the disorders, this is a complex decision because the individuals often provide reports that are at least partially accurate, but are likely to include some inaccuracies and omissions. Therefore, it is imperative that attention be given to the types of information being remembered, the manner in which the memory is elicited, and how these interact with the memory abilities of typical individuals with the cognitive disorder. This chapter reviews the most recent research on memory accuracy for each disorder and discusses concerns regarding the accuracy of eyewitness testimony by individuals with these impairments.

MENTAL RETARDATION

Mental retardation is a disorder marked by limitations in most cognitive abilities, particularly memory. Early work focused on identifying memory deficiencies (e.g., Belmont & Butterfield, 1971; Borkowski, 1985; Bray et al., 1998; Ellis, 1970; Turnure, 1985), inadequate use of strategies for remembering (e.g., Brown, 1972), or failure to organize information (e.g., Spitz, 1966) to promote memory. More recent efforts in the field have focused on methods for promoting memory accuracy by structuring the information to be remembered in ways that promote recall (e.g., Carlin, Soraci, Dennis, Chechile, & Loiselle, 2001; Carlin et al., 2005) or by providing situational supports that can be used to promote memory (e.g., Bray et al., 1998). An important consideration in reviewing this literature and that to be discussed subsequently is the comparison group(s) used. There are two types of comparison groups that are commonly used in the research on memory and mental retardation, a chronological age (CA)-matched group or a mental age (MA)-matched group. In comparison to CA-matched comparisons, individuals with mental retardation almost always perform much more poorly on tests of recall and recognition. The MA-matched comparisons are included to determine if the memory deficiency is due to delayed development, in which case performance would be similar to that of the MA-matched group, or to atypical development due to a neurological or other difference between the groups. Comparisons to MA-matched controls have provided a more complicated set of results, with individ-

9.

PSYCHOLOGICAL IMPAIRMENT

263

uals with mental retardation sometimes performing better than, worse than, or equiva, lent to the MA,matched comparisons. Likely sources of this variability across studies include inexact MA matching, level of mental retardation (McDaniel, Foster, Compton, & Courtney, 1998), particular mem, ory skill assessed (e.g., implicit vs. explicit memory; see Vicari, Bellucci, & Carlesimo, 2000), and type of material to be remembered. For example, for the latter variable, Mar, tin, West, Cull, and Adams (2000) showed that, on the subtests of the Rivermead Be, havioural Memory Test, adults with mild mental retardation performed better on visual memory tests (pictures and faces) than on tests involving verbal memory (stories and names). For the present purposes, these are important issues for determining compe, tency to provide testimony in a legal context and for evaluating the validity of the mem, ory report.

The DRM False Memory Paradigm The only study performed to date that used the DRM paradigm (Deese, 1959; Roediger & McDermott, 1995) to assess memory accuracy in individuals with mental retardation is one recently completed by our research group (Sullivan, Carlin, Wakeford, Hasel, & Jakway, 2005). In this study, we assessed memory for old, critical, and novel pictures in three groups of participants: individuals with mild mental retardation, individuals matched for CA, and individuals matched for MA. The memory task was a modified recognition task in which the experimenter asked if the participant had seen a particular ").The partie, object from the previous acquisition list (e.g., "Did you see a ipants were not re,presented with the pictures at test. Results indicated that individuals with mental retardation performed much more poorly than CA,matched comparisons. The individuals with mental retardation reported having seen significantly fewer old items (96% vs. 82%) and had much higher false report rates, for both critical (20% vs. 56%) and novel (1% vs. 32%) items. The individuals with mental retardation under, reported previously experienced items and overreported items that they had actually not seen. The latter finding likely is due, at least in part, to a tendency to respond affir, matively when in doubt. This type of positive response bias, or acquiescence, is common in the literature on mental retardation (Finlay & Lyons, 2002). Relative to the MA, matched group, the individuals with mental retardation again had higher incidences of false reports for both critical (41% vs. 56%) and novel (11% vs. 31 %) items, but the groups were similar with regard to accuracy for old items. When looking at a measure of overall accuracy (i.e., % Correct Reports - % False Reports), the results were 92%, 71%, and 46% for the CA,matched comparisons, MA,matched comparisons, and group with mental retardation, respectively. Though the group with mental retardation per, formed similarly to the MA,matched group for correct recognition, the high false report rates for critical and novel items rendered the group with mental retardation signifi, candy less accurate overall. Overall, basic memory research with individuals with mental retardation has dem, onstrated that they tend to report fewer previously encountered items than CA, matched control groups and have higher rates of false reporting. However, memory

264

SORACI ET AL.

accuracy for individuals with mental retardation can be quite variable, depending on level of mental retardation, etiology of mental retardation (e.g., Carlesimo et al., 1996), type of material to be remembered, and availability of memory supports in the environment. These findings demonstrate a pattern of performance on memory tests that is typified by diminished reporting rates, increased false reports, and a tendency to acquiesce and respond affirmatively to questions. The next section reviews research on the generality of these findings to contexts involving eyewitness testimony.

Eyewitness Testimony Evaluating the validity of eyewitness testimony provided by an individual with mental retardation is problematic because of concerns with memory accuracy and interactions between accuracy, questioning style (free recall, open-ended questions, closed-ended questions), and level of mental retardation (Kebbell & Hatton, 1999). In this section, several recent studies are reviewed that assessed these variables and interactions and provide a summary of the best research to date on this topic. Michel, Gordon, Ornstein, and Simpson (2000) performed simulated health checks on children (9-14 years old) with and without (MA-matched and CA-matched groups) mild mental retardation. The health check included 15 items (e.g., check reflexes, measure height) performed by the "doctor." Children's memories were assessed with multiple forms of test. Memory assessment began with a free recall phase and progressed through increasingly specific and closed-ended questioning. Memory accuracy was assessed for percentage of healthcheck features remembered and for false reports of health-check procedures that actually were not performed. For free recall, the group with mental retardation and the MA-matched group performed similarly, recalling about 55% of the health check items immediately after the checkup and 35-40% at a 6-week follow-up. These levels of recall were significantly lower than those of theCA-matched group (80% and 70% correct on the tests). Furthermore, rates of reporting false information in free recall were very low for all groups. When yes-no questioning was used (e.g., "Did the doctor check your eyes?"), percentage correct responses for the three groups exceeded 90% at the immediate interview and 80% at the 6-week follow-up. Thus, much more information was accessed with this form of closed-ended questioning. However, false reports of health-check procedures that did not occur were reported with the use of this form of questioning, particularly by the group with mental retardation and the MA-matched group. Furthermore, the false reports increased dramatically from the immediate interview (10%) to the 6-week interview (about 25%). Overall, the group with mental retardation and the MA-matched comparisons performed similarly, and the CA-matched group was superior to both lowMA groups in terms of completeness of report and susceptibility to false reporting. Henry and Gudjonsson (2003) assessed eyewitness accuracy in groups of children (11-12 years old) with mild and moderate mental retardation, and compared their performances to MA- and CA-matched controls. Memory was assessed for an actual event that was experienced by the children. The event was an unexpected visit to their classroom by a female speaker and male cameraman. Students were asked to recall details of

9.

PSYCHOLOGICAL IMPAIRMENT

265

the event the next day and again 2 weeks later. Memory was assessed with the use of an initial free recall test and then increasingly specific questioning (general open-ended questions, specific open-ended questions, and specific yes-no questions). Free recall results showed that the CA-matched group reported the most information, followed by the group with mild mental retardation. Interestingly, the group with mild mental retardation reported significantly more information than the MA-matched comparison group and the group with moderate mental retardation. For responses to misleading questions, a measure of suggestibility, the moderate mental retardation group scored higher (i.e., greater suggestibility) than the group with mild mental retardation and the CA-matched group. Perhaps the most relevant finding, however, concerned "eyewitness shift." This is a measure of the respondents' likelihood of changing answers to yes/no questioning from the first to the second interview. This measure indicated that participants with mental retardation, mild or moderate, were more likely to shift responses (approximately 21 %) than CA-matched (11 %) or MA-matched (15%) comparisons. Thus individuals with mental retardation seem to be more susceptible to misleading questions and more likely to change responses when questioned repeatedly. Because typical contexts within the justice system involve multiple interviews, this is an area of special concern with this population.

summary The pattern of results from the studies reviewed demonstrates that individuals with mental retardation can provide accurate accounts of witnessed events, particularly in a free recall situation that immediately follows the events. In this circumstance, the information they report is likely to be accurate though incomplete relative to age-matched comparisons (Kebbell & Hatton, 1999). As time passes and the questioning becomes more closed-ended, accuracy of memory reports tends to decrease (Milne, 1999). Particular concerns are tendencies to respond affirmatively when asked about events that did not occur and to change responses under repeated questioning (e.g., Clare & Gudjonsson, 1993; Henry & Gudjonsson, 1999). It is important when evaluating memory accuracy that level of mental retardation be considered. Research has consistently shown that individuals with mild mental retardation provide more complete and accurate descriptions than do individuals with moderate or severe mental retardation. As IQ decreases, so does accuracy of reporting details of past events. Future research should focus on (a) methods of questioning that elicit the most complete and accurate reports from individuals with mental retardation (see Dent, 1992), (b) a broader range of age groups and etiologies of mental retardation, and (c) issues concerning competency to understand the oath and testify in a court proceeding (e.g., Gudjonsson, Murphy, & Clare, 2000). AUTISM

Autism is a disorder similar to mental retardation in that there is a wide range of ability levels encompassed in this population. Most of the research on memory accuracy has

266

SORACI ET AL.

focused on high-functioning individuals with autism, and this review reflects that bias. The literature on autism, false memory, and eyewitness testimony is not as well developed as that on mental retardation and memory. Therefore, we selected representative publications for this review that address issues most relevant for evaluating the reliability and validity of reports of past events from individuals with this diagnosis. Early research on memory in individuals with autism focused on the relationship between autism and amnesic disorders (e.g., Boucher, 1981; Boucher & Warrington, 1976; Renner, Klinger, & Klinger, 2000). However, as the field has developed this hypothesis has been refuted, and current focus is on the relationship between memory abilities and executive functioning (e.g., Benneto, Pennington, & Rogers, 1996; Hala, Rasmussen, & Henderson, 2005) or memory functioning and failure to use organizing strategies (e.g., Minshew & Goldstein, 2001; Williams, Goldstein, & Minshew, 2005) or semantic structure (e.g.,Toichi & Kamio, 2003; Tager-Flusberg, 1991). It has become clear that there is no across-the-board memory deficit in individuals with high-functioning autism. On many tests of memory they perform comparably to CA-matched comparisons. However, as more demand is placed on executive resources, memory problems emerge in this population. Minshew and Goldstein (2001) compared individuals with high-functioning autism with a group of CA-matched comparisons on a series of memory tasks. Groups performed comparably on tests of short-term memory (10-s and 20-s delays) for words and paired-associates learning, but differences emerged for recall of story elements, geometrical figures, and mazes. The authors concluded that memory problems become apparent in individuals with autism when stimulus complexity increases (e.g., geometric figures, stories) or when semantic relatedness could be used (list recall). Bennetto et al. (1996) had taken a similar approach to the study of memory functioning in this population and showed equivalence across groups for tests of short- and long-term recognition memory and cued recall. However, there were significant differences between the groups with and without autism on tests involving free recall and memory for temporal order and source. This pattern of results, in conjunction with a finding of group differences on several measures of executive functioning, led the authors to the conclusion that memory deficits in individuals with autism are similar to those of other groups with executive processing deficiencies. This executive functioning deficit would explain the lack of use of semantic structure in list recall and reductions in memory performance as task or stimulus complexity increases. A series of studies by Toichi and Kamio (2001, 2002, 2003) assessed basic memory functioning and semantic structure in individuals with autism. Toichi and Kamio (2001) compared a group with high-functioning autism to a group with learning disabilities who were approximately matched on level of functioning. They used a word-completion priming task for both related and unrelated items. Results showed an advantage for both groups in the related condition, but no group differences in semantic priming. This contrasted with other studies showing semantic processing deficits in individuals with autism when episodic memory tasks were employed, and led them to the hypothesis that the focal deficit may be in the link between semantic processing and long-term memory. Toichi and Kamio (2002) tested this link in a study assessing levels of processing (LOP)

9.

PSYCHOLOGICAL IMPAIRMENT

267

effects in this population. The three tasks used to represent the differing levels of processing were perceptual (what font was used?), phonological (is the pronunciation simi?), and semantic (e.g., is it a drink?). Overall correct recognition rates did lar to not differ between groups. However, the CA-matched comparisons demonstrated the expected LOP effect with recognition accuracy higher for the semantic condition than the other two conditions, whereas the group with autism performed equivalently on all three tasks. Interestingly, the group with autism had higher recognition scores for the graphic condition than the control group. This indicates enhanced rote memory for visual characteristics of words in this group. Overall, the pattern of results suggests that semantic memory does not influence episodic memory in individuals with autism as it does in other populations.

The DRM False Memory Paradigm The only study to use the DRM paradigm with individuals with autism is that of Beversdorf et al. (2000). Given the findings cited above regarding semantic structure and the link between semantic and episodic memory, it would be expected that individuals with autism would show a pattern of results different from that of a control group. Specifically, if there is a disruption or disconnect between these systems, then individuals with autism would not be expected to show memory effects due to semantic relatedness or spreading activation as do individuals without autism (e.g., Roediger, Watson, McDermott, & Gallo, 2001). Thus, the prediction would be that individuals with autism would have lower rates of false memories for critical items in this paradigm. Beversdorf et al. (2000) compared individuals with autism with a CA-matched control group on a standard DRM task. Lists were presented via audiotape, and recognition was tested with a questioning technique (e.g., Did you hear "thimble"?) after each list. The authors used d' as a measure of ability to discriminate true and false items. Results showed that the hypothesis was supported; individuals with autism were significantly better able to discriminate the two types of items. Thus, their overall accuracy was better than that for the CA-matched group. Additional analyses demonstrated that the key difference between the groups was that individuals with autism had much lower rates of false memory for critical items. These results support the general hypothesis that individuals with autism do not use semantic structure to support episodic memory. Though a decrease in false reports in this paradigm may seem like a strength for this population, it appears be an artifact of an underlying memory problem. Failure to use contextual information (i.e., the semantic structure of the lists) to support memory leads to better performance in this form of test but is cause for concern in most forensic contexts in which an individual is asked to recall information about a complex and multifaceted stimulus, a scene.

Eyewitness Testimony Though no specific studies of eyewitness testimony have been published for this population, several studies closely related to such situations provide useful information about

268

SORACI ET AL.

the likely outcome for the types of stimuli encountered in those circumstances. The first set of studies in this section focuses on studies of source monitoring in autism, and the second set focuses on memory for faces and scene information. Farrant, Blades, and Boucher (1998) assessed reality monitoring in individuals with autism. Reality monitoring entails differentiating whether the source of some stimulation is oneself or another person. In this experiment, they compared memory for words that were spoken by the participant (internal source) or by an experimenter (external source). The group with autism was compared with a CA-matched control group and a control group with mental retardation. A surprise recognition test required the participants to indicate whether the word was on the previous memory list, and if so, whether it had been spoken by them or the experimenter. Results showed no differences across groups in reality monitoring. Millward, Powell, Messer, and Jordan (2000) compared memory for events experienced by oneself or another person. The context of the experiment was a 25-minute walk in the vicinity of the participant's school. Walks occurred with the experimenter or with the experimenter and another participant (the companion). Mter each of the two walks, participants were asked questions about their experiences, or the experiences of their companion, during the walk. Questioning took two forms: free recall (open questions) and cued recall (leading questions). Surprisingly, results showed that individuals with autism had better memory for events experienced by others than for events they experienced themselves. This is opposite of the typical pattern observed. Furthermore, memory scores for self-experienced events were significantly lower for the group with autism for open-ended and leading questions. A subsequent study showed that this same pattern held true when performance was compared with that of a MA-matched group with learning disabilities. Thus, this decreased memory for personally experienced events seems specific to autism and not simply lower mental age. Hala, Rasmussen, and Henderson (2005) recently published a study attempting to resolve the ambiguities in the literature regarding source monitoring in autism. They compared performances of individuals with autism with that of a MA-matched group on tests of reality monitoring, internal-source monitoring, and external-source monitoring. The tasks required participants to differentiate between self- and experimentergenerated words (reality monitoring), words spoken by two different experimenters (external monitoring), or words that were said by the participant or simply thought by the participant (internal-source monitoring). There were no differences between groups for recognition accuracy. The two groups were equally able to distinguish old words from new words on the recognition tests. However, when asked to identify the source of the words, the MA-matched group outperformed the group with autism for all three sourcemonitoring conditions. Furthermore, the group with autism performed above chance for the reality-monitoring task only and performed most poorly in the internal monitoring task. Though the latter comparison was complicated by a floor effect, it was consistent with the finding of Millward et al. (2000) that individuals with autism have poorer memory for events regarding themselves than for events regarding other people. Finally, Williams, Goldstein, and Minshew (2005) assessed auditory and visual memory in individuals with high-functioning autism. There were no differences on the auditory tasks that required recall of story details or unrelated word pairs. However, sig-

9.

PSYCHOLOGICAL IMPAIRMENT

269

nificant differences between the group with autism and the CA-matched comparisons were apparent on the visual tasks that are most similar to circumstances encountered in legal contexts. One visual task required participants to view a set of photographs and later select these faces from a set including an equal number of unviewed faces. Participants with autism were less likely to correctly recognize previously viewed faces. Unfortunately, no information was provided to specify whether differences were due to lower correct recognition rates or increased false recognition rates. The second relevant visually based task (Family Pictures) required participants torecall details about the location and activities of family members in a previously displayed scene. Again, individuals with autism scored significantly lower than age-matched peers. Thus participants with autism scored lower than comparison groups on two tasks that most closely approximated the memory demands of typical environments about which they would be questioned in legal contexts. When asked to recognize a photograph of previously seen individuals, the group with autism performed more poorly. When asked to recall information about the location and actions of people in a scene, they again scored lower than controls. This poor recall does not seem to indicate a generally poorer memory, because these same individuals performed as well as controls on other tests of memory.

Summary Whereas individuals with mental retardation tend to perform poorly in general on tests of memory, individuals with autism demonstrate a more complicated pattern of abilities and limitations. The literature reveals circumstances in which individuals with autism perform as well or even better than age-matched comparisons (e.g., the DRM false memory paradigm), and other circumstances in which they perform poorly relative to comparison groups. Many researchers have hypothesized that the pattern of deficits is consistent with disruptions of executive functioning in this population (e.g., Bennetto et al., 1996). This executive functioning deficit does not have an impact on memory tasks involving simple stimuli or designed memory supports (e.g., cued recall). However, with more complex stimuli (e.g., faces, scenes) or need for use of contextual information (e.g., semantic relatedness), performances of individuals with autism are typically worse than performances of control groups. Furthermore, research showing poorer memory for faces, details of scenes, and self-experienced events (Millward et al., 2000) raise serious concerns about their ability to serve as reliable eyewitnesses in legal proceedings. Clearly, however, there is a need for systematic research on eyewitness testimony in this population. No studies have used methodologies that closely match typical eyewitness contexts. Furthermore, most of the studies in this literature have focused on high-functioning individuals with autism in a fairly restricted age range. Future studies should focus on these variables, questioning techniques, and how they affect memory performance and eyewitness accuracy in this population. SCHIZOPHRENIA

Schizophrenia is considered to be one of the most severe psychiatric illnesses in terms of both its deleterious effects on individuals' functioning and its costs to society. The

270

SORACI ET AL.

symptoms of this disorder are frequently categorized into positive and negative manifestations. The positive symptoms reflect an excess of normal processes, and the negative symptoms are due to a reduction or loss of normal processes. The most prevalent positive symptoms are hallucinations, delusions, and disorganized speech. The negative symptoms include flattened affect, reduced or absent speech, and a diminution of goaldirected behavior. Memory impairments and false memories are associated with some of the positive symptomatology, especially delusions. In this section, we review some of the current literature on memory impairments in individuals with schizophrenia. We then examine source monitoring deficits in schizophrenic patients, review false memory research involving individuals with schizophrenia, and discuss implications for the role of memory distortions in schizophrenia symptomatology.

Schizophrenia and Memory Impaired memory performance in individuals with schizophrenia is a well-researched and well-replicated finding. A meta-analysis of 70 studies published between 1975 and 1998 (Aleman, Hijman, de Haan, & Kahn, 1999) found strong empirical support for the presence of memory dysfunction in individuals with schizophrenia. They found stronger evidence for disruption of memory assessed through recall rather than recognition, although impairments in both have been reported frequently. Findings of memory dysfunction in schizophrenic individuals were not significantly related to age, education, medication effects, duration or severity of illness, or level of positive symptoms; however, the authors did find a small but significant positive relationship between level of negative symptoms and memory impairment. Moritz, Heeren, Andresen, and Krausz (2001) demonstrated that this relationship is unlikely to be due to distractibility or other attentional dysfunction. Brebion, Amador, Smith, and Gorman (1997) assessed the roles of encoding, storage, and retrieval in the memory decrements of schizophrenic patients for verbal materials. They compared recall and recognition indices for individuals with and without schizophrenia in multiple conditions. Participants were presented with nonorganizable and organizable lists and were tested immediately or after 1 minute. Organizable lists included 16 words with 4 words from each of four categories. Nonorganizable lists included 16 words that were not associated with each other. Results showed that schizophrenic patients did not benefit from the list organization, whereas the controls did show a benefit. Furthermore, the participants with schizophrenia showed similar effects of serial position for the organized and nonorganized lists. Brebion et al. concluded that the schizophrenic participants did not use the list organization to benefit memory, but rather used a sequential strategy for both list types. No group differences emerged across delay conditions (immediate versus 1 minute). In total, the results of this study indicated that the memory deficits of patients with schizophrenia were primarily due to encoding differences across groups. The encoding deficit hypothesis, particularly one involving deficient elaborative processing at encoding, is preferred by many investigators (e.g., Danian, Rizzo, & Bruant, 1999) of memory in schizophrenic patients.

9.

PSYCHOLOGICAL IMPAIRMENT

271

Schizophrenia and Source Monitoring Reality monitoring refers to the ability to distinguish between information that was generated internally and information that was perceived in the external environment. Source monitoring is a form of reality monitoring and refers to the process of attributing the origin of memories, knowledge, and beliefs to specific sources (Johnson, Hashtroudi, & Lindsay, 1993). An example of a source monitoring error is failing to remember whether you read a fact in an article or were told the fact by another person. Accurate source monitoring relies on an intricate balance of numerous cognitive processes, such as encoding information correctly, consolidating it with previous memories, reactivating and retrieving those memories, and evaluating them for accuracy and source (Johnson & Raye, 1998). One or more of these steps can become corrupted and contribute to the formation of false memories. Researchers have suggested that some of the positive symptoms of schizophrenia could be explained by problems in internal/external source monitoring. In this framework, hallucinations are internally generated events (e.g., inner speech) that are mistakenly attributed to an external source (for a review, see Morrison & Haddock, 1997). Research examining the integrity of source monitoring in individuals with schizophrenia has generally found evidence for an external attributional bias. That is, in cases of ambiguity, schizophrenia patients are likely to attribute information to external sources. Vinogradov et al. (1997) found that individuals with schizophrenia, compared with healthy individuals, tended to attribute new and self-generated material to an external source. Furthermore, they found that this attributional bias was stable over a 2-year period and that it was independent of fluctuations in medication and psychiatric status, at least for a subset of their subjects. Morrison and Haddock (1997) investigated both immediate and delayed source monitoring for positive, negative, and neutral words in individuals with schizophrenia and healthy controls. Their data indicated that in the immediate condition, individuals experiencing auditory hallucinations demonstrated an external attributional bias. This pattern of responses was significantly different from those of both healthy controls and schizophrenic individuals without auditory hallucinations. There were no group differences in the delayed condition. These results suggest that auditory hallucinations may involve a disruption in the attribution of thoughts or perceptions in the present moment, but that subsequent memories for the source of such attributions are unaffected. Valence of the stimuli did have an effect on source monitoring for schizophrenic participants with and without auditory hallucinations. Errors mistakenly attributing source to the experimenters were greater for positive stimuli, and errors mistakenly attributing source to the participants themselves were greater for negative stimuli. Keefe, Arnold, Bayen, and Harvey (1999) further examined source-monitoring deficits in individuals with schizophrenia. Their results indicated that schizophrenic patients exhibited deficits in discriminating between information obtained from two external sources, two internal sources, and one external and one internal source. Therefore, it appears that source-monitoring difficulties on the part of patients with schizophrenia may be more complex and global than a simple external attributional bias.

272

SORACI ET AL.

However, in agreement with the previously reviewed literature, they found that when individuals with schizophrenia were uncertain about the source of material, they demonstrated an external attributional bias. Moritz, Woodward, and Ruff (2003) found that individuals with schizophrenia not only had poorer source memory for internally and externally generated words, but that they also demonstrated higher confidence than controls for their incorrect responses. The cumulative results of the research on source monitoring and schizophrenia suggest that individuals with schizophrenia tend to exhibit an external attributional bias, such that ambiguous source information is attributed to an external source. These results lend credence to the theory that positive symptoms, such as auditory hallucinations, may be the result of incorrectly attributing internally generated thoughts or speech to the external environment. The exact nature of this bias may be more complicated than once thought. Immediate versus delayed testing appears to influence results. Patients also tend to exhibit impairments in discriminating between two external sources, compared with healthy individuals. However, given the importance of source monitoring in the formation of accurate memories, this research suggests that one must consider possible impairments of source monitoring when evaluating reported memories of individuals with schizophrenia.

Schizophrenia and False Memory Given that individuals with schizophrenia reliably demonstrate decrements in performance on a wide range of memory tests, and that the positive symptoms of the disorder implicate problems with reality testing, it is plausible that such individuals may demonstrate a heightened susceptibility to the formation of false memories. Indeed, this area of research has recently garnered significant attention. As reviewed below, most of the studies of false memory in individuals with schizophrenia have examined intrusions on tests of recall and false positives on tests of recognition memory. Danian et al. (1999) tested their hypothesis that individuals with schizophrenia demonstrate episodic memory problems due to "defective binding between separate aspects of events" (p. 640). Participants were shown pairings of two objects. The pairings of objects were made by the experimenter ("watch") or by the participant ("perform"). On a subsequent recognition test, individuals with schizophrenia had lower hit rates to old pairings and increased false positives to novel pairings. Interestingly, patients reported significantly more "know" judgments than "remember" judgments for correct correction and were less able to correctly report the source (self or experimenter) of the pairings. Furthermore, recognition performance for patients was much worse in the watch condition than in the perform condition. This study, therefore, supported the previous findings of impaired recognition for objects and sources in individuals with schizophrenia. Danian et al. concluded that schizophrenia patients are less able to form cohesive memories of events with their sources and attributed the memory deficiency to an encoding deficit. Weiss, Dodson, Goff, Schacter, and Heckers (2002) examined the ability of individuals with schizophrenia to suppress false recognition of nonpresented items. Their re-

9.

PSYCHOLOGICAL IMPAIRMENT

273

suits indicated that individuals with schizophrenia were more likely than healthy con~ trois to commit false positive errors on the recognition task. However, as seen in healthy populations, picture encoding decreased the rates of false recognition relative to words. This finding suggests that the individuals with schizophrenia were able to suppress false identification in conditions that provided more salient perceptual features. This result was against prediction and suggests that individuals with schizophrenia are able to rely on what is called the distinctiveness heuristic; that is, they are able to utilize vivid per~ ceptual details present at encoding to distinguish between veridical and false memories. In a study implementing a recognition task for presented and nonpresented words, Moritz and Woodward (2002) found that schizophrenia patients committed significantly more recognition and source attribution errors than controls, and that patients rated their incorrect responses with significantly higher confidence than controls. The authors suggest that this finding may indicate that inflated confidence in erroneous judgments may underlie the development of false beliefs in the phenomenology of schizophrenia. Moritz, Woodward, Cuttler, Whitman, and Watson (2004) investigated the perfor~ mance of schizophrenic patients and healthy controls on the DRM paradigm, using the word lists reported in Roediger and McDermott's (1995) study. False memory errors occurred with equal rates in schizophrenic individuals and healthy controls. However, within the schizophrenia patients only, there was a very high positive correlation (r = . 76) between correct recognition for presented items and false recognition of the critical lure. The authors suggest that the healthy controls demonstrated false memory due to spreading activation during the process of encoding, whereas the individuals with schizo~ phrenia were relying on a familiarity or gist~based form of memory upon recognition. They attribute this difference across populations to poorer memory for item~specific information in individuals with schizophrenia. In accord with their previous study (Moritz & Woodward, 2002), the authors also found that the patients with schizophre~ nia demonstrated significantly higher confidence for their incorrect responses than did controls. In another study using the DRM paradigm, Elvevag, Fisher, Weickert, Weinberger, and Goldberg (2004) assessed recall and recognition accuracy and found that individu~ als with schizophrenia had lower hit rates and lower false~positive rates for critical lures. However, the individuals with a diagnosis of schizophrenia had higher rates of unrelated intrusions, thus rendering overall false report rates equal to those of the comparison group. Elvevag et al. also reported a positive correlation between false alarm rate and positive symptoms of schizophrenia. The studies of Moritz et al. (2004) and Elvevag et al. (2004) that used the DRM paradigm demonstrate approximately equal probabili~ ties of false reporting in individuals with schizophrenia.

Conclusions and Future Research The research literature on schizophrenia and memory suggests that individuals with schizophrenia have impairments in memory, and that these impairments may particu~ larly involve difficulty in encoding and source monitoring. The data suggest that indi~ viduals with schizophrenia rely more heavily on familiarity or gist when evaluating the

274

SORACI ET AL.

accuracy of a memory because they do not remember item-specific details as well as individuals without schizophrenia. Despite this seeming overreliance on gist memory and poorer memory for source information, schizophrenia patients do not demonstrate significantly higher rates of false memories on many recognition tests. Identifying the bases of this counterintuitive finding should be a primary focus of research. This result may be due to factors such as failure or diminished ability to use list organization (e.g., Brebion et al., 1997) or reduced semantic organization (e.g., Goldberg, Dodge, Aloia, Egan, & Weinberger, 2000). Further research on false memory in schizophrenia should use paradigms that more directly examine false memory, such as the DRM paradigm (e.g., Moritz et al., 2004). In addition, because the vast majority of the research in this area has been performed with words as stimuli, future research should use other types of stimuli. By researching the cognitive underpinnings of false memory in patients with schizophrenia, techniques may be developed to help these individuals monitor the source of their memories.

ALZHEIMER'S DISEASE

Perhaps the most basic feature of the Alzheimer's diagnosis is a deterioration of memory abilities. Even from the initial stages of the disease, memory impairments in episodic memory can be quite severe (e.g., Dudas, Clague, Thompson, Graham, & Hodges, 2005; Hodges, 2000). Thus, individuals with this form of dementia typically cannot provide reliable retrospective reports. Though no research has been published assessing eyewitness accuracy in this population, numerous studies using the DRM paradigm have been published, and they are the focus of this section. Balota et al. (1999) used the DRM paradigm to explore veridical and false memories in individuals with Alzheimer's disease (AD) by comparing five groups: (a) three healthy aging groups (young, young-old, and old-old) and (b) two groups of individuals with AD, mild and very mild. These groups were compared with the use of both recall and recognition tests. Results indicated a reliable decrease in performance for veridical recall across groups (from healthy young to mild AD), whereas false recall rates remained relatively stable across groups. For the group with mild AD, false recall rates were actually higher than veridical recall, indicating a higher level of susceptibility to false memories among individuals with AD than among healthy adults. There was a tendency among old-old and individuals with AD to produce high levels of noncritical intrusions in the recall task. The majority of these noncritical intrusions were semantically related items that were not on the lists and items from previous lists. Veridical recognition results (although contaminated by the earlier recall task) produced a pattern similar to that of the recall results, in that performance decreased across groups. However, a different pattern emerged for the critical items, in which false recognition rates were lower for individuals with mild AD. Budson, Daffner, Desikan, and Schacter (2000) confirmed and extended these findings. The DRM paradigm was again used to compare three groups of participants, two healthy groups (young and old) and one group of patients with AD, with the use of a

9.

PSYCHOLOGICAL IMPAIRMENT

275

recognition task. Results were compared after single and multiple list presentations and indicated that although the level of false recognition was lower for AD patients after a single list presentation, the level of false recognition increased for AD patients after multiple list presentations. The authors suggested that AD patients tend to rely more on gist information than do control subjects, an interpretation that would account for the increase in false recognition in the multiple list presentation condition. Watson, Balota, and Sergent-Marshall (2001) investigated false memories for critical lures by utilizing not only semantic associates, but also phonological associates (e.g., code, told, fold, old, culled, called, scold, sold, hold, coal, colt, and polled were phonological associates of the word cold). Hybrid lists of both semantic and phonological associates were also used (e.g., chill, told, warm, old, shiver, called, winter, sold, freezer, coal, snow, and polled formed the hybrid list for cold). Five groups of participants were tested: three groups of healthy adults (young, old, and old-old) and two OAT groups, very mild and mild. There were two lists of each of the three types (semantic, phonological, and hybrid), and all groups were given free recall tests following the visual presentation of each list. Whereas results for the semantic associates were similar to those found by Balota et al. (1999) and Budson et al. (2000), the results for the phonological associates and the hybrid lists showed a very different pattern. For the phonological lists, three groups (old-old, very mild and mild OAT) produced higher levels of false recall than veridical recall, and for the hybrid lists, all five groups produced higher levels of false recall than veridical recall. The authors discussed these results in relation to spreading activation, suggesting that there is an additive effect of spreading activation for semantic and phonological information in the hybrid lists. A modified version of this paradigm was used to test recognition (Budson, Sullivan, Daffner, & Schacter, 2003). Semantic, phonological, and hybrid lists were visually presented to healthy adults and AD patients, prior to recognition testing. Whereas, unlike the Watson et al. (2001) study, no differences were found between the three types of lists, AD patients displayed lower levels of both veridical and false memory. These results are consistent with both the Balota et al. (1999) study and the Budson et al. (2000) study. In order to investigate whether false memory effects in AD patients are due entirely to semantic deficits, Budson, Desikan, Daffner, and Schacter (2001) used specially constructed perceptually related abstract objects, which reduced semantic processing, to investigate false recognition differences between AD patients and healthy adults. The finding that AD patients had lower levels of both veridical and false memory for these novel items indicated that deficits were not entirely semantically based. The authors hypothesized that AD patients have degraded gist memory. Additional insight into how semantic memory works in AD patients has been obtained following a study of pictures versus words (Budson, Sitarski, Daffner, & Shacter, 2002). It had previously been established that older adults are able to reduce false recognition rates by studying pictures of semantic associates and making use of the distinctiveness heuristic (Schacter, Israel, & Racine, 1999). Budson et al. (2002) extended this work to AD patients and found the opposite effect. After studying pictures, AD patients produced higher rates of false recognition, suggesting that they are unable to use

276

SORACI ET AL.

the distinctiveness heuristic. The authors suggest that viewing the pictures may actually enhance gist memory in AD patients. With improved gist memory, false recognition rates are increased. This finding is consistent with the Budson et al. (2000) finding that false memory rates increased for AD patients after multiple list presentation. As noted by Schacter and Dodson (2002), it takes those multiple trials for AD patients to accumulate semantic gist information, and it is this gist information that is responsible for recognition errors. Taken together, these studies confirm that AD patients suffer serious memory deficits. In all studies, AD patients showed lower levels of veridical recall or recognition than did healthy adults. The pattern of findings for false reports is less clear. On tests of recall, false memories for critical items were similar for individuals with AD and the controls. However, on tests of recognition, individuals with AD tended to produce fewer false memories. This effect was qualified by several methodological changes, however. With repeated lists or use of pictures, false report rates of individuals with AD increased on tests of recognition memory. It appears that group differences may result from differences in the use of gist information. The individuals with AD may have degraded gist memory and therefore lower rates of false recognition, unless procedures that enhance gist memory (e.g., list repetition, pictorial stimuli) are used. These enhancements have the opposite effect for individuals without AD.

DISSOCIATION

The Diagnostic and Statistical Manual (DSM-IV-TR) describes dissociation as a disruption in the integration of consciousness, memory, identity, and perception (APA, 2000). There are individual differences in dissociation among healthy individuals, but higher levels are implicated in a variety of psychological disorders, ranging from the dissociative disorders to acute stress disorder, post-traumatic stress disorder (PTSD), and somatization disorder. A review of the literature on trauma and pathological dissociation is beyond the scope of this chapter (but see the chapter in this volume by Smith & Gleaves); hence, the focus of the studies discussed below will be on individual differences in dissociation in healthy individuals. Dissociation is thought to comprise three factors: depersonalization/derealization, absorption or imaginative involvement, and dissociative amnesia (Bernstein & Putnam, 1986). One of the most commonly used inventories to measure dissociation is the Dissociative Experiences Scale (DES; Bernstein & Putnam, 1986; Carlson & Putnam, 1993), a 28-item self-report questionnaire in which individuals rate on a scale from 0% to 100% how often they experience certain events related to dissociation. The DES items measure dissociative symptoms via three major subscales that tap into the factors described above. In addition, a fourth subscale, the dissociative taxon (DES-T), is an eightitem subset of the DES that measures a pathological form of dissociation. Higher scores on this subscale are usually seen only in those with trauma or dissociative disorders (Eisen, Morgan, & Mickes, 2002). Although healthy individuals dissociate to some extent with little effect on their daily lives (e.g., highway hypnosis, becoming absorbed in a task), higher levels of dissoci-

9.

PSYCHOLOGICAL IMPAIRMENT

277

ation may disrupt memory processes and subsequent accuracy for an event's occurrence. The consequences of such interference may result in high dissociators having little to no memory for an event, or they may extract the gist of a situation and remember super, ficial material rather than specific detail. Reliance on gist,based phenomena can be a precursor to false memory (see Reyna & Brainerd, 1995, and this volume, for a discus, sion of fuzzy trace theory). Dissociation and its effect on memory have been assessed with such laboratory tasks as the DRM paradigm, and for memory of autobiographical and historical events. Discussed below are several studies that have demonstrated how dissociation may have a negative impact on memory.

Dissociation and False Memory in the Laboratory and for Real-World Events Winograd, Peluso, and Glover (1998) examined the relationships among individual dif, ferences in several measures of personality, including the DES, and memory performance in a DRM paradigm. They found a significant relationship between DES scores and a tendency for false recall and recognition. Specifically, they found a positive correlation between number of noncritical intrusions on the recall task and the depersonalization and absorption subscales of the DES. In recognition, individuals who demonstrated a high tendency for dissociative amnesia on that DES subscale also showed a greater propensity for identifying words on a recognition test as "old," and having the subjective experience of remembering them being presented, whether or not they were previously studied. Winograd et al. suggested that high scorers on the DES are less capable in their source,monitoring abilities (e.g., Johnson, Hashtroudi, & Lindsay, 1993), which may in turn contribute to memory error in a DRM task. Hyman and Billings (1998) interviewed participants two separate times about real childhood events, as well as one event that had never occurred. Twenty,seven percent of the participants had created at least a partial false memory for the experimenter, suggested event by the second interview, and the experimenters found a significant pos, itive correlation between scores on the DES and false memory creation. Hyman and Billings posited that people with higher dissociative tendencies are more prone to incor, porating external false information into their own autobiographical memories. Platt, Lacey, Iobst, and Finkelman (1998) sought to determine whether results from laboratory tasks that elicit false memory extend to memory distortion in the real world. They examined the relationships among trait dissociation, performance on the DRM paradigm, and memory for a historical event, the 0. J. Simpson trial verdict. Participants from a previous study had recounted how they had learned about the verdict the night after its occurrence and were contacted at a later time (6, 8, or 12 months) for this study. Although they did not find a correlation between the DES and false memory for either of the tasks, they found a negative correlation between scores on the Tellegen Absorp, tion Scale (TAS; Tellegen & Atkinson, 1974) and memory distortion for learning about the 0. J. Simpson verdict. TheTAS is a scale that examines the extent to which people become heavily engaged in everyday events. Findings indicated that less absorption was associated with greater accuracy for this historical event. A final observation was that

278

SORACI ET AL.

participants who were more accurate in their recount of learning about the 0. J. Simp, son verdict were less likely to exhibit false memory for word lists in the laboratory. Platt et al. concluded that laboratory and real,life memory distortion may be closely related. However, Wilkinson and Hyman (1998) found that memory errors observed in the laboratory may not generalize to the real world. In Experiment 1, they found that par, ticipants' subjective experiences for childhood "known" events (i.e., events that partie, ipants are aware of happening, but have no conscious recollection) could be altered closer to the "remember" (i.e., conscious recollection) end of the remember,know continuum when they were asked to imagine the known event in question. They also found that the DES, especially the absorption subscale, was correlated with the know, remember continuum change. In Experiment 2, they replicated Experiment 1 but added a DRM paradigm component. Wilkinson and Hyman again found that participants' "known" recollections were rated more toward the "remember" continuum after an imagery exercise, but the researchers did not find a relationship between the know, remember continuum and false recall or recognition for the DRM lists. They also found a positive relationship between the know,remember continuum and the depersonalization subscale of the DES. Wilkinson and Hyman suggested that laboratory tasks are not reliable predictors of everyday memory and indicated that different memory processes may underlie autobiographical and laboratory tasks. Together, the above studies suggest that dissociation, or some of the individual fac, tors that it comprises (i.e., depersonalization, absorption, and dissociative amnesia), is associated with increased rates of false memory in laboratory tasks (e.g., the DRM para, digm) and for experimenter,suggested childhood events, historical event inaccuracy, and malleability in subjective feelings about memory judgments. Discussed briefly below are several studies that examine dissociation in a more forensically relevant context.

Dissociation and Memory in Eyewitness Situations Yuille and Daylen (1998) proposed seven different patterns of recall that could affect memory. It is relevant that they suggested that dissociation during an event can result in memories consisting of great or little detail, depending on the focus of attention. People who have internal focus and attend to things other than the event will most likely have poor recall for it, although accuracy of memory for their subjective state at the time can be high. Conversely, Kindt and van den Hout (2001) observed that participants who disso, dated while watching an aversive film, compared with a low,dissociative group, reported higher rates of perceived memory fragmentation for the film. The high, dissociative group (dissociators) reported this perception of fragmented memory, even though their memories for the film's content were as accurate as that of the low,dissociative group. This finding suggests that memory for negative material may remain intact for dissociators, even though they subjectively believe that their memory is poor. Eisen and Carlson (1998) studied how suggestibility is related to dissociation, ab, sorption, and history of abuse in a normal college sample. They staged a scenario and had participants answer misleading questions that were pertinent to central or peripheral

9.

PSYCHOLOGICAL IMPAIRMENT

279

details about the staged event. Answers to misleading questions were indicative of sug~ gestibility. They found that dissociation, as measured by the DES, was positively correlated with suggestibility. Higher DES scores were associated with more errors on central misleading questions, but not on peripheral ones. The authors suggested that because centrally related questions are more likely to be encoded in memorable events, these errors were good indicators of suggestibility. Dissociation, however, was not related to mistakes on questions that were not misleading. In a later study also examining healthy individuals, Eisen, Morgan, and Mickes (2002) found that overall scores on the DES were not related to inaccurately answering misleading questions, but that a subscale of that test that measures pathological dissociation (DES-T) was marginally related to susceptibility to making errors on misleading questions.

Conclusions Pathological levels of dissociation are commonly associated with deficits in cognitive processing (DSM-IV-TR, 2000). However, normal levels of dissociation in healthy individuals also may interfere with attention and memory. Examination of dissociation and memory in the laboratory suggests that people who report more dissociative behaviors on the DES are more susceptible to memory error (although see Kindt & van den Hout, 2001), possibly because their lack of attention to externally presented stimuli leaves them vulnerable to suggestion. Hence, they assimilate nonevents into an existing cognitive framework. Although everyday dissociation can be inconsequential, the above studies indicate that higher levels of normal dissociation may be associated with the incorporation of nonexistent childhood events into memory, memory distortion for historical events, and suggestibility to misleading questions. All of these errors have implications in a forensic setting. TRAUMA

Although still a controversial topic today, trauma's effect on memory has roots in psy~ chology extending back to the nineteenth century. Freud is arguably the most important investigator of repressed and recovered memories, and in his own lifetime and up to the present, his theories of repressed memories of childhood sexual abuse (CSA) have been harshly criticized. Freud's theory that his patients were abused as children was developed prior to his patients' "recovering" such memories, making it appear as if he used suggestive techniques and guided his patients into "uncovering" repressed memories of abuse. After Freud discarded his theory, he no longer found his patients making such claims (see McNally, 2003, for a review; also Read & Connolly and Smith & Gleaves in this volume), suggesting that Freud was influential as a psychoanalyst in eliciting claims that fit the theory he developed. In the past three decades there has been a resurgence of recovered memory claims of abuse, giving rise to continued debates in clinical and experimental psychology, and

280

SORACI ET AL.

proving consequential to patients, the accused (usually family members), and the courts. Such claims are made sincerely and need to be considered seriously, but without inde~ pendent corroboration it becomes a war of allegations and denial. The psychological lit~ erature has empirically shown that memory is prone to errors of omission and distortion (e.g., Schacter, 1999), and that misleading questions, hypnosis, guided imagery, and sug~ gestion can distort and even create memories for events that never occurred.

DRM Paradigm and Trauma Bound by obvious ethical constraints, researchers cannot experimentally assess whether memory abnormality follows trauma in humans, and if so how it manifests. Although false memory for real~world traumatic events cannot be empirically tested in the labora~ tory, researchers have used the DRM paradigm to study trauma groups and the false recall or recognition of words. Clancy, Schacter, McNally, and Pitman (2000) used a modified DRM paradigm to study false recognition in four groups: women who reported recovered memories of CSA, referred to as the recovered memory group; women who believed they were sexually abused as children, but have no memories of the abuse (repressed~memory group); women who were sexually abused as children and always remembered the abuse (continuous~ memory group); and women with no history of CSA (control group). Results showed that there was no difference between the groups for true recognition of the stimulus items; however, when looking at false recognition rates, recovered~memory participants showed higher rates of false recognition than the other groups. Bremner, Shobe, and Kihlstrom (2000) also studied recall and recognition in self~reported sexually abused women with post~traumatic stress disorder (PTSD) and non~PTSD controls. They found lower rates of correct recall, equal rates of critical lure false recall, and a higher inci~ dence of noncritical intrusions (extra~ list items that are falsely recalled but not the crit~ ical foils) in the PTSD group compared with the control groups. For recognition, the PTSD group falsely recognized more critical lures and noncritical foils than the con~ trol groups. Zoellner, Faa, Brigidi, and Przeworski (2000) examined memory in trauma~ tized participants with PTSD, traumatized participants without PTSD, and a non~ traumatized control (NTC) group. Immediate free recall tests showed that PTSD and non~PTSD traumatized participants generated more critical foils than the NTC group. No significant difference was detected between groups for critical foil recognition, al~ though there was a trend for the PTSD group to falsely recognize critical foils more than the NTC group. Clancy, McNally, Schacter, Lenzenweger, and Pitman (2002) investigated memory distortion in people reporting abduction by aliens. What makes this research unique among the studies discussed above is that (a) we can be reasonably confident that any reports of experiencing such a phenomenon could be explained by other means, as dis~ cussed below, and (b) these individuals might not be officially recognized as a trauma~ tized group. Studying three groups, Clancy et al. found that the participants who re~ ported having recovered memories of abduction by aliens had greater false recall and false recognition rates than participants who believed they had a repressed memory of

9.

PSYCHOLOGICAL IMPAIRMENT

281

alien abduction, who in turn had greater rates of false memory than a control group. Individual differences revealed that, compared with the control group, the repressed and recovered memory groups scored higher on measures of hypnotic suggestibility, perceptual aberration, and magical ideation, which could lead to vulnerability for false beliefs. The authors maintained that reports of alien abduction could result from hypnopompic (i.e., upon awakening) states in which people experience auditory, tactile, and visual hallucinations during a state of sleep paralysis, and cited work by Hufford (1982) claiming that at least 15% of the general population has experienced sleep paralysis. Although it is difficult to ascertain what the participants actually experienced, there seems to be a plausible alternative explanation to alien abduction. Conclusions Together, results from different research groups have indicated that the DRM paradigm can distinguish between groups of people who have experienced trauma from those who have not; those who have developed PTSD from the result of a trauma from those who have not; and people who believed they experienced an event that might be considered traumatic (i.e., being abducted by space aliens) from those who have not made such claims. Findings from this research do not suggest that traumatized individuals fabricate, or are more likely to fabricate, memories about events in their everyday lives, or about their specific traumatic event. Rather, they show that traumatized populations, especially those with PTSD, may be more prone to incorporate suggested information into their memories. Studying memory and false memory in traumatized populations may shed light on thought processes and memory biases associated with traumatization that extend beyond information provided in the environment. Readers are strongly cautioned that the paradigm's utility cannot generalize past the laboratory, and that these studies make no claims against the veracity of trauma victims' experiences. It may be that traumatic events did occur and they may have affected the cognitive processing of the above-mentioned trauma groups, hence making them susceptible to the false memory effect in the laboratory (Clancy et al., 2002). Memory is reconstructive rather than reproductive in nature; thus we are limited in our ability to determine whether an event occurred, and if so, what the precise details of it were. The laboratory research discussed above indicates that both healthy individuals and trauma populations are susceptible to memory error. Studying false memory in a controlled laboratory setting provides an objective way of determining the cognitive and external components that may be responsible for false memory generation, thus allowing for the potential oflaboratory research informing clinical practice and the legal system. DRUGS AND ALCOHOL

Eyewitness researchers are frequently asked by police investigators, as well as prosecution and defense counsel, whether memories for an event are compromised by the consumption of various drugs (e.g., alcohol, marijuana). Typically, the drugs being referred

282

SORACI ET AL.

to are those that are taken for recreational or addictive goals. It is recognized, of course, that psychoactive drugs vary enormously in their consequences, from depressants to tranquilizers, stimulants, analgesics, antipsychotics, and psychedelics, and it is reasonable to assume that cognitive processes including memory will be affected. The question itself is also relevant because it is well known that events of forensic importance, indeed the majority of violent assaults and other crimes, very often occur in the context of substances that alter psychological and physiological processes of perpetrators, victims, and witnesses (e.g., Bean, 2004; Corwyn & Benda, 2002; Gerson, 1978; Graham & West, 2001; Harrison, Erickson, Adlaf, & Freeman, 200; Yuille, Tollestrup, Marxsen, Porter, & Herve, 1998). However, the psychological literature on the effects of these substances on recollections of complex events is sparse, and, as a result, the specific memorial consequences of the ingestion of most drugs are not well understood. Despite the very limited empirical literature, we have further restricted the scope of this review by excluding those studies in which researchers (psychopharmacologists, psychiatrists, psychologists, and neuropsychologists) have focused on the perceptual, cognitive, and physiological consequences of sustained abuse of alcohol, cocaine, amphetamine, and similar substances. The investigation of the long-term effects of substance abuse has not been examined at all within eyewitness contexts. Sustained use of any drug causes neuronal adaptation and neurotransmitter-receptor changes that affect tolerance for and memorial consequences of the drug (e.g., Brokate, Eling, & Lanz, 2004; Curran, 2000; Curran & Weingartner, 2002; Hildebrandt, Julien, 2005). As an example, alcoholic blackouts are known to occur with excessive or persistent abuse of alcohol, and, by definition, a person who experienced a blackout for a time period in which a crime was committed (and witnessed) will provide neither accurate nor reliable recollection of the event (Goodwin, 1995; Sweeney, 1989) and, in some cases, may develop false memories of a crime to fill in the blackout period (see Gudjonsson, 2003; Kassin, 1997, 2005). To date, there have been no research studies examining blackout-prone participants for their recollections of a standardized and controlled complex event, although there has been examination of episodic word recall by these individuals (e.g., Miller, Hertel, Saucedo, & Hester, 1994) and of executive cognitive functions among heavy-drinking college students (Blume, Marlatt, & Schmaling, 2000). Furthermore, we assume that if individuals with such pathologies and addictions were witnesses to a specific and forensically relevant event, other issues such as their competency to give testimony or stand trial and assignment of criminal responsibility (if defendants) may be raised long before the witness/defendant attends court and the court hears relevant expert opinion on eyewitness memory issues. To study properly the effects of drugs on eyewitness memory, we require controlled situations in which imminent witnesses ingest or receive a substance prior to an event and are subsequently questioned about the event, while either under the influence or sober. The number of such studies is small, and, unfortunately, they do not likely mimic either the variety or the intensity of the psychological states often achieved by witnesses in forensic contexts. There are several reasons for the paucity of controlled research: first, many legitimate ethical concerns surround the psychological and physical health of research participants who are exposed to drugs in significant dosages. Second, there are

9.

PSYCHOLOGICAL IMPAIRMENT

283

methodological constraints imposed by participants' past experience with drugs that greatly restrict the range of critical levels of a substance in terms of both awareness and tolerance. For example, the use of a balanced placebo design is effective at low dosages of alcohol (Marlatt & Rosenhow, 1980) but can be compromised with larger dosages. At larger dosages, experienced drinkers receiving placebo drinks quickly become aware that they have not consumed the equivalent of four to six 1~ounce drinks in the preceding hour, an awareness that quickly subverts the purpose of the placebo design. Investigators generally assume that the effects of drugs upon memory encoding, storage, and retrieval increase as consumption levels increase, possibly in a linear fashion, but perhaps in a curvilinear manner with amount and time, known as the dose~response curve. A general negative relationship (increased dosage, lower performance) is likely to be accurate, but the precise shape of the function across a broad range of dosages is much more difficult to determine. Most controlled and illegal substances as used in sod~ ety are often contaminated with other substances. For example, even trademarked alco~ hoi contains ingredients other than ethanol. As a result, when the effects of alcohol, marijuana, and other street drugs on cognition and memory are assessed, it is their active ingredients that should be studied, not the fillers and contaminants that frequently ac~ company them. Finally, the use of drugs in society is very often complicated by the fact that people take more than one drug at a time and the interactive effects of these drugs make behavioral, psychological, and memorial outcomes very unpredictable. Although there is no reason to anticipate that the consumption of any psychoactive drug would not be relevant to cognitive processing and ultimately, to memory perfor~ mance, and little reason to suppose that effects of drugs would generally have anything but detrimental effects on memory, the number and focus of pertinent investigations are too limited to make these broad generalizations. It is certainly the case that surveys of eyewitness experts (e.g., Kassin, Tubb, Hosch, & Memon, 2001) and of potential jurors (Read, Desmarais, Rojas, & Vander Veen, 2005) demonstrate a widespread expectation that consumption of alcohol, marijuana, and other drugs impairs memory for a witnessed event. However, the outcome of controlled research sometimes tells us otherwise. For example, with certain kinds of information, it appears that consumption of alcohol may not impair processing at all and, if ingested soon after a specific event, may even serve to enhance (relative to a control condition) recollection of event details (Bruce & Pihl, 1997; Wilson, 1982). Similarly, with medical (but nonetheless, often abused) drugs, mem~ ory enhancements have also been reported (Curran & Weingartner, 2002). The complexity of these effects likely arises from the variety of cognitive processes involved in or related to memory. To quote Curran and Weingartner, "no drug currently exists that only affects memory" (2002, p. 125). Typically, memory researchers and theo~ rists distinguish between encoding, storage, and retrieval processes in memory (e.g., Roediger, McDermott, & Goff, 1997; Schacter, 1996). If someone is inebriated at the time of encoding, it is likely that impaired perception of or insufficient attention to the event details will limit the event's encoding in memory, making eventual, accurate, and complete retrieval of the event details highly unlikely. On the other hand, if a moderate dose of alcohol is consumed soon after an event is witnessed, alcohol's ability to reduce arousal and limit the amount of attention paid to potentially interfering activities over

284

SORACI ET AL.

the retention interval may mean that eventual recall is actually better (more accurate and complete) than if sobriety had been maintained throughout the interval following the event (Bruce & Pihl, 1997). Similarly, retrieval of information will benefit if there is a matching of contexts and cues available at encoding with those at retrieval (Schacter, 1996). That is, if the psychological effects of a drug are present at encoding, a witness may better retrieve the information when under the influence of the drug than when sober (e.g., Eich, 1980; Weingartner, Putnam, George, & Ragan, 1995). As the reader may be aware, this phenomenon, called state-dependent memory, is sometimes observed with alcohol ingestion, whereby recall is greater when alcohol status (inebriated or not) at retrieval matches that at study. However, if encoding occurred in the absence of a drug, retrieval of the information while under the influence of a drug is generally unlikely to be affected (but see below). The nature of the event and the types of information it includes, along with a host of situational and personal variables, as well as the manner by which memory is tested, all play roles in determining whether information will be subsequently remembered. Furthermore, drugs can show differential sensitivity to these variables: to various types of information (and resultant memory processes), to different stages of memory (i.e., shortterm or long-term recall), and to different forms of testing. First a distinction that is relevant here is that between automatic and controlled processes: several studies have demonstrated that the ingestion of alcohol may produce a dissociation in memory performance as a result of the two types of processing. For example, Tracy and Bates (1999) found that whereas recall of words that varied in their frequencies of presentation in a word-list learning experiment was significantly impaired by alcohol, the accuracy of estimation of the words' frequencies by participants was unaffected. The argument made is that because the encoding of frequency information requires little effort, it is automatically encoded (see Hasher & Zacks, 1984; Zacks & Hasher, 2002), and, therefore, once it is encoded, alcohol intoxication will have little effect on its retrieval. However, because free recall requires sufficient conceptual and elaborative encoding to be successful and this is a controlled or effortful process, alcohol significantly impaired encoding (and ultimately, retrieval) of the word themselves on the free recall test. When this distinction is applied to an eyewitness situation, one might speculate that encoding of frequency information about repeated events (e.g., number of gunshots, number of people observed) may be relatively unaffected by alcohol consumption. Second, when eyewitness researchers refer to memory, they are generally referring to long-term memory recall, the recollection of something that happened at least minutes, likely hours or days, following an event, rather than only seconds later. Generally, drugs do not affect short-term (immediate recall, only seconds later) memory, but many do affect encoding and storage, resulting in impaired retrieval of an event from long-term memory (hours or days later). Various tranquilizers in the category of drugs called benzodiazepines (e.g., Valium, Ativan) are well known to have these amnestic effects upon long-term but not short-term recall (Curran & Weingartner, 2002; Parkin, 1987). Third, most lay readers assume a test of memory to be one in which a person is asked to recall some information, a test that is said to be explicit because the participant is asked directly about the event he or she witnessed. However, there are also implicit tests

9.

PSYCHOLOGICAL IMPAIRMENT

285

of memory in which the witness is not asked for recall of event-relevant information, but instead to respond to a variety of stimuli, some of which are conceptually or semantically related to the event. For example, the participant might complete a word stem (e.g., A L C----) or make a rapid decision about whether a letter sequence is a meaningful word. Implicit tests are rarely used in forensic investigations because they have little face validity for the trier-of-fact, and they would need to be individually constructed carefully for each event and pretested to rule out bases for responding in some manner other than implicit knowledge of an event. For example, some polygraph researchers have recommended the use of the Guilty Knowledge Test (GKT) in place of a procedure called the Control Question Test (e.g., Elaad, 1998). For our purposes here, the GKT could be used as an implicit test because examinees who are asked to respond to every question with a denial will show differences in reaction time and other physiological measures to the questions that relate to guilty knowledge, even if such knowledge is only implicitly known. In any case, we do know that following consumption of some drugs, performance on the two types of tests can be very different, with explicit tests showing little recall and substantial memory impairment, but with implicit tests demonstrating perfectly normal "memory" for the experience (Ray, Bates, & Bly, 2004). These dissociations have been seen with alcohol on laboratory episodic (list-learning) tasks but not yet in regard to memory for complex events. The dissociation tells us that although information may be unavailable to conscious recall, its effects can nevertheless be detected with more sensitive (implicit or unconscious) tests. For example, Ray et al. (2004) used implicit (or indirect) and explicit (or direct) tests of memory to test selectively for information acquired unintentionally or intentionally. Their results revealed that alcohol intoxication affected explicit (e.g., free recall) memory but not implicit (e.g., semantic priming) memory (see also recent work by Fillmore, Vogel-Sprott, and Gavrilescu, 1999, and Kirchner and Sayette, 2003). A related distinction has to do with the phenomenology of memory experiences. In particular, when recalling details of an event, we can often recollect or "relive" our presence at the event, or sometimes we simply know that we had been there, but we have no recollective details with which we can support our claim. Many drugs produce dissociations in these experiences: for example, after consuming alcohol, participants may be unaffected in the level of accuracy of their recognition and recall decisions about an event, but they are much less likely to say they can actually recollect the experience. Instead, they just know it happened (e.g., Curran & Hildebrandt, 1999). From a forensic perspective these descriptions are important because an experience reported by a witness as "known" rather than "recollected" may be viewed as having much less credibility to the trier-of-fact. For the most part, investigations of the effects of drugs on memory have focused on the encoding and storage stages, rather than retrieval. The reason is simple: the largest effects of drugs are observed at these stages. Similarly, for our purposes, forensic questions would more likely surround the state the witness was in at the time of observing the event, rather than what he or she did later. We described above a number of caveats as to the restricted nature of this review. We restrict the review even further to alcohol and marijuana alone because studies that meet our purposes here (i.e., engage normal

286

SORACI ET AL.

participants in moderate consumption under controlled settings and recalled following a specific retention interval) have only been completed with alcohol and marijuana. Nonetheless, investigators, lawyers, and researchers need to keep in mind the variety of different memory processes, tests, and experiences discussed above that likely are relevant to all drugs and recollection of particular eyewitness events. We cannot do justice here to the detailed psychopharmacological research on drug effects: the relevant research literatures are enormous in size, and there is simply very little work completed with paradigms that closely approximate eyewitness events.

Effects of Alcohol That the consumption of alcohol can impair perceptual judgments, awareness, memory, and decision-making has been well documented (Cashon, Curtin, Lang, & Patrick, 2003; Curran & Weingartner, 2000; Heather, Peters, & Stockwell, 2001). All of these cognitive processes would be implicated in the observation and recollection of a complex event. The conclusion has been that, in general, depending upon the individual, as consumption (and concomitant blood alcohol levels) and task complexity increase, the extent of cognitive and motor impairment increases. However, depending upon the individual's past experience with alcohol, physical condition, and size, race, and health, the impairment effects can vary dramatically (Bruce & Pihl, 1997; Cox, Yeates, Gilligan, & Hosier, 2001). Not every individual with blood alcohol levels at or above a specific reading would demonstrate impairments in all of these areas, whereas others may simply be unable to complete the specific laboratory tasks. Briefly, with standard cognitive experimental paradigms, alcohol has been shown to impair both the encoding and retrieval of verbal and pictorial information. However, depending upon the individual's focus of attention, levels of cognitive control needed to complete the task, and level of arousal, impairment may be apparently nonexistent or substantial (Cashon et al., 2003). Meissner, Sporer, and Schooler (2005) have examined some effects of alcohol on person identification tasks in particular. Here we examine event memory as well. Steele and Josephs (1990) have argued that with increased consumption, the drinker demonstrates increasing signs of "alcohol myopia," that is, an increased tendency to narrow one's attention to salient features of a situation to the relative exclusion of other information. As a result, details attended to in this increasingly smaller focal area are maintained reasonably well in memory, but, relative to control (placebo) drinkers, memory for peripheral information will be drastically reduced. Dysart, Lindsay, MacDonald, and Wicke (2002) argued that this myopic process should have consequences for performance in a person identification situation because drinkers may not encode the necessary features of a perpetrator to allow them to discriminate a target's face from other faces who share many of the same, but not all, features. As a result, they anticipated that drinkers would have more difficulty with a Target-Absent lineup (or show-up) identification procedure. Their study supported this perspective because identification performance declined with increasing blood alcohol; that is, "innocent" foils were selected more frequently as consumption rose. Their study was carried out in the field, and ingeniously completed at that, but it does not allow a distinction to be

9.

PSYCHOLOGICAL IMPAIRMENT

287

made between amount consumed and individual differences in memory and identification abilities. Yuille and Tollestrup (1990) initially explored the effects of alcohol consumption with witnesses to a staged theft. The consumption of alcohol reduced the amount of information recalled about the event and the person on both an immediate and delayed (1 week later) test. Alcohol consumption reduced the amount of scoreable information in the free recall (cognitive interview) task (immediate and delayed), and alcohol drinkers reported significantly less correct information, although both group performances were high, and the numerical difference (2%) was very small. As in the study by Dysart et al. (2002), alcohol participants were under the influence at the time of the immediate test. When one of the actors in the staged event was included within a photo lineup, alcohol had no effect on identification performance. However, as in the study by Dysart et al., performance on a Target-Absent lineup was significantly worse for the alcohol than for control participants. Read, Yuille, and Tollestrup (1992) tested the alcohol-myopia hypothesis in two experiments with a "crime paradigm" wherein participants were perpetrators who were (as compared with passive witnesses) required to steal money from a professor's office under either low or high arousal (or stress). Memory was assessed 1 week later when participants were no longer under the influence of alcohol. Under low stress, person identification by the research participant was significantly impaired compared with placebo participants; however, when the participants experienced high arousal, the debilitating effects of alcohol were completely eliminated in person identification. As part of the scenario, participants interacted with two target persons, one central to the crime and one peripheral to the crime. Alcohol was anticipated to exaggerate the difference in identification performance on the two targets, but it did not. Instead, evidence for the focusing effects of alcohol myopia upon event recall was clearly seen in recollection of the sequence of events experienced by the participants. Those who consumed alcohol were less able to recall details of the central (intruder) person (Experiment 1) as well as the correct sequence of their own actions during the crime (Experiment 2). Increased arousal, on the other hand, certainly had myopic effects on person identification, because the two people were remembered equally well under low stress but very differently under high arousal, with significant impairment for the peripheral person (see the Reisberg & Heuer chapter, this volume, for a discussion of these kinds of effects). Overall, however, the important message of this study is that increasing someone's level of arousal or stress appeared to completely eliminate the debilitating effects of moderate alcohol consumption on memory performance. Thus, there appears to be some support for the lay idea that fear or stress can serve to "sober up" someone who has been drinking, at least at moderate levels of consumption. As mentioned above, the scientific investigation of the effects of drugs, including alcohol, on cognition and behavior is done in such a manner as to control for the participants' expectancies about the effects of the drug. Accordingly, placebo groups always receive the "carrier" for the drug (a drink) that has an appearance (and odor) identical to that of the true alcohol drink. Dozens of alcohol studies using such designs have demonstrated that participants who receive a placebo also demonstrate many of the same

288

SORACI ET AL.

behavioral changes, impairments, and subjective feelings of intoxication as the alcohol condition participants. In short, our expectations about the effects of a drug provide for similar kinds of behavior, and these expectations are experientially and culturally driven (e.g., Julien, 2005; Moore, 2003). From a memory perspective, the effect of an alcohol placebo may be to reduce one's attentiveness to an event, to exert less effort in recall, or to make use of a different (more liberal) criterion for the acceptance of post-event information. Indeed, Assefi and Garry (2003) recently demonstrated that subjects whoreceived a placebo were more likely to be misled by post-event misinformation than were non-placebo control condition participants. As well, the placebo group responses were presented with greater confidence. We assume that similar expectations exist for people who choose to consume alcohol in naturalistic settings, and, as a result, the impact of post-event information from other sources (witnesses and the media) may well influence their recall of an event as well. Finally, Read et al. (1992) manipulated alcohol expectations at both encoding (the crime event) and retrieval (the cognitive interview). An effect was obtained in which recall was best in the interview when what participants were told at encoding (alcohol or no alcohol-placebo) was inconsistent with what they were told at the interview about what they had received at encoding. The effects were complex but, along with the observations of Nelson, McSpadden, Fromme, and Marlatt (1986), the results do demonstrate that witnesses' retrospective beliefs about their consumption of alcohol can affect the amount of correct information recalled. Within some interviewing and interrogation situations, evidence presented as to the amount of alcohol consumed may, as found by Assefi and Garry (2003), affect both responses to memory questions and confidence in those responses. The simulated eyewitness studies, together with our review of laboratory episodic tasks, lead to several conclusions. First, moderate quantities of alcohol are likely to reduce subsequent recall, particularly for information that was peripheral to the meaning of the event. With increased dosages we would anticipate the effects to be dramatically worse. Second, the more sensitive implicit tests of memory suggest that alcohol does not affect all types of information equally, with sparing of information that is automatically encoded (possibly frequency and spatial information), and that with such tests memory or knowledge of other details may be revealed. Third, one consistent effect of alcohol that is important in court proceedings is the reduced recollective phenomenology associated with an event, a reduction that may make a witness appear less confident and less knowledgeable. Finally, it is likely that individuals who have consumed alcohol may be less resistant to the effects of suggestion and post-event information. If true, more precautions need to be taken during their interviews and interrogation.

Effects of Marijuana As with alcohol, marijuana intoxication appears to have its primary effects at the encoding and storage stages (Heishman, Arasteh, & Stitzer, 1997; Parkin, 1987; Wetzel, Janowsky, & Clopton, 1982). There is limited evidence that marijuana intoxication can produce severe memory deficits, but does so very rarely (Stracciari, Guarino, Crespi, & Pazzaglia, 1999). As with alcohol of course, sustained use of the substance may have

9.

PSYCHOLOGICAL IMPAIRMENT

289

cognitive deficits that, as yet, are both poorly documented and understood (Gunderson, Vosburg, & Hart, 2002). These effects are not particularly relevant to the question of interest here: Does marijuana intoxication impair memory for a single witnessed event? Virtually no eyewitness testimony work has been done on the effects of marijuana. Only one study by Yuille et al. (1998) assessed the effects of marijuana on memory for a complex and staged event in which a purported experiment was disrupted by an angry confederate, followed by a heated interaction between researcher and confederate, and witnessed by a group of observers. The experimental group smoked a marijuana cigarette (with a moderate tetrahydrocanabinol (THC) content of 1.24%), and the placebo group smoked a highly similar cigarette with THC effectively at zero. Memory tasks included a full Step-Wise Interview procedure (Yuille, Hunter, Joffe, & Zaparniuk, 1993) and a photo identification task for the confederate. The participant's subjective ratings of the effects of marijuana demonstrated that the marijuana group did indeed feel intoxicated, but in a manner consistent with placebo effects in alcohol research, even the control condition gave intoxication ratings substantially beyond "no effect." The results on the memory tasks demonstrated impairment (21% fewer details recalled) at the immediate interview (while the marijuana participants were still intoxicated); however, this difference proved to be temporary because it was eliminated when all subjects were tested 1 week later. Control participants did not significantly improve their recall from the first to the second interview. There was no measurable effect of marijuana on the identification task. The authors argued that, given the equivalence of recall at the second interview, marijuana affected the retrieval of information on the immediate test, but not the storage of information. The retrieval deficit was hypothesized to result from reduced motivation among the marijuana group to report details. Overall then, consumed in moderation, marijuana exerts effects on memory that appear to be fairly small and temporary.

SUMMARY The research reviewed in this chapter highlights the many challenges of working with special populations in laboratory and applied settings. Because of the methodological challenges associated with these groups, they often are under-studied, as was shown for all of the groups discussed. For many of the populations, basic studies of memory abilities are lacking, and for most, research regarding the reliability and validity of eyewitness reports is nonexistent. This has limited some of the conclusions that can be drawn, but directs future research to areas of need for increasing understanding of basic and applied issues in memory. It is clear that many of the questions most pertinent to the participation of these populations in legal contexts are as yet unanswered. The foci of this chapter were those variables that explain individual differences in memory accuracy. For legal settings, decisions most often must be made about the competency and accuracy of an individual witness. It is typical that individual differences in special populations are larger than in normative samples, though most research on memory involves group patterns of performance. In each section of this chapter, therefore, independent variables that explain some portion of the individual differences within a population were highlighted. Important considerations regarding level of impairment

290

SORACI ET AL.

(e.g., level of mental retardation, stage of Alzheimer's disease, level of intoxication), questioning context (free recall vs. closed-ended), and other variables likely to be critical for legal decision-making were featured. The research reviewed demonstrates that members of these populations can provide accurate recollect reports for at least parts of the experienced event(s). Of greatest concern are the completeness of the reports and the reporting of false information. Devising techniques for enhancing completeness of recollection that minimize inaccurate responding is a challenge for the field. Particular attention must be given to defining the ideal method(s) for eliciting accurate responses from each population and individual. In general, free recall, or open-ended questioning, provides the most accurate, though incomplete, reports. When closed-ended questioning is used more information will be reported, but at the expense of a greater proportion of false reports. This is critical for legal contexts in which the individual may be questioned multiple times, by different interviewers, at varying intervals from the event, with the use of different questioning techniques. Providing expert judgment on decisions regarding competency to testify and accuracy of memory reports from special populations represents an area in which psychological science can provide a great public service. The populations considered in this chapter often have higher rates of contact with the legal system, particularly as witnesses and victims, and often do not enjoy the same access to the system because of their seeming diminished capacity to participate fully. It is hoped that this chapter has demonstrated the challenges of this field of inquiry, and how psychologists can use information gained from studying individual differences and situations that affect cognition to assess the veracity of memory, especially when there is uncertainty about the precise details of an event and there are limited witnesses to its occurrence.

ACKNOWLEDGMENTS This chapter is dedicated to the memory of the first author, Sal Soraci, who passed away during the preparation of the manuscript. The conceptualization and structure of the chapter were Sal's, and we hope the final version of the chapter reflects his thinking and goal of demonstrating the importance of basic research for applied issues in cognition. Preparation of this manuscript was partially supported by National Institute of Child Health and Human Development (NICHD) grant HD42559 awarded to the Eunice Kennedy Shriver Center of the University of Massachusetts Medical School. The contents of the chapter are solely the responsibility of the authors and do not represent the official views of the NICHD.

REFERENCES Aleman, A., Hijman, R., deHaan, E. H. F., & Kahn, R. S. (1999). Memory impairment in schizophrenia: A meta-analysis. American Journal of Psychiatry, 156 (9), 1358-1366. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC: Author.

9.

PSYCHOLOGICAL IMPAIRMENT

291

Assefi, S. L., & Garry, M. (2003). Absolut® memory distortions: Alcohol placebos influence the misinformation effect. Psychological Science, I4, 77-80. Bean, P. (2004). Drugs and crime (2nd. ed.). Cullompton: Willan Publishing. Balota, D. A., Cortese, M. 1., Duchek, 1. M., Adams, D., Roediger, H. L., McDermott, K. B., & Yerys, B. E. (1999). Veridical and false memories in healthy older adults and in dementia of the Alzheimer's rype. Cognitive Neuropsychology, I6, 361-384. Bean, P. (2004). Drugs and crime (2nd. ed.). Cullompton: Willan Publishing. Belmont, 1· M., & Butterfield, E. C. (1971). Learning strategies as determinants of memory deficiencies. Cognitive Psychology, 2, 411-420. Bennetto, L., Pennington, B. F., & Rogers, S. 1. (1996). Intact and impaired memory functions in autism. Child Development, 67, 1816-1835. Bernstein, E., & Putnam, F. (1986). Development, reliability, and validity of a dissociation scale. Journal of Nervous and Mental Disease, I74, 727-735. Beversdorf, D. Q., Smith, B. W., Crucian, G. P., Anderson, 1. M., Keillor, 1. M., Barrett, A.M., et al., (2000). Increased discrimination of "false memories" in autism spectrum disorder. Proceedings of the National Academy of Sciences, 97, 8734-8737. Borkowski, 1. G. ( 1985). Signs of intelligence: Strategy generalization and metacognition. In S. R. Yussen (Ed.), The growth of reflection in children (pp. 105-144). New York: Academic Press. Boucher, 1. (1981). Immediate free recall in early childhood autism: Another point of behavioural similarity with the amnesic syndrome. British Journal of Psychology, 72, 211-215. Boucher, 1., & Warrington, E. K. (1976). Memory deficits in early infantile autism: Some similarities to the amnesic syndrome. British Journal of Psychology, 67, 73-87. Bray, N. W., Reilly, K. D., Fletcher, K. L., Huffman, L. F., Grupe, L.A., Villa, M. F., et al. (1998). Memory competencies and deficiencies: A conceptual framework and the potential of connectionist models. In S. A. Soraci & W. 1. Mcllvane (Eds.), Perspectives on fundamental processes in intellectual functioning (Vol. 1, pp. 3-44). Greenwich, CT: Ablex/JAI Press. Brebion, G., Amador, X., Smith, M. 1., & Gorman, 1. M. (1997). Mechanisms underlying memory impairment in schizophrenia. Psychological medicine, 27, 383-393. Bremner, 1· D., Shobe, K. K., & Kihlstrom, 1. F. (2000). False memories in women with selfreported childhood sexual abuse: An empirical study. Psychological Science, II (4), 333-337. Brown, A. (1974). The role of strategic behavior in retardate memory. InN. R. Ellis (Ed.), International review of research in mental retardation (Vol. 7, pp. 55-101). New York: Academic Press. Bruce, K. R., & Pihl, R. 0. (1997). Forget "drinking to forget": Enhanced consolidation of emotionally charged memory by alcohol. Experimental and Clinical Psychopharmacology, 5, 242-250. Budson, A. E., Daffner, K. R. Desikan, R., & Schacter, D. L. (2000). When false recognition is unopposed by true recognition: Gist-based memory distortion in Alzheimer's disease. Neuropsychology, 14, 277-287. Budson, A. E., Desikan, R., Daffner, K. R., & Schacter, D. L. (2001). Perceptual false recognition in Alzheimer's disease. Neuropsychology, I5, 230-243. Budson, A. E., Sitarski, 1., Daffner, K. R., &Schacter, D. L. (2002). False recognition of pictures versus words in Alzheimer's disease: The distinctiveness heuristic. Neuropsychology, 16, 163-173. Budson, A. E., Sullivan, A. L., Daffner, K. R., & Schacter, D L. (2003). Semantic versus phonologial false recognition in aging and Alzheimer's disease. Brain & Cognition, 5, 251-261. Carlin, M. T., Soraci, S. A., Dennis, N. A., Chechile, R. A., & Loiselle, R. C. (2001). Enhancing free-recall rates of individuals with mental retardation. American Journal on Mental Retardation, 106, 314-326. Carlin, M. T., Soraci, S. A., & Strawbridge, C. P. (2005). Generative learning during visual search for scene changes: Enhancing free recall of individuals with and without mental retardation. American Journal on Mental Retardation, I10, 13-22. Carlson, E. B., & Putnam, F. W. (1993). An update on the dissociative experiences scale. Dissociation, 6, 16-27.

292

SORACI ET AL.

Cashon, T. S., Curtin, J. J., Lang, A. R., & Patrick, C. J. (2003). Deleterious effects of alcohol intoxication: Diminished cognitive control and its behavioral consequences. Journal of Abnormal Psychology, 112, 476-487. Clancy, S. A., McNally, R. J., Schacter, D. L., Lenzenweger, M. F., & Pitman, R. K. (2002). Memory distortion in people reporting abduction by aliens. Journal of Abnormal Psychology, 111 (3), 455-461. Clancy, S. A., Schacter, D. L., McNally, R. J., & Pitman, R. K. (2000). False recognition in women reporting recovered memories of sexual abuse. Psychological Science, 11 (1), 26-31. Clare, I. C. H., & Gudjonsson, G. H. (1993). Interrogative suggestibility, confabulation, and acquiescence in people with mild learning disabilities (mental handicap): Implications for reliability during police interrogations. British Journal of Clinical Psychology, 32, 295-301. Corwyn, R. F., & Benda, B. B. (2002). The relationship between the use of alcohol, other drugs, and crime among adolescents: An argument for the delinquency syndrome. Alcoholism Treatment Quarterly, 20, 35-49 Cox, W. M., Yeates, G. N., Gilligan, P. A. T., &Hosier, S. G. (2001). Individual differences. In N. Heather, T. J. Peters, & T. Stockwell (Eds.). International handbook of alcohol dependence and problems (pp. 357-375). Indianapolis, IN: John Wiley & Sons. Curran, H. V. (2000). Psychophamacological perspectives on memory. In E. Tulving and F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 539-554). Oxford: Oxford University Press. Curran, H. V., & Weingartner, H. (2002). Psychopharmacology of human memory. In A. D. Baddeley, M.D. Kopelman, &B. A. Wilson (Eds.), The handbook of memory disorders (pp. 123141). Chichester: John Wiley & Sons. Danion, J.-M., Rizzo, L., & Bruant, A. (1999). Functional mechanisms underlying impaired recognition memory and conscious awareness in patients with schizophrenia. Archives of General Psychiatry, 56(7), 639-644. Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 38(1), 17-22. Dent, H. (1992). The effects of age and intelligence on eyewitnessing ability. In H. Dent & R. Flin (Eds.), Children as witnesses. Chichester, UK: John Wiley & Sons. Dudas, R. B., Clague, F., Thompson, S. A., Graham, K. S., & Hodges, J. R. (2005). Episodic and semantic memory in mild cognitive impairment. Neuropsychologia, 43, 1266-1276. Dysart, J. E., Lindsay, R. C. L., MacDonald, T. K., & Wicke, C. (2002). The intoxicated witness: Effects of alcohol on identification accuracy from showups. Journal of Applied Psychology, 87, 170-175. Eich, J. E. (1980). The cue-dependent nature of state-dependent retrieval. Memory & Cognition, 8, 157-173. Eisen, M. L., & Carlson, E. B. (1998). Individual differences in suggestibility: Examining the influence of dissociation, absorption, and a history of childhood abuse. Applied Cognitive Psychology, 12, S4 7-S61. Eisen, M. L., Morgan, D. M. & Mickes, L. (2002). Individual differences in eyewitness memory: Examining relations between acquiescence, attention, and resistance to misleading information. Personality and Individual Differences, 33(4), 553-571. Elaad, E. (1998). The challenge of the concealed knowledge polygraph test. Expert Evidence, 6, 161-187. Ellis, N. R. (1970). Memory processes in retardates and normals. InN. R. Ellis (Ed.), International Review of Research in Mental Retardation (Vol. 4, pp. 1-32). New York: Academic Press. Elvevag, B., Fisher, J. E., Weickert, T. W., Weinberger, D. R., & Goldberg, T. E. (2004). Lack of false recognition in schizophrenia: A consequence of poor memory? Neuropsychologia, 42, 546-554. Farrant, A., Blades, M., & Boucher, J. (1998). Source monitoring by children with autism. Journal of Autism and Developmental Disorders, 28,43-50.

9.

PSYCHOLOGICAL IMPAIRMENT

293

Femina, D. D., Yeager, C. A., & Lewis, D. 0. (1990). Child abuse: Adolescent records vs. adult recall. Child Abuse and Neglect, 14, 227-231. Fillmore, M. T., Vogel-Sprott, M., & Gavrilescu, D. (1999). Alcohol effects on intentional behavior: Dissociating controlled and automatic influences. Experimental and Clinical Psychopharmacology, 7, 372-378. Finlay, W. M. L., & Lyons, E. (2002). Acquiescence in interviews with people who have mental retardation. Mental Retardation, 40, 14-29. Gerson, L. W. (1978). Alcohol-related acts of violence: Who was drinking and where the acts occurred. Journal of Studies on Alcohol, 39, 1294-1296. Goldberg, T. E., Dodge, M., Aloia, M., Egan, M. F., & Weinberger, D. R. (2000). Effects of neuroleptic medication on speech disorganization in schizophrenia: Biasing associative networks toward meaning. Psychological Medicine, 30, 1123-1130. Goodwin, D. W. (1995). Alcohol amnesia. Addiction, 90, 315-317. Graham, K., & West, P. (2001). Alcohol and crime: Examining the link. InN. Heather, T. J. Peters, & T. Stockwell (Eds.). International handbook of alcohol dependence and problems (pp. 4394 70). Chichester: John Wiley & Sons. Gudjonsson, G. H. (2003). The psychology of interrogations and confessions: A handbook. Chichester, UK: Wiley & Sons. Gudjonsson, G. H., Murphy, G. H., & Clare, I. C. H. (2000). Assessing the capacity of people with intellectual disabilities to be witnesses in court. Psychological Medicine, 30, 307-314. Gunderson, E. W., Vosburg, S. K., & Hart, C. L. (2002). Does marijuana use cause long-term cognitive deficit? Journal of the American Medical Association, 287, 2652. Hala, S., Rasmussen, C., & Henderson, A.M. E. (2005). Three types of source monitoring by children with and without autism: The role of executive function. Journal of Autism and Developmental Disorders, 35, 75-89. Harrison, L. D., Erickson, P. G., Adlaf, E., & Freeman, C. (2001). The drugs-violence nexus among American and Canadian youth. Substance Use & Abuse, 36, 2065-2086. Hasher, L., & Zacks, R. T. (1984). Automatic processing of fundamental information: The case of frequency of occurrence. American Psychologist, 39, 1372-1388. Heather, N., Peters, T. J., & T. Stockwell, T. (Eds.). (2001). International handbook of alcohol dependence and problems. Chichester: John Wiley & Sons. Heishman, S. J., Arasteh, K., & Stitzer, M. L. (1997). Comparative effects of alcohol and marijuana on mood, memory, and performance. Pharmacology, Biochemistry & Behavior, 58, 93-101. Henry, L.A., & Gudjonsson, G. H. (1993). Eyewitness memory and suggestibility in children with mental retardation. American Journal on Mental Retardation, 104, 491-508. Henry, L.A., & Gudjonsson, G. H. (2003). Eyewitness memory, suggestibility, and repeated recall sessions in children with mild and moderate intellectual disabilities. Law and Human Behavior, 27,481-505. Hildebrandt, H., Brokate, B., Eling, P., & Lanz, M. (2004). Response shifting and inhibition, but not working memory, are impaired after long-term heavy alcohol consumption. Neuropsychology, 18,203-211. Hodges, J. R. (2000). Memory in the dementias. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 441-459). New York: Oxford University Press. Hyman, I. E., Jr., & Billings, J. F. (1998). Individual differences and the creation of false childhood memories. Memory, 6, 1-20. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114(1), 3-28. Johnson, M. K., & Raye, C. L. (1998). False memories and confabulation. Trends in Cognitive Sciences, 2(4), 137-145. Julien, R. M. (2005). A primer of drug action: A comprehensive guide to the actions, uses, and side effects of psychoactive drugs (1Oth ed.). New York: Worth Publishing.

294

SORACI ET AL.

Kassin, S.M. (1997). The psychology of confession evidence. American Psychologist, 52, 221-233. Kassin, S. M. (2005). On the psychology of confessions: Does innocence put innocents at risk? American Psychologist, 60, 215-228. Kassin, S.M., Tubb, V. A., Hosch, H. M., & Memon, A. (2001). On the "general acceptance" of eyewitness testimony research: A new survey of the experts. American Psychologist, 56, 405-416. Kebbell, M. R., & Hatton, C. (1999). People with mental retardation as witnesses in court: A review. Mental Retardation, 37, 179-187. Keefe, R. S. E., Arnold, M. C., Bayen, U. J., & Harvey, P. D. (1999). Source monitoring deficits in patients with schizophrenia: a multinomial modelling analysis. Psychological Medicine, 29(4), 903-914. Kindt, M., & van den Hout, M. (2003). Dissociation and memory fragmentation: Experimental effects on meta-memory but not on actual memory performance. Behaviour Research and Therapy, 41, 167-178. Kirchner, T. R., & Sayette, M.A. (2003). Effects of alcohol on controlled and automatic memory processes. Experimental and Clinical Psychopharmacology, 11, 167-175. Marlatt, G. A., & Rosenhow, D. J. (1980). Cognitive processes in alcohol use: Expectancy and the balanced placebo design. InN. K. Mello (Ed.), Advances in substance abuse: Behavioral and biological research (pp. 155-199). Greenwich, CT: JAI Press. Martin, C., West, J., Cull, C., & Adams, M. (2000). A preliminary study investigating how people with mild intellectual disabilities perform on the Rivermead Behavioural Memory Test. Journal of Applied Research in Intellectual Disabilities, 13, 186-193. McDaniel, W. F., Foster, R. A., Compton, D. M., & Courtney, A. S. (1998). A strategy for screening memory functions in individuals with mental retardation. Research in Developmental Disabilities, 19, 317-325. McNally, R. J. (2003). Remembering trauma. Cambridge: Belknap Press. Michel, M. K., Gordon, B. N., Ornstein, P. A., & Simpson, M.A. (2000). The abilities of children with mental retardation to remember personal experiences: Implications for testimony. Journal of Clinical Child Psychology, 29, 453-463. Miller, W. R., Hertel, P., Saucedo, C., & Hester, R. K. (1994). Effects of alcohol and expectancy on episodic memory in individuals reporting alcoholic blackouts. Experimental and Clinical Psychopharmacology, 2, 161-166. Millward, C., Powell, S., Messer, D., & Jordan, R. (2000). Recall for self and other in autism: Children's memory for events experienced by themselves and their peers. Journal of Autism and Developmental Disorders, 30, 15-28. Milne, R. (1999). Interviewing children with learning disabilities. In A. Memon & R. Bull (Eds.), Handbook of the psychology of interviewing (pp. 165-180). Chichester, UK: John Wiley & Sons. Minshew, N.J., & Goldstein, G. (2001). The pattern of intact and impaired memory functions in autism. Journal of Child Psychology and Psychiatry, 42, 1095-1101. Moore, D. (2003). The anthropology of drinking. InN. Heather, T. J. Peters, & T. Stockwell (Eds.), International handbook of alcohol dependence and problems (pp. 471-487). Chichester, UK: John Wiley & Sons. Moritz, S., Heeren, D., Andresen, B., & Krausz, M. (2001). An analysis of the specificity and the syndromal correlates of verbal memory impairments in schizophrenia. Psychiatry Research, 101,23-31. Moritz, S., & Woodward, T. S. (2002). Memory confidence and false memories in schizophrenia. The Journal of Nervous and Mental Disease, 190(9), 641-643. Moritz, S., Woodward, T. S., Cuttler, C., Whitman, J. C., & Watson, J. M. (2004). False memories in schizophrenia. Neuropyschology, 18, 276--283. Moritz, S., Woodward, T. S., & Ruff, C. C. (2003). Source monitoring and memory confidence in schizophrenia. Psychological Medicine, 33, 131-139.

9.

PSYCHOLOGICAL IMPAIRMENT

295

Morrison, A. P., & Haddock, G. (1997). Cognitive factors in source monitoring and auditory hallucinations. Psychological Medicine, 27(3), 669-679. Nelson, T. 0., McSpadden, M., Fromme, K., & Marlatt, G. A. (1986). Effects of alcohol intoxication on metamemory and retrieval from long-term memory. Journal of Experimental Psychology: General, 115, 257-274. Parkin, A. (1987). Memory and amnesia: An introduction. Oxford: Basil Blackwell. Platt, R. D., Lacey, S.C., Iobst, A. D., & Finkelman, D. (1998). Absorption, dissociation, and fantasy-proneness as predictors of memory distortion in autobiographical and laboratorygenerated memories. Applied Cognitive Psychology, 12, S77-S89. Ray, S., Bates, M. E., & Bly, B. J. (2004). Alcohol's dissociation of implicit and explicit memory processes: Implications of a parallel distributed processing model of semantic priming. Experimental and Clinical Psychopharmacology, 12, 118-125. Read, J.D., Desmarais, S., Rojas, E., & VanderVeen, S. (2005, March). Beyond skepticism: Juror knowledge and expert testimony on eyewitness issues. Paper presented at the meetings of the American Psychology-Law Society, La Jolla, CA. Read, J. D., Yuille, J. C., & Tollestrup, P. (1992). Recollections of a robbery: Effects of arousal and alcohol upon recall and person identification. Law and Human Behavior, 16, 425-446. Reich, R. R., Goldman, M.S., & Noll, J. A. (2004). Using the false memory paradigm to test two key elements of alcohol expectancy theory. Experimental and Clinical Psychopharmacology, 12, 102-110. Renner, P., Klinger, L. G., & Klinger, M. R. (2000). Implicit and explicit memory in autism: Is autism an amnesic disorder? Journal of Autism and Developmental Disorders, 30, 3-14. Reyna, V. F., & Brainerd, C. J. (1995). Fuzzy-trace theory: An interim synthesis. Learning and Individual Differences, 7, 1-75. Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory and Cognition, 21 (4), 803-814. Roediger, H. L., McDermott, K. B., & Goff, L. M. (1997). Recovery of true and false memories: Paradoxical effects of repeated testing. In M. A. Conway (Ed.), Recovered memories and false memories (pp. 118-149). Oxford: Oxford University Press. Roediger, H. L., Ill, Watson, J. M., McDermott, K. B., & Gallo, D. A. (2001). Factors that determine false recall: A multiple regression analysis. Psychonomic Bulletin & Review, 8, 385407. Schacter, D. L. (1996). Searching for memory: The brain, the mind, and the past. New York: Basic Books. Schacter, D. L. (1999). The seven sins of memory: Insights from psychology and cognitive neuroscience. American Psychologist, 54(3), 182-203. Schacter, D. L., & Dodson, C. S. (2002). Misattribution, false recognition and the sins of memory. In A. Baddeley, M. Conway &J.P. Aggleton, (Eds.), Episodic memory: New directions in research (pp. 71-85). London: Oxford University Press. Schacter, D. L., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger and older adults: The distinctiveness heuristic. Journal of Memory and Language, 40, 1-24. Spitz, H. H. (1966). The role of input organization in the learning and memory of mental retardates. InN. R. Ellis (Ed.), International Review of Research in Mental Retardation (Vol. 2, pp. 29-54). New York: Academic Press. Steele, C. M., &Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous effects. American Psychologist, 45, 921-933. Stracciari, A. S., Guarino, M., Crespi, C., & Pazzaglia, P. (1999). Transient amnesia triggered by acute marijuana intoxication. European Journal of Neurology, 6, 521-523. Sullivan, K., Carlin, M. T., Wakeford, Y., Hasel, L., &Jakway, A. (2005, March). False memory in individuals with mental retardation. Poster presented at the 38th Annual Gatlinburg Conference on Research and Theory in Intellectual & Developmental Disabilities, Annapolis, MD.

296

SORACI ET AL.

Sweeney, D. F. (1989). Alcohol versus Mnemosyne-Blackouts. Journal of Substance Abuse Treatment, 6, 159-162. Tager-Flusberg, H. (1991). Semantic processing in the free recall of autistic children: Further evidence for a cognitive deficit. British Journal of Developmental Psychology, 9, 417-430. Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences ("absorption"), a trait related to hypnotic susceptibility. Journal of Abnormal Psychology, 83, 268-277. Toichi, M., & Kamio, Y. (2001). Verbal association for simple common words in high-functioning autism. Journal of Autism and Developmental Disorders, 31, 483-490. Toichi, M., & Kamio, Y. (2002). Long-term memory and levels-of-processing in autism. Neuropsychologia, 40, 964-969. Toichi, M., & Kamio, Y. (2003). Long-term memory in high-functioning autism: Controversy on episodic memory in autism reconsidered. Journal of Autism and Developmental Disorders, 33, 151-161. Tracy, J. L., & Bates, M. E. (1999). The selective effects of alcohol on automatic and effortful memory processes. Neuropsychology, 13, 282-290. Turnure, J. E. (1985). Communication and cues in the functional cognition of the mentally retarded. InN. R. Ellis & N. W. Bray (Eds.), International review of research in mental retardation (Vol. 13, pp. 43-77). New York: Academic Press. Vicari, S., Bellucci, S., & Carlesimo, G. A. (2000). Implicit and explicit memory: A functional dissociation in persons with Down syndrome. Neuropsychologia, 38, 240-251. Vinogradov, S., Willis-Shore, J., Poole, J. H., Marten, E., Ober, B. A., & Shenaut, G. K. (1997). Clinical and neurocognitive aspects of source monitoring errors in schizophrenia. American Journal of Psychiatry, 154(11), 1530-1537. Watson, J. M., Balota, D. A., & Sergent-Marshall, S.D. (2001). Semantic, phonological, and hybrid veridical and false memories in healthy older adults and in individuals with dementia of the Alzheimer type. Neuropsychology, 15, 254-268. Weingartner, H. J., Putnam, F., George, D. T., & Ragan, P. (1995). Drug-state dependent autobiographical knowledge. Experimental and Clinical Psychopharmacology, 3, 304-307. Weiss, A. P., Dodson, C. S., Goff, D. C., Schacter, D. L., & Heckers, S. (2002). Intact suppression of increased false recognition in schizophrenia. American Journal of Psychiatry, 159(9), 1506-1513. Wetzel, C. D., Janowsky, D. S., & Clopton, P. L. (1982). Remote memory during marijuana intoxication. Psychopharmacology, 76, 278-281. Wilkinson, C., & Hyman, I. E., Jr. Individual differences related to two types of memory errors: Word lists may not generalize to autobiographical memory. Applied Cognitive Psychology, 12, S29-S46. Williams, D. L., Goldstein, G., & Minshew, N.J. (2005). Impaired memory for faces and social scenes in autism: Clinical implications for memory dysfunction. Archives of Clinical Neuropsychology, 20, 1-15. Wilson, G. T. (1982). Alcohol and anxiety: Recent evidence on the tension reduction theory of alcohol and abuse. In K. R. Blankenstein & J. Polivy (Eds.), Self-control and self-modification of emotional behavior (pp. 117-141). New York: Plenum. Winograd, E., Peluso, J. P., & Glover, T. A. (1998). Individual differences in susceptibility to memory illusions. Applied Cognitive Psychology, 12, S5-S27. Yuille, J. C., & Daylen, J. (1998). The impact of traumatic events on eyewitness memory. In C. P. Thompson, D. J. Herrmann,]. D. Read, D. Bruce, D. G. Payne, & Michael P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 155-178). Mahwah, NJ: Lawrence Erlbaum Associates. Yuille, J. C., Hunter, R., Joffe, R., & Zaparniuk, J. (1993). Interviewing children in sexual abuse cases. In G. S. Goodman and B. Bottoms (Ed.), Child victims, child witnesses: Clinical, developmental, and legal implications (pp. 95-115). New York: Guilford Press.

9.

PSYCHOLOGICAL IMPAIRMENT

297

Yuille, J. C., & Tollestrup, P. A. ( 1990). Some effects of alcohol on eyewitness memory. Journal of Applied Psychology, 75, 268-273. Yuille, J. C., Tollestrup, P. A., Marxsen, D., Porter, S., & Herve, H. F. M. (1998). An exploration of the effects of marijuana on eyewitness memory. International Journal of Law and Psychiatry, 21, 117-128. Zacks, R. T., & Hasher, L. (2002). Frequency processing: a twenty-five year perspective. In P. Sedlmeier & T. Betsch (Eds.), Frequency processing and cognition (pp. 21-36). Oxford: Oxford University Press. Zoellner, L. A., Foa, E. B., Brigidi, B. D., & Przeworski, A. (2000). Are trauma victims susceptible to "false memories"? Journal of Abnormal Psychology, 109(3), 517-524.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

10 Recovered Memories Steven M. Smith and David H. Gleaves Texas A&M University

The subject of recovered memories is at the center of the so-called false memory debate, sometimes referred to as the "recovered memory debate," and even the "memory wars," all terms referring to an intellectual and emotional conflict that has captured both scholarly and popular interest and concern. A web search using the search term recovered memory yields well over a million results dealing with the subject. A recent bibliography of articles listed over 800 published papers relevant to the topic. In many ways the subject of recovered memories seems to be less about memory per se and more about validation, that is, on the one hand, validation of victims' memories of traumatic events that truly occurred, and, on the other hand, validation of the innocence of people who have been wrongfully accused of being abusers. On the front lines of the memory wars have been clinicians and therapists who typically (but by no means always) have been more vocal about concerns with victims of abuse who recover memories, versus experimentalists who typically (but not always) have expressed more concern over those who have been wrongfully accused based on false memories. Because both of these concerns are so important, we will try to address both as we discuss what recovered memories are, what is known about them, what is controversial about recovered memories, and what is not yet known about them. Here we ask a number of questions to clarify what is known about recovered memories, what is not known, and what needs to be known. Do recovered memories exist? If so, what causes them, and how common are they? Does repression exist? Are recovered memories valid representations of past events, and are they less (or more) accurate than are continuous memories, that is, memories that were never inaccessible? Is recovery of traumatic memories different from recovery of nontraumatic memories? Another goal is to clarify a number of terms and concepts, such as recovered memories, suppression, inhibition, interference, normal forgetting, repression, dissociation, discovered memories, memory triggers, amnesia, trauma, reminiscence, mood dependence, and hypnosis. We will offer a taxonomy of a number of phenomena related to recovered

299

300

SMITH AND OLEAYES

memories, distinguishing, for example, continuous memories from recovered and discovered memories. We review the history of these issues, discuss the available research, and point to issues that should be addressed by future research. In some cases, we make policy recommendations about recovered memories.

HISTORICAL REVIEW

Apparent amnesia and the subsequent recovery of memory were originally studied by researchers and writers such as Sigmund Freud and Pierre Janet. Freud's (1896/1962) "Aetiology of Hysteria" is perhaps the best known early discussion of this topic. In his 1896 address, Freud presented his "seduction theory," in which he argued that "hysteria" resulted from repressed memories of childhood sexual trauma, and that memory recovery (i.e., bringing these memories into consciousness) would lead to alleviation of the hysterical symptom. He subsequently abandoned his seduction theory in favor of his theory of childhood sexual fantasy. Why Freud abandoned his theory and whether or not his initial discoveries were genuine has been a topic of recent debate (see Gleaves and Hernandez, 1999, for a discussion of these issues). In 1889, Janet published fautomatisme psychologique, in which he reported his investigations into the processes that linked traumatic experiences and psychopathology. Janet is viewed by many (e.g., van der Kolk & van der Hart, 1989) as the first to systematically study dissociation as the psychological process with which people react to overwhelming experiences. Like Freud, Janet also studied "hysteria" and argued that the primary mode of adaptation in hysterical individuals was the dissociation of memories related to frightening experiences. After Freud and Janet, the concept of amnesia for trauma continued to receive attention, particularly in the literature on wartime trauma. Sargant and Slater (1941) described a World War II account of 1,000 consecutive admissions to a neurological unit. Over 14% of the sample exhibited amnesia, and the extent of amnesia appeared to be associated with the severity of the trauma (e.g., 35% of those exposed to severe stress exhibited significant amnesias). The memory disturbances reported in the wartime literature included all subtypes of amnesia. In describing many of these psychiatric reactions to war trauma, Grinker and Spiegel (1945) wrote, "There may be total amnesia, including both events on the battlefield and the patient's previous life, or memory for part of the battle experience may be retained, with a gap involving the actual precipitating traumatic factors and the events which followed. The majority of patients make persistent attempts to recover their lost memories, and in many instances their efforts may be successful without an aid from the therapist" (p. 10). Retrieval of these memories of trauma was described as essential to recovery and was accomplished through the use of psychotherapy, hypnosis (Kardiner &Spiegel, 1947), or even drugs such as sodium pentothal (see also Grinker & Spiegel, 1945). This topic became controversial in the 1980s and 1990s when reports of amnesia and memory recovery associated with child sexual abuse became more common (Sivers et al., 2002).

10.

RECOVERED MEMORIES

301

TERMS AND DEFINITIONAL ISSUES

The terms and concepts used in discussions of recovered memories are of great importance, both because the language must be used clearly and unambiguously, and because some nomenclature can carry along baggage or implied meanings that are not intended to be included by a verbal label. An example is the term false memory, a term that has been used alternatively to refer to misremembered details or completely fabricated memories, two very different phenomena whose effects can vary greatly in terms of their consequences (DePrince, Allard, Oh, & Freyd, 2004). Although the term recovered memory is most frequently used in the context of trauma or abuse, it more generally refers to a conscious memory of an event or an episode that one was previously unable to remember. The generally accepted criteria for a recovered memory are that (1) the event or episode in question must have been successfully encoded, (2) memory for the encoded events must be inaccessible for a time, and (3) conscious memory must occur sometime after the period of inaccessibility. To some, the term recovered memory also suggests that (1) the events in question were traumatic, (2) repression was the cause of temporary inaccessibility, and (3) unconscious processes work to repress and recover lost memories of trauma. It is important, however, to distinguish these implications from the definition of recovered memories, which does not involve trauma, repression, or unconscious mechanisms. A recovered memory can be contrasted with a continuous memory, that is, a memory of an event or episode that was never inaccessible. Similar to the term recovered memories is the term discovered memories, that is, surprising or even shocking memories of experiences that have not been remembered for a long period of time. Although similar, the term discovered memory carries less conceptual baggage for some than the term recovered memory (e.g., Schooler, Ambadar, & Bendiksen, 1997; Schooler, 2001), because it does not imply that the memories in question were repressed, only that they were not remembered. The surprising and shocking nature of discovered memories can be understood sometimes in terms of sudden changes in an individual's understanding of the meaning of the remembered events. Encoding refers to forming a mental representation of an event and having that representation stored in long-term memory. Encoded memories of events are said to be available in memory. An experience that is encoded and later forgotten might be recoverable, whereas experiences that are not successfully encoded could never be recovered. Although the initial encoding of an experience may occur immediately, a period of consolidation, lasting hours or even days after the event, is believed by some to be necessary for the encoded event to become stored in a more permanent fashion. Accessibility refers to one's ability to bring a thought or a memory into one's conscious mind. Access is possible only for memories that are available, that is, memories that are already encoded in long-term memory. Access of a memory means the same thing as retrieval of a memory. Retrieval is a process involved in explicit memory in which memories of encoded experiences are brought to mind and is similar to the term recollection, which refers to consciously inspecting or reexperiencing parts or aspects of previously encoded episodes. A memory can be said to be inaccessible if attempts to retrieve a

302

SMITH AND GLEAYES

memory fail. Not everyone, however, adopts the same criteria for such failures. That is, when motivation is present, and efforts to retrieve a memory are persistent and focused, there is clearer evidence for inaccessibility than if retrieval efforts are feeble or nonexistent. It is difficult to determine that a memory was inaccessible if appropriate cues were never provided, or if no attempts to retrieve the memory in question occurred. We will return to this point below. A number of terms in the area of recovered memory deal with various forms of forgetting, including suppression, inhibition, amnesia, dissociation, repression, and normal forgetting. Here we briefly explain the meanings of these important terms, and we try to distinguish each term's denotations from its connotations. The terms suppression and inhibition both refer to cognitive mechanisms for excluding ideas and experiences from one's conscious mind. Suppression refers to a temporary banishing of thoughts or memories from one's conscious mind through a deliberate act of one's will (e.g., Wenzlaff & Wegner, 2000). Inhibition, in the memory domain, is the long-term forgetting that results from repeated bouts of suppression (e.g., Anderson, Bjork, & Bjork, 1994; Anderson & Spellman, 1995; Anderson & Green, 2001). Use of the term normal forgetting implies that there are also special mechanisms that do not involve normal forgetting, that is, special mechanisms that can cause periods of amnesia for certain memories, mechanisms beyond simple interference (competition and confusion among memories), inhibition (retrieval difficulty or retrieval failure), and decay (loss of memory over time) of memories. One special mechanism is referred to as dissociation, a disruption in the integration of various parts of the mind. Dissociation separates groups of mental processes that are normally integrated, resulting in independent functioning of these processes, such as an abnormal separation of affect from cognition. In most general terms, dissociation is the opposite of association. The current version of the Diagnostic and Statistical Manual of Mental Disorders (DSM; American Psychiatric Association, 2001) defines psychological dissociation as "a disruption in the usually integrated functions of consciousness, memory, identity, or perception" (p. 519). It may also be defined as a complex mental state during which there is a change in a person's consciousness that disturbs the normally connected functions of identity, memory, thoughts, feelings, and experiences. A distinction is made between nonpathological dissociation (e.g., absorption and imaginative involvement) and pathological dissociation (e.g., amnesia, depersonalization, and derealization; Waller, Putnam, & Carlson, 1996). There are many types of amnesia, but all may be characterized by an extraordinary degree of forgetting, such as a loss of memory or a gap in one's memory. The term can be confusing, not only because there are various types of amnesia, but also because the term sometimes implies conclusions about amnesia that may be in contention, such as the amnesia's reversibility, or how amnesia differs from normal forgetting. In most general terms, amnesia refers simply to an inability to remember. The DSM defines dissociative amnesia (previously known as psychogenic amnesia) as a "reversible memory impairment in which memories of personal experience cannot be retrieved in verbal form" (APA, p. 520). The fact that it is reversible implies that the lost memories can be recovered. This is in contrast to amnesias of strictly organic bases, which are more generally not reversible.

10.

RECOVERED MEMORIES

303

In the DSM and trauma literature, several different types of amnesia are recognized.

Localized refers to an inability to remember events that occurred during a circumscribed period of time (e.g., an hour or two surrounding a specific trauma). Selective refers to being able to only recall bits and pieces of events that took place in a defined period of time. For example, an abuse victim may recall only some parts of a series of events around his or her abuse. Generalized amnesia would be diagnosed when a person's amnesia encompasses his or her entire life. There are sometimes rare but genuine cases who appear as a John or Jane Doe. Continuous amnesia occurs when the individual has no memory for events beginning from a certain point in the past continuing up to the present. Systematized amnesia is characterized by a loss of memory for a specific category of information. A person with this disorder, for example, might be missing all memories about one specific family member. Amnesia for a series of abusive experiences would qualify as systematized amnesia. It is important to note that the apparent amnesia associated with trauma generally only applies to declarative (autobiographical knowledge) memory and not to nondeclarative (feelings, sensations, and habits) memory. In fact, the effects of trauma may be the opposite for these two types of memory. As Jacobs and Nadel (1998) noted, "It [traumatic stress] can cause amnesia for the autobiographical context of the stressful event and hypermnesia for the emotional memories produced by them" (p. 1115). Thus, a person might feel the intense emotion and pain associated with a trauma but not understand the origins of those feelings. One special mechanism postulated to cause amnesia for traumatic experiences is repression, an unconscious defense mechanism that supposedly banishes unacceptable ideas, fantasies, affects, or impulses from consciousness. Even when an experience cannot be consciously recalled, memory of repressed material is theorized to emerge implicitly, that is, in certain responses or behaviors related to the inaccessible memories. Repression is a key concept in psychodynamic theory, explaining how the unconscious wards off from conscious awareness threatening and uncomfortable content. That repressed memories might subsequently enter consciousness, perhaps as a product of psychotherapy, has historically served as the basis of recovered memories. Although there is now ample scientific evidence that supports the existence of recovered memories, the same cannot be said for the existence of this unconscious form of repression, which has proved elusive to scientific scrutiny. Whereas the concept of repression is embraced by some (e.g., Edelson, 1990), others dismiss the concept completely (e.g., Holmes, 1990; Kihlstrom, 2002). Still others have endeavored to revamp the concept of repression, showing that straightforward cognitive processes, such as memory inhibition (e.g., Anderson, Bjork, & Bjork, 1994; Anderson & Green, 2001), interference and retrieval bias (Smith, Gleaves, et al., 2003), and implicit memory (e.g., Bower, 1990), can explain ways that unwanted memories can be banished from consciousness. These cognitive explanations do not rely on an unconscious mechanism that manages emotional memories by keeping them out of conscious awareness. In general, the term trauma refers to a highly stressful event. However, depending on how it is defined, the term may also be quite subjective. That is, what is traumatic to one

304

SMITH AND GLEAVES

individual may not be traumatic to another. In an attempt to maintain the reliability of the PTSD diagnosis, the DSM offers an operational definition of a traumatic event that "involves actual or threatened death or serious injury, or other threat to one's physical integrity; or witnessing an event that involves death, injury, or a threat to the physical integrity of another person; or learning about unexpected or violent death, serious harm, or threat of death or injury experienced by a family member or other close associate .... The person's response to the event must involve intense fear, helplessness, or horror" (p. 463). Examples include, but are not limited to, military combat, violent personal attack, natural or man-made disasters, and torture. For children, sexually traumatic events may include age-inappropriate sexual experiences without violence or injury. The term betrayal trauma (Freyd, 1996) refers to a kind of trauma that is independent of the reaction of the event. According to Freyd (2003), "betrayal trauma occurs when people or institutions we depend on for survival violate us in some way."

LABORATORY RESEARCH AND THEORETICAL UNDERPINNINGS

A good deal of empirical research on memory for events has examined mechanisms for long-term forgetting. These mechanisms can be grouped into two general categories, mechanisms based upon inadequate encoding of memories and those based on processes that take effect after encoding. Forgetting due to inadequate encoding of experiences can result from a number of causes, such as inadequate attention to an event, or apprehending only superficial aspects of events. Experiences that are inadequately encoded are said to be unavailable and cannot be recovered. Successful encoding of an event can be verified if the event is explicitly or consciously remembered after some period of time has passed; generally, memory after more than a few minutes is considered evidence of an adequately encoded episode. It should be noted, however, that consolidation processes, which occur after initial encoding (such as during sleep the night after an experience), may be needed to make an event memory permanently available. That is, a memory that was initially encoded, but not successfully consolidated, might not be recoverable. The most commonly researched cognitive mechanisms that could cause forgetting of successfully encoded memories are interference and inhibition. In interference, events are forgotten because competing memories come to mind. Experimental psychology boasts a long and rich history of laboratory research on the nature of memory interference (e.g., Anderson & Neely, 1996; Postman & Underwood, 1973). Inhibition, in contrast, causes forgetting by making memories less accessible independently of competing memories. Although the two are difficult to distinguish experimentally, there is now considerable evidence of forgetting that is explainable by an inhibitory forgetting mechanism, but not by interference (e.g., Anderson & Spellman, 1995; Anderson & Green, 2001). A single period of thought suppression may be immediately followed by a "rebound" period during which a formerly suppressed thought becomes hyperaccessible (e.g., Wegner, 1989). Increasing the number of times that someone tries not to think of a response has been found to decrease the chance that the response will be remembered (e.g., Anderson

10.

RECOVERED MEMORIES

305

& Green, 2001). Studies that show this decreased accessibility find that responses are remembered more poorly than would be expected even when independent memory cues are used, which is suggestive evidence that the forgetting effects are caused by inhibition rather than distraction or interference. Individually, each of these mechanisms could potentially play a role in the occurrence of recovered memories. Collectively, these cognitive mechanisms fix the limits of the spectrum of phenomena that have been termed normal forms of forgetting. Those limits, however, might be surprising to some. Research on most normal forgetting phenomena, such as directed forgetting, retrieval-induced forgetting, part-list cuing inhibition, and inhibition from the think/no-think method show that experimental manipulations can cause recall levels to drop by about 5-20%. Such effects, although reliable and well replicated, do not seem to account for the magnitude of forgetting effects one might associate with repression or amnesia. Recently, however, reversible laboratory forgetting effects of 35-60% have been found (Smith, 1995; Smith et al., 2003; Smith and Moynan, 2004) with the use of experimental methods based on interference and cue-dependent memory mechanisms. Memory of the Forget Group for the three targeted word lists has been found to be up to 62% worse than recall of those three lists by the Control Group. These forgetting effects are rather remarkable for simple laboratory behavioral procedures involving a single hour-long session, yet they clearly do not involve any special cognitive or affective mechanisms. Perhaps even more remarkable is the finding that those same laboratory interference procedures produce large forgetting effects even for a list of expletives (Smith & Moynan, 2004); that is, subjects were made to forget that they read and wrote down an entire list of curse words! This research is described further below. Another body of experimental research has been devoted to memory recovery. Spontaneous recovery refers to the recurrence of a conditioned response after extinction has occurred; it is one of the most basic phenomena in Pavlovian conditioning. Extinction refers to the finding that if conditioned and unconditioned stimuli are not paired for many trials, an individual will stop giving the conditioned response. Recovery can occur after a period of nonexposure to the conditioned stimulus, and it is called spontaneous when the causes of the recovered response may not be obvious. Spontaneous recovery is one laboratory analogue of a recovered memory in that encoding (conditioning) is initially successful, then a period of inaccessibility occurs (extinction), followed by recovery of the original memory (spontaneous recovery). Examples of spontaneous recovery include recovery of extinguished fears and phobias (e.g., Jacobs & Nadel, 1985) and return of extinguished alcohol cravings (Rodd, Bell, Sable, Murphy, & McBride, 2004). Hypermnesia and reminiscence are also laboratory analogues of recovered memories (e.g., Erdelyi, 1990; Shaughnessy, 2002); both terms refer to findings that people sometimes recall more after repeated recall attempts. Reminiscence refers to the elicitation of memories that were not remembered earlier, and hypermnesia is the related finding that repeated tests can increase the total amount recalled. Events forgotten on one test might be recovered, that is, successfully recalled on a later test. Hypermnesia and reminiscence are some of the oldest reported phenomena in the experimental literature on memory (e.g., Ballard, 1913) and have been shown with a wide variety of materials,

306

SMITH AND GLEAVES

including lists of words, pictures, and stories (Payne, 1987). Although there are different theories of the causes of reminiscence and hypermnesia, most are based on the notion that multiple retrieval attempts do not all produce the same memories. Therefore, later recall attempts sometimes will bring memories to mind that were not previously remembered. Tip-of-the-tongue (TOT) states, and recovery from them, represent clear everyday examples of blocked and recovered memories. The TOT is defined as a situation in which one cannot recall a word or name that seems to be in memory, and it feels as if recall of the missing name or word is imminent (Smith, 1994). Most TOT experiences are thought to occur for names and words that are actually known, but are temporarily and inexplicably inaccessible (e.g., A. Brown, 1990; Burke, McKay, Worthley, & Wade, 1992). Most naturally occurring TOTs are eventually resolved, some after a short time, but some take days or longer to resolve. The cognitive mechanisms that give rise to TOT states include memory blocks (e.g., Jones, 1989; Jones & Langford, 1987; Choi & Smith, 2005) and incomplete activation of memories (e.g., A. Brown, 1990; Burke et al., 1992). Both of these theories acknowledge that TOTs are surprising retrieval failures for memories that have been successfully encoded and that TOT states are often accompanied by partial retrieval of the target memories and by strong feelings that the target memory, though temporarily inaccessible, is actually available (encoded) in memory. It is also important to note that illusory TOTs can occur (e.g., Schwartz, Travis, Castro, & Smith, 2000), that is, strong feelings that never-encoded memories are actually available. These false metamemory experiences and others can result from an inappropriate use of inferential mechanisms, such as cue familiarity (e.g., Metcalfe, Schwartz, & Joachim, 1993; Schwartz & Metcalfe, 1992), social pressure (Widner, Smith, & Graziano, 1996), or retrieval of material related to the target (e.g., Koriat, 1993; Schwartz & Smith, 1997). Another phenomenon studied in the experimental literature is mood-dependent memory, a type of cue-dependent memory. Cue dependence is the reliance of memory on appropriate hints or associations, typically stimuli that were present or that were thought about when encoding of the events in question originally took place. Memory accuracy is generally better when the conditions of the original encoding are better reinstated (e.g., McGeoch & Irion, 1952), including reinstatement of verbal cues (e.g., Light & Carter-Sobell, 1970; Tulving & Thomson, 1973), encoding environments (e.g., Smith, 1979; Smith, Glenberg, & Bjork, 1978; Smith & Vela, 2001), drug states (e.g., Eich & Birnbaum, 1982), and mood states (e.g., Bower, 1981; Eich & Metcalfe, 1989). It is a possibility that stressful and traumatic experiences are accompanied by unusual emotional states, and those rare states, once reexperienced, might trigger memories that would otherwise be inaccessible. Such an explanation of recovered memories, although logically consistent, has not been experimentally investigated because of ethical problems with inducing such states in experimental volunteers. This mood-dependent theory, if true, would apply only to cases in which memories were experienced and encoded in unusual emotional states. Hypnosis is often defined as a state of inner absorption, concentration, and focused attention. It is also a clinical tool used to treat a variety of problems. Hypnosis is sometimes used in the treatment of dissociative disorders (Maldonado, Butler, & Spiegel,

10.

RECOVERED MEMORIES

307

2002) and has been studied as a tool for retrieval of memories (see Kihlstrom & Barnhardt, 1993). However, in 1985, a committee commissioned by the American Medical Association cautioned against the systematic use of hypnosis for recollection for both its unreliability (the possibility for example of confabulation) and its potential to create vivid false memories with an artificially induced sense of certainty. "The Council finds that recollections obtained during hypnosis can involve confabulations and pseudomemories and not only fail to be more accurate, but actually appear to be less reliable than non hypnotic recall" (AMA, 1985, p. 1918). Because of this potential problem, many states in the United States and several countries now confine legal testimony to that obtained prior to any systematic hypnotic treatment. Despite these concerns, in naturalistic research, use of hypnosis has not been found to be substantially associated with recovered memories of sexual abuse (Leavitt, 2001). Consistent with classical models of learning, stimuli associated with original conditioning events may also elicit the same response as that event. Memories may be triggered after a period of amnesia by situations that are similar to the original traumatic event, general emotional activation, relationships, hearing about others' recent traumatic events, sexual encounters, or seeing someone associated with the trauma or with the time period of the trauma (Brown, Scheflin, & Hammond, 1998; Harvey, 1999; Schooler, 1999). Interestingly, some research has found that recovered memories are least likely to occur in the context of psychotherapy (Brown et al., 1998). When they do occur in therapy, recovered memories are frequently additional memories recovered during work on previously known memories. Exposure treatment sessions, by their very nature, are likely to reactivate additional details of reported memories and trigger previously unrecalled memories (Leskin et al., 1998). Currently, there is no evidence that recovered memories, particularly for the central themes of traumatic events, are any more or less accurate than continuous memories (Brown et al., 1998; Williams, 1995; Dalenberg, 1996; Smith, Gleaves, et al., 2003). However, as described above, extreme trauma can disrupt the storage of the spatialtemporal aspects of a memory. In addition, the process by which somatic fragments are reconstructed into a coherent narrative is imperfect and subject to influence. jacobs and Nadel (1998) described the circumstances, according to their model of the psychophysiology of reconstructed memory, under which recovered traumatic memories are likely to be more or less accurate. They argued that specific memories would be most accurate if a person had experienced one massive traumatic event. With multiple, similar traumas there would be less but still some degree of confidence in specific memories. However, the authors argued that when an individual has experienced numerous different traumatic events (e.g., sexual assaults by different perpetrators, numerous traumatic war experiences, and/or several automobile accidents), they would have little confidence in the memory for specific events because the spatiotemporal context of the varying events could be inaccurately linked to the somata-emotional memory fragments. However, to conclude that the specifics of a memory are questionable is different from concluding that a person has not been traumatized (i.e., that a memory is completely false). In fact, it is the severity, variety, and extent of the trauma that theoretically cause the problems with accurate retrieval and reconstruction. It is also important to note that this model

308

SMITH AND GLEAVES

applies to memory for events that involved extreme amounts of stress. Memories of less stressful events may be blocked and recovered through differing mechanisms. The accuracy of such memories has not been established but has also not been empirically challenged. Experimental studies of blocked and recovered memories have shown large, consistent, and reversible memory blocking effects, demonstrating a laboratory analogue of recovered memories (e.g., Gerkens, 2004; Smith, 1995; Smith, Gleaves, et al., 2003; Smith, Gerkens, Choi, & Hull, 2003; Smith & Moynan, 2004). As noted above, this laboratory method is based upon straightforward cognitive mechanisms that include interference for memory blocking, and cue-dependent memory for recovery. In these experiments students' memories were tested for three targeted lists of words, say, a list of tools, a list of fish, and a list of sports, each containing a dozen items. Initially, subjects all viewed and wrote down the three targeted lists of words mixed in with 18 other lists of words (filler lists), which were also viewed and written down. Both the Control Group and the Forget Group, after a 20-minute delay, tried to recall the lists. The Forget Group, however, saw the 18 filler lists (but not the three target lists, tools, fish, and sports) several more times before the recall test. Both groups saw the three target lists only once, but the Forget Group saw the fillers many more times, causing a very large (30-60%) interference effect for the three targeted lists on the initial recall test. Importantly, the large forgetting effects in these studies were completely reversible. A person who, on the initial recall test, had forgotten about reading the list of tools, for example, was nonetheless able to recall the tool list and as many list items as participants who had continuous memories of the material when appropriate cues were provided. Thus, the method produces a very large and robust recovered memory effect without the use of any special or exotic mechanisms. It has been found that when these memory recovery effects occur, the accuracy (amount recalled) is equivalent for both continuous and recovered memories, and the levels of false memories are also equivalent (Gerkens, 2004; Smith, Gleaves, et al., 2003). The equivalence offalse memories for continuous and recovered memories has been found for completely suggested fabrications, episodically confused details (Smith, Gleaves, et al., 2003), and semantic confusion errors (Gerkens, 2004). Thus, the experimental evidence indicates that the levels of accurate and false memories are not different for continuous and recovered memories.

CLINICAL ISSUES

In this section we deal with issues that concern various clinical aspects of recovered memories. These issues include traumatic amnesia, betrayal trauma, effects of suggestibility in therapy, so-called memory work in psychotherapy, post-traumatic stress disorder, the use of exposure in therapy, and questionable aspects of treatments. It is important to distinguish the phenomenon of traumatic amnesia from the mechanism(s) responsible for it. There are numerous possible mechanisms for reported amnesia, including repression, suppression, dissociation, simple forgetting due to lack of

10.

RECOVERED MEMORIES

309

retrieval cues and interference, avoidance (conscious minimization and denial), lack of discussion (practice) due to shame, lack of state-dependent retrieval cues, and precipitous forgetting of nocturnal experiences. Sivers, Schooler, and Freyd (2002) discussed other possible mechanisms, including some specific to sexual abuse. In some cases, amnesia may be a direct physiological effect of trauma. Extreme stress appears to affect memory both at the neurological level and in terms of subjective experience. A physiological effect of traumatic stress can be a disruption of the mechanisms responsible for encoding the spatiotemporal context of memory (Jacobs & Nadel, 1998). With everyday experiences, autobiographical memories are created such that events, feelings, context, and the relations among them are linked. Moderate levels of stress can produce contextually linked, exceptionally vivid memories because of enhanced encoding of spatiotemporal information (due to the effect of stress hormones on hippocampus function) and enhanced encoding of emotional elements (due to hormonal effects on amygdala function). However, at high levels of stress, the hippocampus-based system appears to shut down, whereas the amygdala-based system is enhanced. This effect can lead to emotional and somatic information being stored separately from the spatiotemporal contexts of a memory (Jacobs & Nadel, 1998). Although the physiological effects of stress may explain some cases of amnesia, they most likely do not explain all cases. Particularly relevant are those associated more with betrayal experiences (e.g., Freyd, 1996). Some child molestation cases, such as the welldocumented case of Ross Cheit (see Freyd, 1996, or Hom, 1993), serve as an example. At the age of 38, Cheit recovered memories of being repeatedly molested between the ages of 10 and 13 years by a camp administrator. The original events, although also emotional, may not have been experienced with the same level and types of emotions as occur with life-threatening events. Cheit wrote (as described in Freyd, 1996), "The concept of trauma never seemed right to me. It didn't fit my story. There were no threats. I never sensed danger. I didn't fear him" (p. 11). Rather than terror, the original emotions were probably more akin to disgust, shame, and a sense of betrayal. Cheit reported feeling "frozen'' at the touch of the molester and later described his memories as disgusting and embarrassing. However, in some cases, both mechanisms may apply. That is, an event may be a betrayal and be life-threatening. Although the underlying mechanism should not be confused with the phenomenon itself, the underlying mechanism involved in a particular case may be important because it may determine the likelihood that a recovered memory is accurate. If, for example, amnesia was due to an encoding failure, it would be impossible to recover a memory of the event. If the event involved extreme levels of stress and the physiological effects described above, although the memory may be recoverable, its accuracy may be in question. In addition to the characteristics described above, betrayal trauma theory predicts that apparent amnesia for abuse will be higher when the relationship between perpetrator and victim involves closeness, trust, and/or care-giving. Freyd (1996) re-analyzed a number of relevant data sets (including the prospective sample reported by Williams [1994, 1995] and the retrospective samples reported by Cameron [1993] and Feldman-

310

SMITH AND OLEAYES

Summers and Pope [1994]) and stated that reported amnesia was more highly associated with incestuous abuse than nonincestuous abuse. Subsequently, Freyd, DePrince, and Zurbriggen (2001) found that physical and sexual abuse perpetrated by a caregiver was related to higher levels of self-reported memory impairment for the events compared with noncaregiver abuse. See Freyd (2003) for additional examples. It appears that men experience more nonbetrayal traumas than do women, whereas women experience more betrayal traumas than do men (Goldberg & Freyd, 2004). Women may seem to have more recovered memories of abuse because they may have more experiences with the sorts of events that lead to forgetting (DePrince & Freyd, 2002). In terms of suggestibility, Leavitt (1997, 1999) examined samples of persons reporting recovered memories of sexual abuse found with the use of the Gudjonsson Suggestibility scale. In the first study, Leavitt found that persons reporting recovered memories were in general less suggestible than a psychiatric comparison group. In the second study, he found that, following 2 years of treatment, the most suggestible of the patients did not recover memories. Those with the lowest levels of suggestibility recovered memories from the same therapeutic practices. These findings obviously do not support the position that suggestibility is responsible for reports of recovered memory. In a series of studies, McNally and colleagues examined how persons reporting recovered memories of sexual abuse differ from those reporting continuous memories (and in some cases, other conditions). McNally, Clancy, Schacter, and Pitman (2000) reported that recovered memory participants scored higher on absorption and dissociation than did those reporting either continuous memories or no abuse history. In a study of reality monitoring, McNally, Clancy, Barrett, and Parker (2005) reported that adults reporting either repressed or recovered memories of childhood sexual abuse (CSA) were less able to discriminate between words they had seen from words they had imagined seeing, as compared with adults reporting either never having forgotten their CSA or adults reporting no history of CSA. However, relative deficits in the ability to discriminate percepts from images were apparent only on some tests. One aspect of this topic that is frequently misunderstood is the reason that a therapist might engage in memory work or memory recovery as part of therapy. Critics of such therapy frequently argue that the justification for such practices is based purely on disproved Freudian concepts. Although Freud's original work was indeed based on the assumption that unconscious material needed to be made conscious, more modern trauma treatments, including those accepted as empirically supported (e.g., Keane, 1998), are actually based largely on Pavlovian models of learning as well as Mowrer's (1960) twofactor theory. That is, thinking about and processing memories of trauma is viewed as a form of imaginal exposure, in which extinction of conditioned emotional responses is the goal. Information/emotional processing theories of PTSD have also expanded on the basic learning model by adopting network models of emotion to understand how fearful associations are maintained and how they can be altered. Foa and colleagues (1989) conceptualized PTSD by using Lang's (1985) bioinformational model of emotions. In this model, information regarding associations between stimuli, responses, and meaning

10.

RECOVERED MEMORIES

311

is stored in memory networks. Foa and Kozak (1986) proposed that successful processing occurs when the memory network is fully activated (stimulus, response, and meaning elements are accessed, and arousal is experienced) and new information is incorporated into the network. Doing so thus involves processing the traumatic memory. As the above illustrates, although based on a variety of theoretical rationales, empirically supported therapies for traumatic memory involve exposing participants to memories and feelings associated with trauma. Research suggests that such interventions are associated with significant symptom reduction. Furthermore (and of most relevance to this chapter), processing of trauma memories quite frequently invokes additional memories. As Leskin, Kaloupek, and Keane (1998) wrote, "One of the expected consequences of exposure treatment sessions is reactivation of additional details for reported memories, as well as triggering previously unrecalled memories" (italics added) (p. 996). Thus, when memory recovery occurs in therapy (which is rare relative to other contexts in which it happens), it is most likely during processing of other existing memories. Based on the existing trauma treatment literature, the use of such treatments per se should not be a cause of concern. In fact, failing to use them in bona fide cases of trauma should be regarded as a greater concern. However, there would be at least two questionable uses of such treatments. One would be prematurely attempting to address traumatic memories in therapy. Phases of psycho-education and stabilization are generally necessary before trauma resolution should be attempted (Zerbe Enns et al., 1998), and premature attempts can clearly cause clients to de-compensate. Furthermore, comprehensive treatment guides by Chu (1998) and Gold (2000) make clear that addressing traumatic memories is far from all that one does with such patients. Thus, a treatment that focuses exclusively on recovery and exposure to traumatic memories should be regarded as inappropriate and potentially harmful, regardless of the authenticity of the memories. The second questionable use of such therapeutic approaches, of significant relevance to this chapter, would be in cases where there is no known history of abuse. That is, to help trauma patients recover memories of trauma so they would then be processed using the means described above would, at best, depart from how these treatments are supposed to be used. Zerbe Enns et al. (1998) provide a variety of guidelines for clients who may have experienced abuse, including the possible need for dealing with the uncertainty.

A TAXONOMY OF PHENOMENA

Because recovered memories are not a unitary phenomenon, but rather refer to several related phenomena and concepts, we present a taxonomy of phenomena that may help distinguish various types of experiences. Thus, blanket statements that appear to refer to all such phenomena can be replaced with more specific, and it is hoped, more accurate statements.

Single Event vs. Series of Abusive Experiences. Recovered memories can refer to either a single event, what Terr (1995) refers to as Type I trauma, or forgetting and

312

SMITH AND OLEAVES

recovery of more long-term repeated experiences, Terr's Type II trauma. For example, the mechanisms that underlie forgetting and recovery of a single physical trauma are undoubtedly different from those that underlie forgetting of repeated abusive experiences over a period of months or years. Whereas a vast body of research over the past century has examined and documented forgetting of single events, there is little experimental research on the forgetting and recovery of experiences that have been repeated over a long period of time, which would involve unethical levels of negative affect.

Traumatic Experiences vs. Post hoc Suppression. Recovered memories of abusive experiences can vary widely in terms of when the mechanism underlying forgetting of those experiences took place. Events originally experienced as highly traumatic, as described above, may be encoded in a manner in which the events are fragmented and inaccurately reconstructed in memory. Recovery of fragments of such traumatic memories may be accurate in content, whereas the themes around which the reconstruction is made can be confabulated. In contrast, events originally experienced as peculiar or unsettling might be subsequently discovered to be shameful or morally shocking; in such cases, the events might be suppressed, inhibited, or otherwise forgotten. There are no theories or empirical results that would lead to the conclusion that these forgotten experiences, if subsequently recovered, would be less accurate than continuous memories that were never inaccessible. Recovered vs. Discovered vs. Continuous Memories. The definition of recovered memories includes the criterion that such memories must undergo a period of inaccessibility to be considered recovered. In contrast, the definition of discovered memories includes no such requirement, such memories may have been inaccessible for a time, or they may have been remembered in a way that creates a new personal understanding of the events (e.g., an adult newly becoming aware that previously misunderstood childhood experiences actually constituted sexual abuse), whether or not the events could have previously been remembered. Continuous memories are those that have been accessible all along. Whereas both recovered and discovered memories can reveal new content that was previously unavailable, continuous memories do not. Traumatic vs. Emotional vs. Unemotional Memories. Virtually no controversy exists over the existence of recovered memories of unemotional experiences. Although there is some skepticism over the notion that emotional experiences can be forgotten and recovered, there have been research reports documenting that emotional events may be forgotten just as unemotional ones can. Whether highly traumatic experiences can be forgotten is more controversial. Memories of Gist vs. Detail. Memories for the gist of incidents differ from memories of the details of those episodes, and the same is true for recovery of memory gist and details. That is, even when the gist of an episode can be recalled, details may remain temporarily inaccessible. Likewise, forgotten gist can be accompanied by remembered details.

10. RECOVERED MEMORIES

313

Suggested vs. Triggered Recovered Memories. It is important to distinguish between recovered memories that result from leading suggestion versus those that are triggered by cues or events in the absence of suggestion. This dichotomy is relevant both for memories recovered in therapy and for those recovered outside of therapeutic settings.

IMPORTANT QUESTIONS: CONSENSUS ANSWERS AND DISAGREEMENTS

Do recovered memories exist? If so, how common are they? Consensus: The consensus appears to be that recovered memories do occur. That is, most scientists and practitioners believe that it is possible for memories of experiences to be encoded, then to undergo a period of inaccessibility, followed by a successful experience of remembering the original episode. Disagreements: The major points of contention are whether emotional (particularly traumatic) experiences are likely to be forgotten, whether the putative mechanism for such forgetting is conscious or unconscious, whether recovered memories constitute accurate representations of past experiences, and whether recovered memories are common or rare. Does repression exist? Do people have ways to avoid remembering traumatic experiences? Consensus: Most would agree that thoughts and memories can be suppressed, avoided, and forgotten, and that formerly forgotten experiences might later be remembered. Most also agree that traumatized individuals are motivated to banish painful thoughts and memories from consciousness. Disagreements: Whereas some believe that unconscious mechanisms of repression commonly keep out of mind unwanted thoughts, such as traumatic memories, others claim that there is no evidence for the existence of such mechanisms and that trauma is far more likely to be remembered than forgotten. Are recovered memories less (or more) accurate than continuous memories? Consensus: Research on this question has not shown any differences in the accuracy of recovered versus continuous memories. Neither forgetting nor false memories have been found to differ for these two types of memories. Disagreements: There is no consensus as to whether recovered memories are valid representations of events, primarily because not everyone agrees that any memories are valid representations of the past. Whereas some believe that memories are generally accurate, others see memory as largely reconstructive, and always subject to inferential errors. Are traumatic memories special? Consensus: Most, but not all, would agree that memories of traumatic experiences differ from memories of nontraumatic experiences. Disagreements: Whereas some maintain that natural coping mechanisms, such as repression and dissociation, routinely act to protect people from reexperiencing traumatic

314

SMITH AND OLEAVES

memories, others believe that emotional experiences, including traumatic ones, are unlikely to be forgotten. A third view has it that traumatic memories may consist of fragments of real experiences that are accurately or inaccurately reconstructed, depending on the appropriateness of the schema used to guide memory reconstruction. A fourth view is that even though traumatic memories may differ systematically from nontraumatic memories in that they may represent events that might be unlikely to come up in daily conversations, or that the events are associated with unusual emotional states that are rarely reexperienced, traumatic memories nonetheless obey normal laws of forgetting, such as inhibition, interference, and cue dependence.

DIRECTIONS FOR FUTURE RESEARCH

Future research should address the question of recovery of gist versus detail. A good deal of memory research (e.g., Brainerd & Reyna, 1993, 1998; Brainerd, Wright, Reyna, & Mojardin, 2003) indicates that gist and detail are encoded as different types of memories. This important distinction must be addressed in research on recovered memories. Continuous memory of gist, for example, does not ensure that memories of details are also continuously accessible, or vice versa. A soldier who can continuously recall only the vague gist of a battle, for example, without being able to recall details, might be shocked to recover or discover memories of those details. Likewise, an individual who can continuously recall a detail of an episode might be surprised to discover the meaning or gist encompassing that detail. One of the stickiest issues in the recovered memory debate is the definition of inaccessibility. To be a recovered memory, the individual in question must have undergone a period of time during which the memory was inaccessible. The history of memory research is replete, however, with studies that show that what is remembered with one type of test is different from a different test, what is recalled with one cue is different from another cue's influence, and what is retrieved on one occasion is going to be different from retrieval on other occasions (e.g., Shaughnessy, 2002). It is therefore impossible to prove that a memory is 100% inaccessible without the use of all types of tests, cues, and occasions to verify the inaccessibility. A consensual operational definition of inaccessibility would help enlighten those engaged in the recovered memory debate. Are there unconscious mechanisms involved in recovered memories? Although much has been learned about unconscious mechanisms in memory and perception, little is known about the role of unconscious mechanisms with respect to banishing memories from consciousness. The role of unconscious mechanisms is also unclear in terms of implicit memory of experiences prior to their conscious recovery, and in terms of how unconscious mechanisms might trigger memory recovery. Although the role of affect in memory is beginning to be understood, its role in the occurrence of recovered memories is not. In particular, it is not clear whether highly emotional memories are likely to be truly forgotten. Whereas one body of research indicates that emotional experiences are more likely than nonemotional ones to be remembered, other work indicates that one's personal belief in the superiority of his or her emotional memories is not supported by

10.

RECOVERED MEMORIES

315

objective evidence. There is also a dearth of evidence on the ways in which traumatic memories can be forgotten and recovered. Whether there are individual differences in the susceptibility to recovered memories is not known. Are some individuals more predisposed than others to forget traumatic experiences, and to later recover those lost memories? It is also not clear how individ~ uals who experience recovered memories are subsequently affected in their thinking and behavior. That is, are their memories accurate? Is the recovery of those memories thera~ peutic for those individuals?

SUMMARY, CONCLUSIONS, AND POLICY RECOMMENDATIONS

Recovered memories are not unitary phenomena, but rather are caused by a number of mechanisms, many of which are normal in the sense that all people experience them. Although the existence of repression, particularly a form that involves unconscious mechanisms, is highly controversial, the existence of recovered memories is not in doubt, nor does the existence of recovered memories rely upon the reality of repression. Some have expressed doubt, however, that emotional experiences, particularly traumatic ones that have been repeated over long periods of time, are likely to be forgotten. Recovered memories have not been found to differ from continuous memories in terms of the amount recalled or the level of false memories. Although recovered memories are defined in part by a period of inaccessibility, even in cases in which partial memory is continuous, it is possible for additional material to be recovered at a later time, and it is possible for surprising new interpretations of memories to be discovered. The issue of inaccessibility can have important legal implications if a recovered memory of an event appears as new evidence that was not previously available for a case. As we previously noted, it is difficult to know whether a particular memory is inaccessible, because research so often shows that an event that is not remembered on one test might be accessed on a different test. Clearly, it is impossible to conduct an infinite set of tests to determine that a memory is truly inaccessible, nor is there any goldstandard test that can be used to determine true inaccessibility. Rather than questioning the previous inaccessibility of a recovered memory, one should question whether the eyewitness withheld from evidence the memory of an event of which he or she was aware. Potential memories of experiences cannot be held in evidence; only conscious awareness of those memories can allow them to be used as evidence. Thus, memories should be considered as new evidence if they were not known and withheld. Such cases include memories that had never been previously recalled; those that had been recalled prior to the evidence-gathering, but that could not be recalled when requested; as well as memory discoveries that involve new realizations or understandings of old memories. Having just defended the validity of once inaccessible memories, we must be quick to point out that the fact that memories are recovered or discovered confers no special validity to those new memories. We have every reason to believe that, like memories of any other experiences, recovered memories are likely to be just as susceptible as

316

SMITH AND GLEAYES

continuous memories to the effects of memory decay over time, distortion, suggestion, false inference, and interference. Recovered memories of events that were originally experienced as traumatic would be susceptible to the distortion and/or fragmentation that can occur for memories formed under high levels of stress. Given that all long-term memory of events is susceptible to forgetting and false memories, it is clear that recovered memories, just like continuous memories, are fallible entities whose accuracy can be bolstered by corroborating evidence.

REFERENCES American Medical Association Council on Scientific Affairs. (1985). The status of refreshing recollection by the use of hypnosis. Journal of the American Medical Association, 253, 1918-1923. American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders, (4th ed., text revision). Washington, DC: Author. American Psychiatric Association Board of Trustees. (1994). Statement of memories of sexual abuse. International]ournal of Clinical and Experimental Hypnosis, 42, 261-264. Anderson, M. C., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Journal of Experimental Psychology: Learning, Memory & Cognition, 20, 1063-1087. Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by executive control. Nature, 410, 366-369. Anderson, M. C., & Neely, J. (1996). Interference and inhibition in memory retrieval. In E. L., Bjork & R. A. Bjork (Eds.), Memory handbook of perception and cognition (pp. 23 7-313). San Diego: Academic Press. Anderson, M. C., & Spellman, B. A. (1995). On the status of inhibitory mechanisms in cognition: Memory retrieval as a model case. Psychological Review, 102(1), 68-100. Ballard, P. B. (1913). Oblivescence and reminiscence. British Journal of Psychology Monograph Supplements, I, 1-82. Bower, G. H. (1981). Mood and memory. American Psychologist, 36(2), 129-148. Bower, G. H. (1990). Awareness, the unconscious, and repression: An experimental psychologist's perspective. In J. L. Singer (Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health. Chicago, IL: University of Chicago Press. Brainerd, C. J., & Reyna, V. F. (1993). Memory independence and memory interference in cognitive development. Psychological Review, 100, 42-67. Brainerd, C. J., & Reyna, V. F. (1998). When things that were never experienced are easier to "remember" than things that were. Psychological Science, 9(6), 484-489. Brainerd, C. J., Reyna, V. F., Wright, R., & Mojardin, A. H. (2003). Recollection rejection: False-memory editing in children and adults. Psychological Review, I10(4), 762-784. Braun, B. G. (1988). The BASK model of dissociation. Dissociation, I (1), 4-23. Brewen, C. R., & Andrews, B. (1988). Recovered memories of trauma: Phenomenology and cognitive mechanisms. Clinical Psychology Review, IB, 949-970. Brown, D., Scheflin, A. W., & Hammond, D. C. (1998). Memory, trauma, treatment, and the law: An essential reference on memory for clinicians, researchers, attorneys, and judges. New York: W.W. Norton & Company. Brown, D., Scheflin, A., & Whitfield, C. (1999). Recovered memories: The current weight of the evidence in science and in the courts. The Journal of Psychiatry & Law, 27(1), 5-156. Burke, D. M., MacKay, D. G., Worthley, J. S., & Wade, E. (1991). On the tip of the tongue: What causes word finding failures in young and older adults? Journal of Memory & Language, 30(5), 542-579.

10.

RECOVERED MEMORIES

317

Cardefia, E., & Gleaves, D. H. (2003). Dissociative disorders: Phantoms of the self. In S. M. Turner & M. Hersen (Eds.), Adult psychopathology and diagnosis (4th ed). New York: Wiley. Choi, H., & Smith, S. M. (2005). Incubation and the resolution of tip-of-the-tongue states. Journal of General Psychology: Experimental, Physiological, and Comparative Psychology, 132(4), 365-376. Chu, ]. A. (1998). Rebuilding shattered lives: The responsible treatment of complex post traumatic and dissociative disorders (pp. 179-203). New York: Wiley & Sons. Chu, ]. A., & Dill, D. I. (1990). Dissociative symptoms in relation to childhood physical and sexual abuse. American Journal of Psychiatry, 149,887-893. Chu, ]. A., Frey, L. M., Ganzel, B. L., & Matthews,]. A. (1999). Memories of childhood abuse: Dissociation, amnesia, and corroboration. American Journal of Psychiatry, 156(5), 749-755. Dalenberg, C.]. (1996). Accuracy, timing and circumstances of disclosure in therapy of recovered and continuous memories of abuse. Journal of Psychiatry and Law, 24, 229-276. Edelson, M. (1990). Defense in psychoanalytic theory: Computation or fantasy? In J. L. Singer (Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health. (pp. 33-60). Chicago: University of Chicago Press. Eich, E., & Birnbaum, I. M. (1982). Repetition, cuing, and state-dependent memory. Memory & Cognition, 10(2), 103-114. Eich, E., & Metcalfe, J. A. (1989). Mood dependent memory for internal versus external events. Journal of Experimental Psychology: Learning, Memory, & Cognition, 15(3), 443-455. Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99, 20-35. Foa, E., Steketee, G., & Rothbaum, B. (1989). Behavioral/cognitive conceptualizations of posttraumatic stress disorder. Behavior Therapy, 20, 155-176. Freud, S. ( 1962). The aetiology of hysteria. In J. Strachey (Ed., T ransl.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 3, pp. 191-221). London: Hogarth Press (original work published 1896). Freyd, J. ]. (1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge, MA: Harvard University Press. Freyd, J. ]. (2003). Memory for abuse: What can we learn from a prosecution sample? Journal of Child Sexual Abuse, 12(2), 97-103. Freyd, J. ]., & Goldberg, L. R. (2004). Gender difference in exposure to betrayal trauma. Paper presented at the 20th Annual Meeting of the International Society for Traumatic Stress Studies, New Orleans, LA, November 14-18, 2004. Gerkens, D. R. (2004). Are recovered memories accurate? Unpublished doctoral dissertation, Texas A&M University, College Station, TX. Gleaves, D. H. (1996). The evidence for "repression": An examination of Holmes (1990) and the implications for the recovered memory controversy. Journal of Child Sexual Abuse, 5, 1-19. Gleaves, D. H., & Hernandez, E. (1999). Recent reformulations of Freud's alleged discoveries of childhood sexual abuse: Historical/scientific clarification or a continued assault on truth? History of Psychology, 2, 1-31. Gleaves, D. H., Smith, S.M., Butler, L. D., & Spiegel, D. (2004). False and recovered memories in the laboratory and clinic: A review of experimental and clinical evidence. Clinical Psychology: Science and Practice, 11, 3-28. Gold, S. N. (2000). Not trauma alone: Therapy for child abuse survivors in family and social context. Philadelphia: Brunner/Routledge. Grinker, R. R., & Spiegel, J. P (1945). War neuroses. Philadelphia: Blakiston. Harvey, M. R. (1999). Memory, research, and clinical practice: A critique of three paradigms and a framework for psychotherapy with trauma survivors. In L. M. Williams & V. L. Banyard (Eds.), Trauma and memory (pp. 19-29). Thousand Oaks, CA: Sage. Holmes, D. S. (1990). The evidence for repression: An examination of sixty years of research. In J. L. Singer (Ed.), Repression and dissociation: Implications for personality theory, psychopathology, and health (pp. 85-102). Chicago: University of Chicago Press.

318

SMITH AND GLEAVES

Horn, M. (1993). Memories lost and found. U.S. News & World Report, November 29, pp. 52-63. Horowitz, M. J. (1986). Stress response syndromes. New York: Aronson. Jacobs, W. J., & Nadel, L. (1985). Stress induced recovery of fears and phobias. Psychological Review, 92,512-531. Jacobs, W. J., & Nadel, L. (1998). Neurobiology of reconstructed memory. Psychology: Public Policy and Law, 4, 1110-1134. Janet, P.M. F. (1889). L'automatisme psychologique. Thesis for doctor of philosophy. Paris: Felix Alcan. Jones, G. V. (1989). Back to Woodworth: The role of interlopers in the tip of the tongue phenomenon. Memory & Cognition, 17,69-76. Jones, G. V., & Langford, S. (1987). Phonological blocking in the tip of the tongue state. Cognition, 26, 115-122. Kardiner, A., & Spiegel, H. (1947). War stress and neurotic illness. Oxford, England: Hoeber. Keane, T. M. (1998). Psychological and behavioral treatment of post-traumatic stress disorder. In P. E. Nathan &J. M. Gorman (Eds.), A guide to treatments that work (pp. 398-407). London: Oxford University Press. Kihlstrom, John F. (2002). No need for repression. Trends in Cognitive Sciences, 6(12), 502. Kihlstrom, J. F., & Barnhardt, T. M. (1993). The self-regulation of memory, for better and for worse, with and without hypnosis. In D. B. Wegner & J. J. Pennebaker (Eds.), Handbook of mental control (pp. 88-125). Englewood Cliffs, NJ: Prentice-Hall. Kluft, R. P. (1995). The confirmation and disconfirmation of memories of abuse in dissociative identity disorder patients: a naturalistic study. Dissociation, 4, 253-258. Kluft, R. P. (1996). Dissociative identity disorder. In L. K. Michelson & W. J. Ray (Eds.), Handbook of dissociation: Theoretical, empirical, and clinical perspectives (pp. 336-337). New York: Plenum Press. Koriat, A. (1993). How do we know that we know? The accessibility model of the feeling of knowing. Psychological Review, 100(4), 609-639. Lang, P. J. (1985). The cognitive psychophysiology of emotion: Fear and anxiety. In A.H. Tuma & J. Maser (Eds.), Anxiety and the anxiety disorders (pp. 131-170). Hillsdale, NJ: Lawrence Erlbaum Associates. Leavitt, F. (2001). Iatrogenic recovered memories: Examining the empirical evidence. American Journal of Forensic Psychology, 19, 21-32. Leskin, G. A., Kaloupek, D. G., & Keane, T. M. (1998). Treatment for traumatic memories: Review and recommendations. Clinical Psychology Review, 18, 983-1002. Light, L. L., & Carter-Sobell, L. (1970). Effects of changed semantic context on recognition memory. Journal of Verbal Learning & Verbal Behavior, 9(1), 1-11. Loewenstein, R. J. (1991). An office mental status examination for complex chronic dissociative symptoms and multiple personality disorder. Psychiatric Clinics of North America, 14 (3), 567-604. Loewenstein, R. J. (1993). Dissociation, development, and the psychobiology of trauma. American Academy of Psychoanalysis, 21 (4), 581-603. Maldonado, J., Butler, L. D., & Spiegel, D. (2002). Treatments for dissociative disorders. In P. E. Nathan & J. M. Gorman (Eds.), A guide to treatments that work (2nd ed., pp. 463-496). London: Oxford University Press. McGeoch, John A., & Irion, Arthur L. (1952). The psychology of human learning (2nd ed.). Oxford, England: Longmans, Green & Co. McNally, R. J., Clancy, S. A., Barrett, H., M., & Parker, H. A. (2005). Reality monitoring in adults reporting repressed, recovered, or continuous memories of childhood sexual abuse. Journal of Abnormal Psychology, 114, 147-152. McNally, R. J., Clancy, S. A., Schacter, D. L., & Pitman, R. K. (2000). Personality profiles, dissociation, and absorption in women reporting repressed, recovered, or continuous memories of childhood sexual abuse. Journal of Consulting & Clinical Psychology, 68, 1033-103 7.

10.

RECOVERED MEMORIES

319

Metcalfe, J., Schwartz, B. L., &Joaquim, S. G. (1993). The cue-familiarity heuristic in metacognition. Journal of Experimental Psychology: Learning, Memory, & Cognition, 19 (4), 851-864. Mower, 0. H. (1960). Learning theory and behavior. New York: Wiley. Nadel, L., &Jacobs, W. J. (1998). Traumatic memory is special. Current Directions in Psychological Science, 7, 154-157. Payne, D. G. (1987). Hypermnesia and reminiscence in recall: A historical and empirical review. Psychological Bulletin, 101 (1), 5-27. Pennebaker, J. W. (2003). Writing about emotional experiences as a therapeutic process. In P. Salovey & A. J. Rothman (Eds.), Social psychology of health. Key readings in social psychology (pp. 362-368). New York: Psychology Press. Pennebaker, J. W. (2004). Writing to heal: A guided journal for recovering from trauma and emotional upheaval. Oakland, CA: New Harbinger Press. Postman, L., & Underwood, B. J. (1973). Critical issues in interference theory. Memory & Cognition, 1, 19-40. Reiss, S. (1980). Pavlovian conditioning and human fear: An expectancy model. Behavior Therapy, 11' 380-396. Resick, P. A., & Calhoun, K. S. (2001). Posttraumatic stress disorder. In D. H. Barlow (Ed.), Clinical handbook of psychological disorders: A step-by-step treatment manual (3rd ed., pp. 60113). New York: Guilford Press. Rodd, Z. A., Bell, R. L., Sable, H. J., Murphy, J. M., & McBride, W. J. (2004). Recent advances in animal models of alcohol craving and relapse. Pharmacology, Biochemistry & Behavior, 79(3), 439-450. Ross, C. A., Duffy, C. M. M., & Eliason, J. W. (2002). Prevalence, reliability, and validity of dissociative disorders in an inpatient setting. Journal of Trauma and Dissociation, 3(1), 7-17. Ross, C. A., Heber, S., Norton, G. R., Anderson, G., & Barchet, P. (1989). The Dissociative Disorders Interview Schedule: A structured interview. Dissociation, 2, 169-189. Sands, S. H. (1994). What is dissociated? Dissociation, 7(3), 145-152. Sargant, W., & Slater, E. (1941). Amnestic syndromes of war. Proceeding of the Royal Society of Medicine, 34, 757-764. Saxe, G. N., van der Kolk, B. A., Berkowitz, R., Chinman, G., Hall, K., Liegerg, G., et al., (1993). Dissociative disorders in psychiatric inpatients. American Journal of Psychiatry, 150, 103 7-104 2. Schooler, J. W. (1999). Seeking the core: The issues and evidence surrounding recovered accounts of sexual trauma. In L. M. Williams & V. L. Banyard, Trauma and memory (pp. 1929). Thousand Oaks, CA: Sage. Schooler, J. W. (2001). Discovering memories of abuse in the light of meta-awareness. Journal of Aggression, Maltreatment & Trauma, 4(2), 105-136. Schooler, J. W., Ambadar, Z., & Bendiksen, M. (1997). A cognitive corroborative case study approach for investigating discovered memories for sexual abusue. In J. D. Read & D. S. Lindsay (Eds.), Recollections of trauma-Scientific evidence and clinical practice (pp. 379-387). New York: Plenum Press. Schwartz, B. L., & Metcalfe, J. (1992). Cue familiarity but not target retrievability enhances feelingof-knowing judgments. Journal of Experimental Psychology: Learning, Memory, & Cognition, 18(5), 1074-1083. Schwartz, B. L., & Smith, S. M. (1997). The retrieval of related information influences tip-ofthe-tongue states. Journal of Memory & Language, 36, 68-86. Schwartz, B. L., Travis, D. M., Castro, A.M., & Smith, S.M. (2000). The phenomenology of real and illusory tip-of-the-tongue states. Memory & Cognition, 28, 18-27. Shaughnessy, M. F. (2002). An interview with Henry Roediger III. Educational Psychology Review, 14(4), 395-411. Sivers, H., Schooler, J., &Freyd, J. J. (2002). Recovered memories. In V. S. Ramachandran (Ed.), Encyclopedia of the human brain (Vol. 4, pp. 169-184). San Diego and London: Academic Press.

320

SMITH AND OLEAYES

Smith, S.M. (1995). APA Division 3 Invited Address, Recovery from memory blocks: Experimental evidence. Presented at the meeting of the American Psychological Association, New York. Smith, S.M., & Moynan, S.C. (November, 2004). Forgetting lists of*$%#! words. Presented at the annual meeting of the Psychonomic Society, Minneapolis, MN. Smith, S. M., Glenberg, A.M., & Bjork, R. A. (1978). Environmental context and human memory. Memory & Cognition, 6, 342-353. Smith, S. M., Gleaves, D. H., Pierce, B. H., Williams, T. L., Gilliland, T., & Gerkens, D. R. (2003). Eliciting and comparing false and recovered memories: An experimental approach. Applied Cognitive Psychology, 17, 251-279. Smith, S.M., Gerkens, D. R., Choi, H., & Hull, R. G. (November, 2003). Forgetting and recovery without inhibition. Poster presented at the annual meeting of the Psychonomic Society, Vancouver, BC. Smith, S. M., & Vela, E. (2001). Environmental context-dependent memory: A review and meta-analysis. Psychonomic Bulletin and Review, 8, 203-220. Spiegel, D. (1994). Dissociative disorders. In R. E. Hates, S.C. Yodofsky, &J. A. Talbots (Eds.), The American Psychiatric Press textbook of psychiatry (2nd ed., pp. 633-652). Washington, DC: American Psychiatric Press. Steinberg, M. (1993). Interviewer's guide to the structured clinical interview for DSM-IV dissociative disorders. Washington, DC: American Psychiatric Press. Steinberg, M. ( 1994). Structured clinical interview for DSM-IV dissociative disorders-revised (SCID-D-R) (2nd ed.). Washington, DC: American Psychiatric Press. Terr, L. C. (1995). In G. S. Everly &J. M. Lating (Eds.), Psychotraumatology: Key papers and core concepts in post-traumatic stress (pp. 301-320). New York: Plenum Press. Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 359-380. van der Kolk, B. A., & Fisler, R. (1995). Dissociation and the fragmentary nature of traumatic memories: Overview and exploratory study. Journal of Traumatic Stress, 8, 50 5-5 25. van der Kolk, B. A., & Vander Hart, 0. (1989). Pierre Janet and the breakdown of adaptation in psychological trauma. American Journal of Psychiatry, 146, 1530-1540. Waller, N., Putnam, F. W., & Carlson, E. B. (1996). Types of dissociation and dissociative types: A taxometric analysis of dissociative experiences. Psychological Methods, 1, 300-3 21. Wegner, D. M. (1989). White bears and other unwanted thoughts. New York: Viking. Wenzlaff, R. M., & Wegner, D. M. (2000). Thought suppression. Annual Review of Psychology, 51,59-91. Widner, R. L., Jr., Smith, S.M., & Graziano, W. (1996). Effects of demand characteristics on feeling-of-knowing and tip-of-the-tongue reports. American Journal of Psychology, 525-538. Williams, L. M. (1994). Recall of childhood trauma: A prospective study of women's memories of child sexual abuse. Journal of Consulting and Clinical Psychology, 62, 1167-1176. Williams, L. M. (1995). Recovered memories of abuse in women with documented child sexual victimization histories. Journal of Traumatic Stress, 8, 649-674. Wilsnack, S. C., et al. (2002). Self-reports of forgetting and remembering childhood sexual abuse in a nationally representative sample of US women. Child Abuse & Neglect, 26, 139-147. Zerbe Enns, C., Campbell, J., Courtois, C. A., Gottlieb, M. C., Lese, K. P., Gilbert, M. S., et al. (1998). Working with adult clients who may have experienced childhood abuse: Recommendations for assessment and practice. Professional Psychology: Research and Practice, 29, 245-256.

11 Using Hypnosis in Eyewitness Memory: Past and Current Issues Giuliana Mazzoni University of Plymouth

Steven Jay Lynn Binghamton University

Stemming from the clinical work of Janet (1889/1973), hypnosis has received considerable attention over the years as a method of enhancing memory reports. The underlying assumption, common to both clinical and forensic applications, was that hypnosis can help people recover memories that otherwise would remain inaccessible. Janet was one of the first therapists to use hypnosis to help patients recover memories for traumatic events that he assumed caused their pathological conditions. In a well-known case, Janet's patient "Maria" was age regressed during hypnosis to her childhood. Janet reported that in one instance she was able to remember a traumatic episode of her life (i.e., seeing a child with a facial deformity), and that by consciously reliving the memory of the trauma she was liberated from the symptoms it had caused. The belief that hypnosis can help patients recover buried memories of traumatic events also provided the rationale for hypnoanalysis, which was used after World War I to help soldiers and veterans recover from neurotic symptoms, and what is now called post-traumatic stress disorder. The logic underlying the use of hypnoanalysis was based on the alleged healing power of abreaction, which occurred when patients were hypnotized and instructed to remember the events that purportedly led to their psychological disorder. Patients were also given posthypnotic suggestions (i.e., suggestions given during hypnosis for behavior to be performed after hypnosis is terminated) to remember the events during the waking state and were helped to work through what was remembered in subsequent hypnotic sessions. One of the most prominent features of hypnosis is the substantial variability in responsiveness to suggestion. Most people are moderately hypnotizable (i.e., they respond 321

322

MAZZONI AND LYNN

to some suggestions but not others), whereas other people are either very slightly (1520%) or very highly (15-20%) hypnotizable. Apparent memory recovery is most likely in subjects who fall in the moderate to high range of hypnotic suggestibility, although some low suggestible individuals occasionally report false memories (see Lynn, Meyers, & Malinoski, 1997), as a more thorough review of the literature below will indicate. The difficulty in recovering memories encountered with relatively low suggestible patients spurred the development of a technique in which patients were administered relatively high doses of barbiturates in order to facilitate memory recovery (Horsley, 1936). Narcoanalysis, as the new technique was called, was based on the same fundamental belief that guided the use of hypnosis. That is, much, if not all, information that is apparently forgotten is not truly irretrievable, but is instead buried in the unconscious, where it can be excavated when the right methods are applied. According to Horsley, the founder of narcoanalysis, these techniques could be used to extract confessions from people when they were guilty but not when they were innocent (for a more extensive review of these techniques, see Kihlstrom, 1998). The belief that events can reside in the unconscious and be accessed only by way of special memory recovery techniques is still prevalent among certain therapists. These clinicians claim that many forms of psychological problems are manifestations of repressed or dissociated memories of traumatic abuse and neglect during childhood (see Lynn, Knox, Fassler, Lilienfeld, & Loftus, 2004). According to these therapists, hypnosis and age regression represent roads to the retrieval of buried memories, and retrieval helps patients heal by bringing the memories into conscious awareness (e.g., Bass & Davis, 1988; Blume, 1990; Fredrickson, 1992; Herman, 1992; Terr, 1994). For example, survey research (Poole, Lindsay, Memon, & Bull, 1995) reveals that approximately onethird (29% and 34%) of psychologists in the United States who were sampled reported that they used hypnosis to help patients recall memories of sexual abuse. It is notable that only 5% of British therapists sampled used hypnosis for this purpose. The differences reported across countries may well reflect widespread media portrayals in the United States of hypnosis as akin to truth serum and the more prevalent use of hypnosis for a variety of clinical and medical conditions in the United States (Lynn, Fassler, Knox, & Lilienfeld, in press). From the clinical context, the use of hypnosis for memory recovery and enhancement spread to the forensic context, in which witnesses have been hypnotized as a means of enhancing memory reports. Wagstaff (1999) has reported that forensic hypnosis was especially popular in the late 1970s and early 1980s, when many books and papers supporting its use in police investigations were published (Haward & Ashworth, 1980; Kleinhauz, Horowitz, & Tobin, 1977; Hibbard & Warring, 1981; Reiser, 1980). According to Wagstaff (1999), published accounts likely underestimated the use and extent of hypnosis in police interrogations during this time period. He reports that one British investigator (Haward, 1988), who used hypnosis in forensic contexts for 30 years, conducted 17 interrogations in one week at the request of police. The use of hypnosis to facilitate memory has become a topic of considerable controversy. Summarizing the arguments on one side of this issue, the American Society of Clinical Hypnosis (ASCH) sought to advance hypnosis in clinical and forensic contexts

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

323

by issuing clinical practice and forensic guidelines that legitimize the use of hypnosis for memory improvement or recovery (Hammond, Garver, et al., 1995). However, contrary to this optimistic assessment, many academic psychologists (e.g., Karlin, Kihlstrom, Laurence, Lynn, Nash, E. Orne, M. T. Orne, Perry) have argued against the use of hyp~ nosis for memory recovery, instead highlighting the potential of hypnosis to increase suggestibility and create false memories in psychotherapeutic and forensic contexts. The data concerning the negative impact of hypnosis on memory have had important legal ramifications, including, in many states, a per se exclusion of the testimony of wit~ nesses who have been hypnotized. We review the reasons for such skepticism in this chapter and consider the diverse ways in which courts have treated hypnotically elicited testimony.

HYPNOSIS AND MEMORY

From the very beginning, therapists have used hypnotic memory retrieval in two oppo~ site ways, without any apparent awareness of the implications that one use had for the other. On one hand, hypnosis has been used to enhance memory and helped recover autobiographical events that otherwise would remain inaccessible. When Marie (Janet's patient) remembered the deformed face of the boy, she and her therapist concluded that the hypnosis she underwent produced a true memory of a real childhood event. This in~ terpretation was consistent with the view that people remember real events, facts, and details with the help of an hypnotic induction that they would not be able to retrieve otherwise. According to this perspective, hypnosis is an excellent memory recovery tool because it selectively enhances accurate remembrances. On the other hand, Janet and other therapists at that time also used hypnosis to create healing pseudomemories. That is, they created false memories of positive events that never happened in order to replace the "true" traumatic memories that, in many cases, had previously been recovered through hypnosis. For example, Janet helped Marie "remember" that the boy she had recalled in a previous hypnotic session did not have any scary facial features. According to Janet, she accepted this pseudomemory as a true memory, and her symptoms disappeared. The intentional fabrication of memory for therapeutic benefit was not limited to Janet. For example, Bernheim (1888/1973) used hypnosis to create perceptions and memories that he called "retroactive hallucinations." Others wrote about illusions of memory (Burnham, 1889), illusory retroactive memories (Forel, 1906), and "construe~ tions" of the past that were said to have a positive therapeutic value for patients (Freud, 1909/1955). Thus, since the 1800s, clinicians have been aware of the ease with which false memories can be created in hypnosis. Clinicians have also been cognizant of the role of hypnotic suggestibility in the for~ mation of pseudomemories. Janet (1989!1973), for example, claimed that false memo~ ries can be created only in highly hypnotizable people. Despite the awareness that hyp~ nosis can produce believed~in false memories, clinicians expressed no doubt about the veracity of the memories that patients recovered during and after hypnosis.

324

MAZZONI AND LYNN

Deleterious Effects of Hypnosis in Memory: some Empirical Evidence Evidence for the influence of suggestibility and the unreliability of hypnosis as a memory enhancement tool is not limited to clinical case reports. Modern research has indicated that suggestibility is linked with pseudomemories in both hypnotic and nonhypnotic contexts. As hinted at earlier, studies have established that highly and moderately suggestible participants report more false memories in response to misleading information and are more prone to develop believed-in false memories during hypnosis than low-suggestibility individuals (see Lynn & Nash, 1994; Burgess & Kirsch, 1999). In some studies, highly suggestible individuals report more false memories than mediumsuggestible individuals, and in other studies highly suggestible and medium-suggestible individuals perform comparably (see Lynn et al., 1997). The fact that false memories are reported by highly and medium-suggestible persons, who make up the majority of the population, and that even low-suggestible persons occasionally report pseudomemories (E. C. Orne, Whitehouse, Dinges, & M. T. Orne, 1996; Scoboria et al., 2001), implies that many individuals are vulnerable to the influence of misleading and suggestive memory recovery techniques. Apart from suggestibility, many studies have shown that the induction of hypnosis itself can engender illusory memories. Indeed, in the review that follows, we document findings indicating that hypnosis can decrease memory accuracy substantially in certain instances and create new and false memories for events that did not occur. The findings reported here have been instrumental in providing a rationale for the formulation of perse inadmissibility rules in American courtrooms, a topic that is extensively covered in the second section of this chapter. Laurence and Perry (1983) provided one of the earliest and most dramatic empirical demonstrations of the production of a false memory in hypnosis. In their study, during hypnosis, highly suggestible participants were regressed to a night of the previous week, and a suggestion was given that they heard a loud noise that awakened them. If they reported having heard the suggested noise, they were asked to describe it in detail. Sixtythree percent of the participants accepted the suggestion and reported hearing the noise. Of these, 76% subsequently developed false memories of the noise. Half of these remembered the noise clearly and were convinced that it had actually happened, whereas the others reported being somewhat confused about the source of the noise. A number of studies (see Brown, Scheflin, & Hammond, 1998; Hammond, Garver, et al., 1995) have shown that hypnosis increases the number of items that are recalled. Some studies have indicated that hypnosis might even improve recall accuracy to a small extent. These findings have been used to (a) support the view that hypnosis facilitates memory retrieval and (b) to bolster the claims of proponents of the use of hypnosis for memory recovery. However, comprehensive literature reviews reveal that the effect is unreliable, and that it is probably linked to factors that are not specific to hypnosis (Erdelyi, 1994; Lynn, Neuschatz, Fite, & Rhue, 2001; McConkey & Sheehan, 1995; Steblay & Bothwell, 1994).

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

325

For example, Erdelyi (1994) observed that the positive effects of hypnosis on mem, ory are confined to free recall for high,sense material (e.g., recall for poems) and dis, appear when recall concerns lists of words, or when other types of memory tests are used. The small positive effect in recall that is sometimes observed is probably due to the lowering of participants' report criterion (Scoboria et al., 2001), such that more items are accepted and reported as memories. Accordingly, the apparent increase in recall does not represent a genuine enhancement of the accessibility of memory contents. Indeed, enhanced recall is invariably accompanied by an increase in false,positive responses, as would be predicted by a lowering of the report criterion. In interpreting the experimental data, it is important to consider the difference be, tween the amount of material that is recalled and the accuracy of what is recalled. Many studies have shown that hypnosis increases the sheer volume of recall, but this apparent enhancement is the by,product of increases in both correct and incorrect information. Therefore, the key to the effect of hypnosis on accuracy depends on the proportion of correct to incorrect information produced in recall without hypnosis and the propor, tion of correct to incorrect information added by the use of hypnosis. When response productivity is statistically controlled, hypnotic recall is no more accurate than non, hypnotic recall (Erdelyi, 1994; Steblay & Bothwell, 1994). In fact, as implied above, hyp, nosis typically produces a tradeoff of inaccurate information for any advantage in accu, rate information recalled. In general, hypnosis produces more recall errors, more intrusions of uncued errors, and higher levels of memories for false information than nonhypnotic conditions, including context reinstatement (Nogrady, McConkey, & Perry, 1985; Kebbel & Wagstaff, 1997; Sheehan & Tilden, 1983; Steblay & Bothwell, 1994). Accordingly, recall after hypnosis is often less accurate than nonhypnotic recall. Relatedly, the claim that hypnosis accurately reinstates experiences from the distant past (e.g., early childhood) has not been substantiated (Nash, 1987). Early memories are highly malleable, prone to distortion, and vulnerable to the influence of leading ques, tions and suggestive procedures (Loftus, 2003). For example, individuals are willing to accept suggestions about events that did not occur early in their lives. For example, they accept that (a) they got lost for an extended period of time (Mazzoni & Loftus, 1998), (b) they were bullied (Mazzoni et al., 1999), (c) they spilled punch at a wedding (Hyman & Pentland, 1996), and (d) they had a nurse take out a skin sample from their little fin, ger (Mazzoni & Memon, 2003).

Adult Memories of Childhood Events By examining early memory reports, it is possible to study the influence of memory re, covery techniques on the recall of events that cross the threshold of infantile amnesia of about 2 years of age. Most adults' earliest reported memories date back to between 36 and 60 months of age. Indeed, virtually all contemporary memory researchers agree that accurate memory reports of events that occur before 24 months of age are ex, tremely rare. Adult memory reports of events occurring at or prior to this age are likely to represent confabulations, condensations, and constructions of early events, as well as

326

MAZZONI AND LYNN

current concerns and stories heard about early events (Bruhn, 1984; Loftus, 1993; Spanos, 1996). Hence, reports of events that are purported to have occurred before the cutoff of infantile amnesia are unlikely to be real memories. Very early memory reports are vulnerable to suggestive influences via hypnosis and a variety of other suggestive procedures (e.g., dream interpretation, Mazzoni, Loftus, Seitz, & Lynn, 1999; imagery, Mazzoni &Memon, 2003). For example, Sivec, Lynn, and Malinoski (1997a) interviewed 40 hypnotized and 40 nonhypnotized participants about their earliest memories. The first time that participants were asked to report their earliest memory, only 3% of the nonhypnotized participants recalled a memory earlier than 2 years. However, 23% of the hypnotized participants reported a memory earlier than age 2, 20% reported a memory earlier than 18 months, 18% reported a memory earlier than a year, and 8% reported a memory of earlier than 6 months. The second time that nonhypnotized participants were asked for an early memory, only 8% of the nonhypnotized participants reported a memory earlier than 2 years, and only 3% reported memories of 6 months or earlier. In contrast, 35% of the hypnotized participants reported memories earlier than 18 months, 30% reported memories earlier than a year, and 13% reported a memory before 6 months. In another study of hypnosis and early memories (Marmelstein & Lynn, 1999), participants reported earlier memories during hypnosis than they did both prior to hypnosis and prior to (nonhypnotic) instructions to recover memories over a 2-week period. One-third of the participants reported memories prior to the cutoff of infantile amnesia (i.e., age 2) after they received nonhypnotic suggestions that they could recall earlier memories. However, during hypnosis, two-thirds of participants reported such memories. Even after being debriefed and contacted by telephone outside the experimental context, more than a third (37%) of the participants continued to report memories prior to age 2. Hypnotic Age Regression. Hypnosis is often used to facilitate the experience of age regression. Age regression involves regressing a person back through time to an earlier life period. Participants are typically given relaxation instructions and then asked to mentally re-create events that occurred at successively earlier periods in their life, or to focus on a particular event at a specific age, with suggestions that they are to fully relive the target event. If hypnosis is not a viable means of recovering memories of the recent past, then there is no reason to believe that it would be any more effective in recovering memories of childhood events. Nash's comprehensive review of the literature (1987) revealed that there is no special correspondence between the behavior and experience of hypnotized adults and that of children. In fact, as the following two studies indicate, hypnosis can contribute to memory distortions of early life events. In one study, third-party verification was obtained for early memory reports of 14 subjects who had been hypnotized and age regressed and 10 control subjects who pretended to go through hypnotic age regression (Nash, Drake, Wiley, Khalsa, & Lynn, 1986). The recollections of hypnotized subjects matched their parents' reports only 21% of the time, whereas simulators' reports after hypnosis were corroborated by their parents 70% of the time. In another study, Sivec and Lynn (1997b) age-regressed partie-

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

327

ipants to the age of 5 and suggested that they played with a Cabbage Patch doll if they were a girl or a He-Man toy if they were a boy. An important aspect of this study was that these toys were not released until 2 or 3 years after the target time of the age regression suggestion. Half of the subjects received hypnotic age regression instructions and half of the subjects received suggestions to age regress that were not administered in a hypnotic context. None of the nonhypnotized persons were influenced by the suggestion. In contrast, 20% of the hypnotized subjects rated the memory of the experience as real and were confident that the event had occurred.

Hypnosis and Confidence in Memory Reports Bernheim (1888/1973) reported that he suggested to a patient that she went to the bathroom four times, and on her last visit she fell and hit her nose. The patient reported remembering having gone to the bathroom four times because of diarrhea, and having fallen on the fourth visit, hitting her nose. The striking aspect of this memory is that it was impossible for Bernheim to convince her that this memory was false. She was adamantly convinced that it was a veridical memory for an event that had in fact occurred. As this anecdote implies, a potentially serious problem with the use of hypnosis for memory enhancement is that hypnotized individuals are more confident about the content of their recall than nonhypnotized individuals (Steblay & Bothwell, 1994). In one study, for example, hypnosis produced an increase in confidence for responses that had been designated as "guesses" during a prior waking test (Witehouse, Dinges, E. C. Orne, & M. T. Orne, 1988). To date, there are at least 18 studies (see Brown et al., 1998) that have demonstrated that hypnotized individuals are more confident in their recollections than nonhypnotized individuls. In contrast, Brown et al. (1999) cited 10 articles that reported no evidence for inflated confidence following an hypnotic induction. However, the studies cited did not support the contention that hypnosis improves recall accuracy. In fact, three of the studies cited found evidence for less accurate recall in hypnotic versus nonhypnotic conditions (Putnam, 1979; Ready, Bothwell, Brigham, Sanders, & Simmons, 1983), although one of the studies found that hypnosis improved recall in one test condition (Ready, Bothwell, & Brigham, 1997). In summary the available evidence indicates that hypnosis is more likely to inflate confidence in memories than nonhypnotic procedures. However, in cases where hypnosis does not produce unwarranted recall confidence, it is also the case that hypnosis generally does not improve recall and in some cases produces more recall errors. The increase in hypnosis-related confidence has important implications for eyewitness testimony. If participants are asked to recall material in hypnosis, they are more confident that their memories are correct, even when the memories are erroneous and false. Moreover, the effect of inflated confidence is particularly strong when highly hypnotizable individuals are hypnotized (Steblay & Bothwell, 1994). In a forensic context, unwarranted confidence may affect not only the behavior of the witnesses who strongly believe that the memories they report are true (even when they are false), but the perception of those who must decide whether the testimony is accurate or not as well.

328

MAZZONI AND LYNN

Confidence can determine how a jury perceives eyewitness testimony. A number of studies have shown that the confidence expressed by a witness is the single most important factor in persuading mock jurors that a witness correctly identified the culprit (e.g., Wells & Brandfield, 1998). Although the effects of confidence are not always present and are not universally large in magnitude, hypnosis by no means selectively increases confidence in accurate memories.

HYPNOSIS IN THE COURTROOM

In the courtroom, the pendulum of judicial opinion has swung between the extremes of denying the admission of hypnotically elicited testimony to treating hypnotic testimony much like any other type of testimony. Over the years, judicial opinion has shifted on the basis of how hypnosis has been perceived by the academic and broader professional community. The landmark case of People v. Ebanks (1897), in which it was opined that "The law of the United States does not recognize hypnotism" (p. 1053), established a precedent of per se inadmissibility, which rendered any testimony elicited during or after hypnosis inadmissible in the courtroom. This decision was defensible in light of the fact that, at the time of the ruling, hypnosis was not recognized by any professional organization as a legitimate therapeutic or scientific tool. Because of this situation, much expert testimony about hypnotically elicited memories was excluded by the Frye v. U.S. (1923) decision, which held that scientific information must be "sufficiently established to have gained general acceptance in the particular field to which it belongs" (p. 1014). As a consequence, hypnotically elicited testimony was virtually eliminated in the courts for many years. It was not until Harding v. State (1968) that certain courts began to soften their position regarding hypnosis, and a variety of approaches began to be adopted by courts across the country. In the Harding case, the court ruled that hypnotically elicited testimony was acceptable because the hypnotist was a well-trained professional and the procedures were neither leading nor suggestive. Moreover, it was adjudged that because hypnosis affected the credibility of the testimony, it was the jury's role to determine the weight that the testimony should be accorded. The open or per se admissibility standard became popular in the period from 1968 to 1978 (Brown et al., 1998) as hypnosis achieved greater respectability. This respectability was abetted by the recognition by major professional societies (e.g., American Psychological Association, American Medical Association) of hypnosis as a useful adjunct to psychotherapy, and by police insisting that hypnosis was a valuable investigative tool. Indeed, Perry, Orne, London, and Orne (1996) have observed that in certain cases, hypnotically elicited recall was seen as more reliable than actual physical evidence. Open admissibility began to spark controversy and was increasingly challenged in the courtroom as hypnotic testimony was adjudged admissible even when it was elicited by untrained laypersons, and as the pitfalls and perils of hypnotically elicited testimony received due recognition. For instance, Orne, Whitehouse, Dinges, and Orne (1988) wrote: "In State v. Mack (1980), a hypnotized person remembered eating a pizza in a restaurant

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

329

that did not serve pizza, seeing tattoos on someone who had none, and having been stabbed with scissors or a knife where there was no evidence that a weapon was involved" (p. 41). Today, only four states in the United States hold to the view that hypnotically elicited testimony is generally admissible. 1 The current state of hypnosis in the courtroom in many states of the United States was decided by People v. Shirley (1982), in which the California Supreme Court adopted a sweeping "per se inadmissible" rule. In this case, the testimony of the victim changed substantially after she was hypnotized the night before she testified. Furthermore, the hypnosis was performed by the deputy district attorney, who obviously had a vested interest in the outcome of the case and the woman's testimony. The California Supreme Court was compelled by a review of the literature on hypnosis and memory and ruled that the testimony was inadmissible and that any case-by-case analysis of hypnotic procedures would be too costly, thereby justifying a per se rule of inadmissibility. Following Shirley, experts increasingly testified that hypnosis was an unreliable recall procedure, that it increased suggestibility and pressures to comply with the hypnotist, that hypnotized persons could not discriminate accurate and false memories, that hypnosis "hardened" or "crystallized" testimony to the point that it rendered the witness immune to cross-examination, and that hypnosis conducted by police and laypersons carried great risks of pseudomemory formation. Today, 27 states 2 in the United States have ruled that individuals who have been hypnotized are not admissible eyewitnesses. In some states this holds even when hypnosis was performed for issues that are not directly connected to the topic of the testimony. However, there might be circumstances in which it is important to admit an individual to testify even after hypnosis. Indeed, there are some exceptions to the total inadmissibility rule (e.g., Rock v. Arkansas, 1987) that seem to pertain to situations in which the hypnotized person is the victim or the defendant, and inadmissibility seems to violate his/her constitutional rights based on the Sixth Amendment to the U. S. Constitution, which guarantees various rights to an accused person in a criminal trial. In such instances, defendants who have been hypnotized for memory enhancement are permitted to testify (Scheflin, Spiegel, & Spiegel, 1999). There are also specific procedural guidelines and alleged safeguards that, if implemented, make the testimony of a hypnotized individual admissible. The guidelines were first elaborated in State v. Hurd (State v. Hurd, 86 N.J. 525, 432 A.2d 86 (1981) (also called "Hurd" or "Orne" rules) and include the following six procedural rules or safeguards: (1) A psychiatrist or psychologist experienced in the use of hypnosis, who should be able to qualify as an expert, must conduct the session. (2) The professional conducting the session should be independent of and not regularly employed by the prosecutor, investigator, or defense. (3) Any information given to the hypnotist by law enforcement personnel or the defense prior to the hypnotic session must be recorded, either in writing North Dakota, Oregon, Tennessee, and Wyoming. Alaska, Arizona, Arkansas, California, Connecticut, Deleware, Florida, Georgia, Hawaii, Illinois, Indiana, Iowa, Kansas, Maryland, Massachussetts, Michigan, Minnesota, Missouri, Nebraska, New York, North Carolina, Oklahoma, Pennsylvania, Utah, Virginia, Washington, and West Virginia. 1

2

330

MAZZONI AND LYNN

or another suitable form. (4) Before inducing hypnosis the hypnotist should obtain from the subject a detailed description of the facts as the subject remembers them. The hypnotist should carefully avoid influencing the description by asking structured questions or adding new details. (5) All contacts between the hypnotist and the subject must be recorded. (6) Only the hypnotist and the subject should be present during any phase of the hypnotic session, including the pre-hypnotic testing and the post-hypnotic interview. Notably, because of the perceived potential for hypnosis to impair recall, Hurd maintains that the burden of establishing admissibility by clear and convincing evidence rests on the party who seeks to introduce hypnotically elicited testimony. Twelve states currently adopt the Hurd approach. 3 A number of state and federal courts that allow exceptions to the "admissibility with safeguards" rule apply a "totality of circumstances" test. In such instances, the case is judged not so much in terms of compliance with specific guidelines, but rather in terms of the "totality of circumstances" that attended the hypnosis session. For instance, in a recently debated case (Karlin & Orne, 1997; Scheflin, 1997), in Borawick v. Shay (68 E3d 597, 607 2d Cir.1995, cer. denied, 116 S.Ct. 1869, 134 LEd.2d 966 [1996]), a 30-year-old plaintiff accused her aunt and uncle of sexually abusing her when she was in their home. The victim reported that she had had no memory of the abuse for 20 years. However, during hypnosis, she reported that she recalled her aunt and uncle had abused her, but she did not remember the event after hypnosis was terminated. Some months after the last hypnotic session, she recalled being sexually abused by her father, and, as a result, in 1992 she filed suit. In evaluating the "totality of circumstances" attending this case, the district court excluded the testimony, stating that the hypnotist was not qualified. The totality of circumstances test is consistent with the Daubert principle in that the trial judge is charged with the responsibility of ruling (under Federal Rule 702) whether the expert is testifying to scientific knowledge, after an assessment of whether the "reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue" (Daubert v. Merrell Dow Pharmaceuticals, Inc., 1993, p. 12). This rule extends the Frye test in that whereas general acceptance in the scientific community is still a consideration, other scientific evidence can be entered into testimony that is deemed by the judge to be "scientific." This judgment is not necessarily straightforward, in that it is not always easy to gauge the extent to which testimony meets the criteria for scientific testimony under Daubert. These criteria include falsifiability of the underlying theory, determination of error rate, peer review and publication, and general acceptance in the relevant scientific community (see Fagiman, Kay, Saks, & Sanders, 1995). Daubert, at least as it was interpreted in the Borawick v. Shay (1995) case, leaves the door open to broad judicial interpretations and discretion about what evidence to admit. Perry et al. (1996) have observed that the Daubert ruling is more liberal than Frye's general acceptance criterion and argue that the fate of hypnotically elicited testimony under Daubert is not altogether clear. One possibility is that it may lead to the gradual erosion Alabama, Colorado, Idaho, Kentucky, Mississippi, New Jersey, New Mexico, Ohio, South Carolina, South Dakota, Texas, and Wisconsin. 3

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

331

of per se exclusions of such testimony. On the other hand, it may mean that hypnosis will continue to be excluded from many courts, not because of "general acceptability grounds, but on Daubert standards of unreliability" (Gianelli, 1995, p. 78). Moreover, as Underwager and Wakefield (1998) have noted, "Errors by a biased or incompetent judge claiming to consider the totality of the circumstances and admitting evidence with no scientific validity or credibility would be very difficult to reverse" (p. 398). The courts in many cases used Daubert to reject scientifically unsound evidence, and it is noteworthy that hypnotically elicited testimony has been excluded because of the recognition of suggestibility effects (Underwager & Wakefield, 1998). Because Daubert is a statutory rather than a constitutional case, it is not binding on the states, which have the option to follow Frye (Gianelli, 1994). Although a sizable minority of states follow the "admissibility with safeguards rule" and the federal courts follow the "totality of circumstances test" (see Brown et al., 1998), as we have seen, the majority of states (Perry et al., 1996) have rejected the idea of admitting hypnotic testimony on a case-by-case basis and considering the totality of circumstances, and, instead, have opted for a broad per se exclusion rule. Admissibility with safeguards standards implies that the accuracy of memories can be improved by following certain procedural guidelines. The Hurd rules (1981) require that certain procedural safeguards be implemented before hypnotically elicited testimony can be brought to the bar. The idea that the totality of circumstances must be considered prior to adjudging testimony as admissible implies that the impact of hypnotic procedures on testimony can be evaluated with respect to the guidelines or procedures followed. In the case of Borawick v. Shay (1995), in which a totality of circumstances rule was adopted, the following determinants of admissibility were noted: (1) whether hypnosis was used for clinical or forensic purposes; (2) "whether the witness received any suggestions from the hypnotist or others prior to or during hypnosis"; (3) "the presence of absence of a permanent record" such as a videotape; (4) the hypnotist's qualifications; (5) "whether corroborating evidence exists to support the reliability of the hypnotically-refreshed memories"; (6) the subject's hypnotizability and "proneness to confabulate"; (7) expert opinion regarding "the reliability of the procedures used in the case"; and (8) pretrial testimony. The American Society of Clinical Hypnosis (ASCH; Hammond, Garver, et al., 1995) had proffered a set of guidelines that combine elements of both Hurd and the federal model. The ASCH guidelines are arguably more stringent in requiring (a) the objective measurement of hypnotizability; (b) a discussion of the imperfections of memory "in and out of hypnosis"; (c) a posthypnotic discussion with investigating officers or attorneys regarding the "potential limitations of additional material that was obtained, and that all of the information must be independently corroborated for it to have credibility" (p. 4 7); and (d) a written report of the session with salient details noted (e.g., observations, "depth of trance, phenomena elicited, hypnotizability score"). What impact these guidelines will have, or whether they will replace the Hurd guidelines as the standard of the field, is difficult to determine, in that there have been no court cases to our knowledge that have used these guidelines to date. Procedural guidelines have not been universally embraced. Critics of the guidelines argue that hypnotically elicited testimony

332

MAZZONI AND LYNN

is so unreliable that no array of procedural safeguards can outweigh the inherently biasing effect of hypnosis. For example, the Zayas court (People v. Zayas, 546 N.E. 2nd at 513 1989) noted that regardless of guidelines being in place, the subject will "inevitably fill in missing details and fantasize about events that never in fact happened. Procedural safeguards do nothing, furthermore, to eliminate the problems associated with bolstered witness confidence and erroneous juror perception." Also rejecting the Hurd guidelines, the Hughes court (People v. Hughes, 59 N.Y. 2d 523, 466 N.Y.S. 2d 255, 453 N.E. 2d 484, 1983), in reference to the problem of confabulation, opined that there is not "any scientific method for detecting this type of 'recollection,' absent objective external verification of the statements generated by the hypnotic session." The observation in People v. Gonzales (108 Mich. App. 145, 310 N.W 2d 306 (1981), aff'd, 415 Mich.615, 329 N.W 2d 743 (1982) that the Hurd guidelines have the potential to confer an unwarranted "aura of reliability" on hypnotic testimony, thereby raising the specter of a miscarriage of justice, captures the essence of objections to the safeguard approach. This criticism can, likewise, be voiced with regard to the guidelines proposed by the "federal" model, developed by the FBI for use in criminal investigations (Ault, 1979) and adopted by many federal agencies. The federal model specifies a somewhat more lenient set of guidelines (e.g., contrary to Hurd, the people in the room during the session may include attorneys, a police artist, a close friend, and so forth) that can be subjected to the same criticisms as Hurd. Whereas it is possible to speculate about the relative merits of one set of guidelines over another, in actuality, it is difficult to evaluate the strengths and limitations of any of the procedures, given the dearth of systematic research. The limited evidence from a number of studies of the effects of prehypnotic "warnings" about the imperfections of hypnotically elicited recall does not inspire a great deal of confidence in their protective effect. In one study (Burgess & Kirsch, 1999), warnings mitigated some of the memory distortions associated with hypnosis but did not improve recall above and beyond a nonhypnotic condition; in another study, warnings minimized memory distortions during but not after hypnosis (Green, Lynn, & Malinoski, 1998); and in a third study (Neuschatz, Lynn, Benoit, & Fite, 2003), repeated warnings (e.g., that hypnotic and nonhypnotic memories are not necessarily accurate), which were taken directly from those recommended by ASCH and delivered before, during, and after hypnosis, did not improve recall relative to a nonhynotic condition. It can be argued that the special status of hypnosis, as exemplified by per se exclusion rules, implies not only that hypnosis affects memory reports negatively, but also that it is more pernicious than other procedures (e.g., the use of misleading questions), the use of which does not result in a per se exclusion. However, comparisons of effects of hypnosis to those of misleading questions do not support this conclusion. Scoboria and his colleagues have reported two studies in which they compared the effects of hypnotic questioning with those of misleading questions (Scoboria et al., 2001, 2004). In these studies, hypnosis was not more detrimental to memory accuracy than the presence of misleading questions. In fact, misleading questions produced significantly more false memories both immediately and after a delay than did hypnosis.

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

333

CONCLUSIONS

The data clearly indicate that hypnosis is not an aid to memory. On the contrary, hypnosis can reduce the accuracy of eyewitness testimony and create an unjustified increase in confidence in what is remembered. These data have led in many states to a per se exclusion of testimony from hypnotized witnesses. Similarly, professional societies have cautioned against the use of hypnosis for memory enhancement. For example, the American Psychological Association-Division 17 (1995) guidelines state that hypnosis should not be used for memory recovery, a conclusion also reached by the Canadian Psychiatric Association (1996). The American Psychological Association (1995) recommended that hypnosis not be used for clients who are attempting to retrieve or confirm recollections of histories of abuse, and the American Medical Association (AMA, 1994) asserted that hypnosis should be used only for investigative purposes in forensic contexts. 4 Similar warnings are in place in other countries. In the United Kingdom, for example, the British Society for Clinical and Experimental Hypnosis has become acutely aware of the danger involved in using hypnosis as a tool to recover memories. UK practitioners, in order to become members of the society, have to state that they will not use hypnosis with their clients to help them recover memories about their past. Even refreshing recall in witnesses through hypnotic suggestions is strongly discouraged. Just as the benefits of hypnosis have been exaggerated, so too have its dangers relative to other suggestive approaches. As noted above, the effects of hypnosis on eyewitness testimony, while detrimental, do not seem to be as great as the effects of misleading questions (Scoboria et al., 2001). These data could be construed as undermining arguments for a per se exclusion rule: prohibiting testimony from witnesses who have been asked leading questions might leave one with no witnesses at all. Additionally, no scientific studies have shown that recall errors associated with the focus of a hypnotic interrogation (e.g., description of a face) render testimony unrelated to the hypnotically elicited testimony (e.g., the sequence of events that occurred) unreliable. Accordingly, a per se exclusion rule that is so broad that it prohibits a person from testifying even about "nonhypnotic" recollections may be seen by some courts as unduly restrictive. Per se exclusion also fails to acknowledge that suggestive procedures that are used in preparing witnesses for trial and in police investigations, for example, have the potential to bias and distort memories (Spanos, Quigley, Gwynn, Glatt, & Perlini, 1991). Combined, these observations imply that it may not be in the interests of courts to bar all testimony from a hypnotized witness. The counterargument is that there is ample justification for the per se exclusion rule, given the well-documented unreliability of hypnotically assisted memories. The mere fact that nonhypnotic suggestive procedures such as leading questions also lead to recall errors does not constitute an argument for the admission of hypnotically elicited 4 Whereas it is possible that false memories induced in a clinical context may be difficult to reverse, our experience as researchers suggests that false memories created in experimental contexts are easily reversed during debriefing, regardless of whether hypnosis or nonhypnotic procedures have been used.

334

MAZZONI AND LYNN

memories to the bar. A broad per se exclusion rule can also be justified on the basis of the cost of hearings to consider the admission of such testimony and the perception that it is important to discourage the use of hypnosis to refresh eyewitness recall. Following this line of reasoning, in order to abolish per se exclusion, it would be necessary for research to ascertain (a) the circumstances, if any, under which hypnosis does not compromise eyewitness memory, and (b) the means of eliminating the pernicious effects of hypnosis on memory. If it could be shown that hypnosis improves recall in ecologically valid situations, such as witnessing an actual or staged crime, a stronger argument could be made for a role for hypnosis in the courtroom. However, no such data are available at this time. In fact, hypnosis does not facilitate the recall of emotional memories, such as the death of Princess Diana, and has been shown to compromise recall compared with a condition in which participants are simply asked to do their best to remember (Krackow, Lynn, & Payne, 2004). Given the current knowledge on the effects of hypnosis on memory, as a general rule hypnosis should not be used to assist recall in a forensic or any other situation. That said, it may be possible to envision certain exceptions. The one positive finding regarding hypnosis and recall is that hypnosis can produce a greater volume of recall (with accurate and inaccurate details) than nonhypnotic methods. In rare instances (e.g., recall of a license plate in a kidnapping) when accurate recall that can be corroborated is a priority, and inaccuracy in detail can be tolerated, then hypnosis may well prove fruitful in criminal investigations after nonhypnotic, nonsuggestive methods have repeatedly failed to yield useful leads. When hypnosis is contemplated in such desperate circumstances, it would be advisable to have a hearing on the viability of the use of hypnosis, although various exigencies, including time and cost pressures, and problems in determining operative criteria for the use of last resort recall enhancement procedures, may render this option impracticable. Still, even in those highly unusual situations in which hypnosis is deemed acceptable, previous guidelines do not go far enough in that they do not recommend assessing participants' confidence in their initial and posthypnotic recall. Without this information, it is impossible to fully evaluate the potentially biasing effect of hypnosis and hypnotic testimony on the witness. It is important to underscore the point that even when hypnosis is used solely for investigative purposes, there are attendant risks. That is, hypnotically elicited inaccurate information could lead investigators to pursue erroneous leads and even to interpret subsequent leads as consistent with initial (and perhaps mistaken) hypnotically generated evidence. Exaggeration of the promises and perils of hypnosis may be related to the mythology that surrounds the phenomenon. The belief that hypnosis can enhance recall far beyond what is ordinary is common not only to therapists but also to many individuals in the general population. This belief may be the source of the memory distortions that result from hypnosis, particularly the increased confidence that accompanies hypnotically "enhanced" recall. Indeed, memory recovery procedures that convey the false expectation that accurate memories of forgotten or poorly encoded events can easily be recovered are likely to increase the sheer volume of memories and bolster confidence in inaccurate as well as accurate memories (e.g., Lynn, Myers, & Malinoski, 1997). Of course,

11. USING HYPNOSIS IN EYEWITNESS MEMORY

335

such memory enhancement procedures should be scrupulously avoided and have no place in the courtroom.

REFERENCES American Medical Association, Council on Scientific Affairs. (1994). Memories of Childhood Abuse. CSA Report 5-A-94. American Psychological Association, Division 17 Committee on Women, Division 42 Trauma and Gender Issues Committee. (1995, July 25). Psychotherapy guidelines for working with clients who may have an abuse or trauma history. Washington, D.C: Author. Bass, E., & Davis, L. (1988). The courage to heal. New York: Harper & Row. Bernheim, H. (1891/1973). Hypnotisme, suggestion, psychotherapie; etudes nouvelles. Paris: Octave Duin (trans. New York, Aronson). Blume, E. S. (1990). Secret survivors: Uncovering incest and its aftereffects in women. New York: John Wiley & Sons. Borawick v. Shay, 68 F.3d 597 607 (2d Cir. 1995) cert. denied, U.S., 116 S.Ct. 1869, 134 L.Ed.2d 966 (1996). Brown, D., Scheflin, A. W., & Hammond, D. C. (1998). Memory, trauma treatment, and the law. New York: W. W. Norton & Company. Bruhn, A. R. (1984). The use of earliest memories as a projective technique. In P. McReynolds & C. J. Chelune (Eds.), Advances in psychological assessment (Vol. 6, pp. 109-150). San Francisco: Jossey-Bass. Burgess, C., & Kirsch, I. (1999). Expectancy information as a moderator of the effects of hypnosis on memory. Contemporary Hypnosis, 16, 22-31. Burnham, W. H. (1889). Memory, Historically and experimentally considered. Ill: Paramnesia. American]oumal of Psychology, 2, 430-464. Canadian Psychiatric Association (1996, March 25). Position statement: Adult recovered memories of childhood sexual abuse. Canadian Journal of Psychiatry, 41, 305-306. Daubert v. Merrel Dow Pharmacueticals, Inc., 113 S. Ct. 2786 (1983). Erdelyi, M. (1994). Hypnotic hypermnesia: The empty set of hypermnesia. Intemational]oumal of Clinical and Experimental Hypnosis, 42, 379-390. Fagiman, D., Kay, D., Saks, M., &. Sanders, J. (Eds.). (1995). Modem scientific evidence: The law and science of expert testimony (pp. 480-527). St. Paul, MN: West Publishing Co. Forel, A. (1906). Hypnotism or suggestion and psychotherapy: a study of psychological, psychophyisiological and therapeutic aspects of hypnosis. New York: Rebman. Fredrickson, R. (1992). Repressed memories. New York: Fireside/Parkside. Freud, S. (1909/1955). Constructions in analysis. In J. Strachey (Ed. and Trans.), The standard edition of the complete work of Sigmund Freud. London: Hogarth Press. Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). Giannelli, P. C. (1995). The admissibility of hypnotic evidence in U.S. courts. International journal of Clinical and Experimental Hypnosis, 43, 212-233. Green, J. P., Lynn, S. J ., & Malinoski, P. (1998). Hypnotic pseudomemories, pre hypnotic warnings, and the malleability of suggested memories. Applied Cognitive Psychology, 12, 431444. Hammond, D. C., Garver, R. B., Mutter, C. B., Crasilneck, H. B., Frischholz, E., Gravitz, M.A., et al., (1995). Clinical hypnosis and memory: Guidelines for clinicians and for forensic hypnosis. Des Plaines, IL: American Society of Clinical Hypnosis Press. Harding v.State, 246 A2d 302 (Md. Ct. Spec. App. 1968). Cert. denied, 395 U.S. 949 (1969). Haward, L. R. C. (1988). Hypnosis by the police. British]oumal of Experimental and Clinical Hypnosis, 5, 33-35.

336

MAZZONI AND LYNN

Haward, L., & Ashworth, A. (1980, August). Some problems of evidence obtained by hypnosis. Criminal Law Review, 469-485 Herman,]. L. (1992). Trauma and recovery. New York, Basic Books. Hibbard, W. S., & Worring, R. W. (1981). Forensic hypnosis: The practical application of hypnosis in criminal investigation. Springfield, IL: Thomas. Horsley,]. S. (1936). Narco-analysis. Lancet, 130, 55-56. Hyman, I. E., & Pentland, J. (1996). The role of mental imagery in the creation of false childhood memories.]ournal of Memory and Language, 35, 101-117. Janet, P. (1889/1973). L'automatisme psychologique. Paris: Alcan. Karlin, R. A., & Orne, M. T. (1997). Hypnosis and the iatrogenic creatin of memory: On the need for a per se exclusion of testimony based on hypnotically influenced recall. Cultic Studies ]oumal, 14, 172-206. Kebbell, M. R., & Wagstaff, G. (1997). Hypnotic interviewing: The best way to interview eyewitnesses? Behavioral Sciences and the Law, 16, 115-129. Kihlstrom, J. F. (1998). Exhumed memory. InS.]. Lynn & K. M. McConkey (Eds.), Truth in memory (pp. 3-31). New York: Guilford. Kleinhauz, M., Horowitz, 1., & Tobin, T. (1977). The use of hypnosis in police investigation: A preliminary communication. ]oumal of the Forensic Science Society, 17, 77-80. Krackow, E., Lynn, S. ]., & Payne, D. (in press). The death of Princess Diana: Hypnosis, context re-

instatement, and task motivation in the recall of flashbulb memories. Imagination, Cognition and Personality. Laurence,]. R., & Perry, C. (1983). Hypnotically created memory among highly hypnotizable subjects. Science, 222, 523-524. Loftus, E. F. (1993). The reality of repressed memories. American Psychologist, 48, 518-537.

Lynn, S. ]., Knox,]., Fassler, 0., Lilienfeld, S. 0., & Loftus, E. (2004). Trauma, dissociation, and memory. In]. Rosen (Ed.), Posttraumatic stress disorder: Issues and controversies. New York: Wiley. Lynn, S. ]., Fassler, 0., Knox,]., & Lilienfeld, S. 0. (in press). Dissociation and dissociative identity disorder: Treatment guidelines and cautions. In]. Fisher & W. O'Donohue (Eds.), Practitioner's guide to evidence based psychotherapy. New York: Kluwer Academic Publications. Lynn, S. ]., Malinoski, P., Green,]., Marmelstein, L., & Stafford,]. (2000). Autobiographical memories, hypnotizability, and suggestion. In V. Pascalis, V. Gheorghieu, I. Kirsch, & P. Sheehan (Eds.), Suggestion and suggestibility. Berlin: International Hypnosis Monographs. Lynn, S.] ., Myers, B., & Malinoski, P. (1997). Hypnosis, pseudomemories, and clinical guidelines: A sociocognitive perspective. In D. Read & S. Lindsay (Eds.), Recollections of trauma: Scientific studies and clinical practice (pp. 305-336). New York: Plenum Press. Lynn, S. ]., & Nash, M. R. (1994). Truth in memory: Ramifications for psychotherapy and hypnotherapy. American ]oumal of Clinical Hypnosis, 36, 194-208. Lynn, S. ]., Neuschatz, ]., Fite, R., & Rhue, J. R. (2001). Hypnosis and memory: Implications for the courtroom and psychotherapy. In M. Eisen & G. Goodman (Eds.), Memory, suggestion, and the forensic interview. New York: Guilford. Malinoski, P., Lynn, S. J., &Sivec, H. (1998). The assessment, validity, and determinants of early memory reports: A critical review. InS.]. Lynn & K. McConkey (Eds.), Truth in memory. New York: Guilford. Marmelstein, L., & Lynn, S. ]. (1999). Expectancies, group, and hypnotic influences on early autobiographical memory reports. International ]oumal of Clinical and Experimental Hypnosis, 47, 301-319. Mazzoni, G., & Loftus, E. F. (1998). Dreaming,believing, and remembering. In J. deRivera & T. R. Sarbin (Eds.), Believed-in imaginings. The narrative construction of reality (pp. 145-156). Washington, DC: American Psychological Association. Mazzoni, G. A., Loftus, E. F., Seitz, A., & Lynn, S. ]. (1999). Creating a new childhood: Changing beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13, 125-144.

11.

USING HYPNOSIS IN EYEWITNESS MEMORY

337

Mazzoni, G., & Memon, A. (2003). Imagination can create false memories. Psychological Science, 14(2), 186-188. McConkey, K., & Sheehan, P. W. (1995). Hypnosis, memory, and behavior in criminal investigation. New York: Guilford. Nash, M. R. (1987). What, if anything, is regressed about hypnotic age regression? A review of the empirical literature. Psychological Bulletin, 102, 42-52. Nash, M. J., Drake, M., Wiley, R., Khalsa, S., & Lynn, S. ]. (1986). The accuracy of recall of hypnotically age regressed subjects. Journal of Abnormal Psychology, 95, 298-300. Neuschatz, J., Lynn, S. J., Benoit, G., & Fite, R. (2003). Hypnosis and memory illusions: An investigation using the Deese/Roediger paradigm. Imagination, Cognition, and Personality, 22, 3-12. Nogrady, H., McConkey, K. M., & Perry, C. (1985). Enhancing visual memory: Trying hypnosis, trying imagination, trying again. Journal of Abnormal Psychology, 94, 195-204. Orne, E. C., Whitehouse, W. G., Dinges, D. F., & Orne, M. T. (1996). Memory liabilities associated with hypnosis: Does low hypnotizability confer immunity? International Journal of Clinical and Experimental Hypnosis, 44, 354-369. People v. Ebanks, 49P 1049 (Cal1897). People v. Hughes, 59 N.Y. 2d 523, 466 N.Y.S. 2d 255, 453 N.E. 2d 484, (1983). People v. Shirley, 641 P.2d 775 (Cal.) cert. denied, 439 U.S. 860 (1982). People v. Zayas, 546 N.E. 2nd at 513 (1989). Perry, C., Orne, M. T., London, R. W., & Orne, E. C. (1996). Rethinking per se exclusions of hypnotically elicited recall as legal testimony. International Journal of Clinical & Experimental Hypnosis, 44, 66-81. Poole, D. A., Lindsay, D. S., Memon, A., & Bull, R. (1995). Psychotherapists' opinions, practices, and experiences with recovery of memories of incestuous abuse. Journal of Consulting and Clinical Psychology, 68, 426-437. Putnam, W. H. (1979). Hypnosis and distortions in eyewitness memory. International}ournal of Clinical and Experimental Hypnosis, 27, 437-488. Ready, D. J., Bothwell, R. K., & Brigham, J. C. (1997). The effects of hypnosis, context reinstatement, and anxiety on eyewitness memory. International Journal of Clinical and Experimental Hypnosis, 45, 55-68. Reiser, M. (1980). Handbook of investigative hypnosis. Los Angeles: Lehi. Rock v. Arkansas, 483 U.S. 44 (1987). Sanders, G. S., & Simmons, W. L. (1983). Use of hypnosis to enhance eyewitness accuracy: Does it work? Journal of Applied Psychology, 68, 70-77. Scheflin, A. (1997). False memories and Buridan's ass: A reply to Karlin and Orne. Cultic Studies Journal, 14, 207-289. Scheflin, A. W., Spiegel, H., & Spiegel, D. (1999). Forensic uses of hypnosis. In A. K. Hess & I. B. Weiner (Eds.), The handbook of forensic psychology (2nd ed., 474-498). New York: John Wiley & Sons. Scoboria, A., Mazzoni, G., & Kirsch, I. (2004). Separate and combined effects of misinformation and hypnosis upon eyewitness memory. Poster presented at the American Psychological Society annual conference, Chichago, IL. Scoboria, A., Mazzoni, G., Kirsch, I., & Milling, L. S. (2001). Immediate and persistent effect of misleading questions and hypnosis on memory reports. Journal of Experimental Psychology: Applied, 8, 26-32. Sheehan, P. W., & Tilden, J. (1983). Effects of suggestibility and hypnosis on accurate and distorted retrieval from memory. Journal of Experimental Psychology: Learning, Memory, & Cognition, 9, 283-293. Sivec, H., & Lynn, S. ]. (1996). Early life events: Hypnotic vs. nonhypnotic age regression. Unpublished manuscript, Ohio University. In Lynn, Neuschatz, Fite et al. (Eds.), Hypnosis and memory: Implications for the courtroom and psychotherapy, in press.

338

MAZZONI AND LYNN

Sivec, H. J., Lynn, S. J., & Malinoski, P. T. (1997). Early memory reports as a function of hypnotic and nonhypnotic age regression. Unpublished manuscript, State University of New York at Binghamton. In Lynn, Neuschatz, Fite et al. (Eds.), Hypnosis and memory: Implications for the courtroom and psychotherapy, in press. Spanos, N. P. (1996). Multiple identities and false memories: A sociocognitive perspective. Washington, DC: American Psychological Association. Spanos, N. P., Quigley, C. A., Gwynn, M. 1., Glatt, R. L., & Perlini, A. H. (1991). Hypnotic interrogation, pretrial preparation, and witness testimony during direct and cross-examination. Law and Human Behavior, 15, 639-653. State v. Hurd, 432 A.2d 86 (N.J. 1981). Steblay, N. M., & Bothwell, R. K. (1994). Evidence for hypnotically refreshed testimony: The view from the laboratory. Law and Human Behavior, 18, 635-651. Terr, L. (1994). True stories of traumatic memories, lost and found. New York: Basic Books. Underwager, R., & Wakefield, H. (1998). Recovered memories in the courtroom. InS. J. Lynn & K. M. McConkey (Eds.), Truth in memory (pp. 394-436). New York: Guilford. Wagstaff, G. F. (1999). Hypnosis and forensic psychology. In I. Kirsch, A. Capafons, E. CardenaBuelna, & S. Amigo (Eds.), Clinical hypnosis and self regulation (pp. 277-310). Washington, DC: American Psychological Association. Wells, G., & Bradfield, A. L. (1998). "Good, you identified the suspect:" Feedback to eyewitnesses distorts their reports of the witnessing experience. Journal of Applied Psychology, 83, 360-376. Whitehouse, W. G., Dinges, D. F., Orne, E. C., & Orne, M. T. (1988). Hypnotic hypermnesia: Enhanced memory accessibility or report bias? Journal of Abnormal Psychology, 97, 289-295.

12 Credibility Assessment in Eyewitness Memory Dorothee Griesel and John

c. Yuille

University of British Columbia

TYPES OF STATEMENTS AND DEFINITION OF LYING

There are three categories of eyewitness statements: (1) a generally correct memory (which may contain some errors), (2) a deceptive attempt, or (3) a historically wrong but subjectively true narrative, that is, an erroneous account provided without intention to mislead. The latter type of memory is called a mistaken memory and can be created, for example, through suggestive interviewing techniques in therapeutic settings (Lindsay & Read, 1995; Loftus, 1993; Porter, Birt, Yuille, & Herve, 2001; Porter & Campbell, 2002). Most deception detection techniques are based on the assumption that the intention to deceive creates cognitive and emotional consequences that are detectable. Because no deceptive intent underlies mistaken memories, they are not the focus of the techniques introduced in this chapter. Thus, this chapter addresses the distinction between true and deceptive accounts. Ekman (1996) defined a lie by two criteria: first, the target of the lie is deliberately misled and, second, the target is not notified about the liar's intention to deceive. Similarly, DePaulo and colleagues (2003) defined deception as "a deliberate attempt to mislead others" (p. 74). These definitions state that the lie is expressed to another person, thereby differing from self-deception (Zuckerman & Driver, 1985). Besides the intentional and interpersonal aspects, Zuckerman and Driver further defined deception as a conscious act, another characteristic differentiating lying from self-deception. Ekman (1997b) also set lies apart from broken promises. It might be excusable or even appropriate for people not to act according to their promises if current circumstances require actions that are different from what was promised in the past. Ekman gave the example of a politician who broke his election campaign promises after the political circumstances had changed. Furthermore, a failure to remember is not a lie, because no choice is involved (Ekman, 1997b). 339

340

GRIESEL AND YUILLE

Lies can be expressed in the form of falsifications or as concealments, both of which qualify as lies, inasmuch as the underlying motive is to mislead. However, there are at least three other ways of lying (Ekman, 1996, 1997b). First, one could be telling the truth falsely, that is, speaking the truth but making it seem like the opposite is actually meant. Second, a half~truth could be told. The third way is the incorrect inference dodge, a term coined by Ekman. An example would be a situation in which a person does not like a piece of art but does not want to express his/her dislike to the artist. The person might comment: "Incredible, I can't believe it!" (Ekman, 1997b). In this chapter the different approaches to and tools of credibility assessment are introduced and evaluated according to the manner in which lies are presented and may betray themselves (e.g., speech style, body language, facial expression, speech con~ tent). Zuckerman, DePaulo, and Rosenthal (1981; as cited in Hess & Kleck, 1997 /1990) noted four different markers of lying: (1) those produced by voluntary control in~ valved in deceptive behavior, (2) psychophysiological markers (such as blinking, pupil dilation), (3) indicators of affective states (such as guilt), and (4) markers due to cog~ nitive demands that arise when a deceptive message is being produced. In this chapter, psychophysiological, behavioral (nonverbal) and verbal lie detection techniques are distinguished. Psychophysiological and behavioral techniques mostly originated in North America and are traditionally concerned with the detection of lies in the strict sense (cf. Yuille, 1989b). Verbal strategies of credibility assessment, on the other hand, examine semantic and stylistic changes in people's statements. These approaches were mostly developed in Europe in the context of assessing allegations of child sexual abuse. These approaches assess for credibility in a broader sense, that is, they are not only concerned with the detection of deliberate lies. The chapter closes with a brief section on factors that modify deception cues, such as characteristics of the liar and the lie recipient.

RATIONALE BEHIND CREDIBILITY ASSESSMENT PROCEDURES

Ekman (1997b) assumed two reasons why lies fail: cognitive load and emotional load. Cognitive load results in clues primarily detectable in the liar's speech, for example, errors, latencies and pauses, as well as in the content of speech. Emotional load is re~ vealed in facial expressions, tone of voice, and/or body language. Emotions can be fabri~ cated or concealed in order to make lies appear truthful. Ekman coined the term leakage to refer to behaviors that betray the lie: "A deception clue answers the question of whether or not the person is lying, although it does not reveal what is being concealed" (Ekman, 2001, p. 40). Two kinds of emotions may betray a lie: a false presentation of an emotion or an emotional consequence of the lie. For example, a person could have a fear of being caught or guilt about the immoral act oflying (DePaulo et al., 2003). Referring to the fear of being caught, Ekman introduced the term detection apprehension; many factors affect this apprehension, including the reputation of the lie detector, the stakes

12.

CREDIBILITY ASSESSMENT

341

related to the lie, the lying experience of the liar, etc. Lies may also be accompanied by positive emotions: pleasure (duping delight), excitement, pride, and relief (at not being caught) (e.g., Ekman, 2001). It is important to bear in mind that no behavior is diagnostic of lying. All credibility assessment techniques reviewed in the following sections detect behavioral changes rather than lies per se. The reason for the behavioral change may be a lie, but the change could be due to many other things. The most difficult problem in lie detection is making the correct attribution of the behavioral change (i.e., whether to a lie or some other cause).

PHYSIOLOGICAL APPROACHES TO CREDIBILITY ASSESSMENT

The polygraph is one of the oldest instruments used in applied psychology research. It is a physiological technique used to detect behavioral changes during investigations of crime suspects. The first attempts to detect deception via physiological measures were made in Italy in the late nineteenth century (Kircher & Raskin, 1992). The physiological measures recorded in a polygraph examination are usually skin conductance, heart rate, blood pressure, and respiration. Today, different polygraph tests can be distinguished: the Relevant-Irrelevant Test, the Control Question Test, the Guilty Knowledge Test, and the Directed Lie Test. The Control Question Test, developed by Reid in the late 1940s, represents the most commonly used polygraph technique (Elaad, 1998; Honts, Kircher, & Raskin, 1995; Honts, Raskin, & Kircher, 1994). The following sections provide an overview of the rationale behind each of these polygraph tests.

Relevant-Irrelevant and Control Question Tests The Relevant-Irrelevant Test (RIT) and the Control Question Test (CQT) are based on comparing responses to two types of questions, which must be answered yes or no. In the RIT, the individual's physiological response to irrelevant questions concerning neutral topics is compared with the reaction shown in response to relevant questions, which are case specific, accusatory, and directly concerned with the matter under investigation (Honts et al., 1995). An example of a relevant question would be, "Did you stab Mr. X?" It is thought that innocent and guilty subjects respond differently to such relevant questions: For guilty subjects, the physiological response to relevant questions ought to be more intense than their response to irrelevant questions, whereas for innocent subjects, the responses to relevant and irrelevant questions should be equal (Kircher & Raskin, 1992). As Honts and colleagues (1995) noted, unfortunately, not only guilty suspects, but also uninvolved, truthful individuals recognize relevant questions as relevant and respond physiologically. Reactions to relevant questions might be caused by factors other than deception, such as the provocative nature of the questions or anxiety over being tested in general (Raskin, Kircher, Horowitz, & Honts, 1989).

342

GRIESEL AND YUILLE

Thus, the RIT is likely to produce a large rate of false positive errors, classifying truthfully innocent examinees as guilty (Honts et al., 1995). The CQT paradigm also involves relevant questions. In contrast to the RIT, however, they are compared with control questions. Control questions are constructed during the pre-polygraph interview. They are intentionally vague, concern a period of the subject's life, and refer to acts that most individuals are likely to have committed but are embarrassed to admit. Thus, the interviewee is expected to answer them with a not fully truthful "no" and, therefore, show some physiological arousal (Kircher &Raskin, 1992). During the pre-polygraph interview, subjects are told that people who committed a crime of the kind under investigation are likely to have committed acts of the kind asked for in the control questions (Honts et al., 1995). An example of a control questions would be, "During the first 22 years of your life, did you ever take something that did not belong to you?" (Raskin et al., 1989, p. 297). Control questions are designed to be similar to the relevant questions to give innocent suspects the opportunity to become more concerned about questions other than the relevant questions (Raskin et al., 1989). It is expected that innocent subjects react the same or more strongly to control questions than to relevant questions (Kircher & Raskin, 1992). A CQT usually involves at least three repetitions of such pairs of questions plus several unevaluated filler questions. In each repetition, the different types of questions are presented in a different order (Honts et al., 1995). Several assumptions are made when the CQT is applied: for example, that the subject can be forced to produce lies in response to the control questions; that a lie in response to a control question elicits less physiological response than a more serious lie in response to a critical question; that responses to all replications of the questions are of equal weight; that questions tailored for specific cases are comparable to each other (Fiedler, Schmidt, & Stahl, 2002). The evaluation should be based solely on the polygraph charts and not include behavioral cues displayed during the procedure or other case information known to the examiner (Raskin et al., 1989). To improve the polygraph's objectivity and reliability, computer-based methods have been introduced to assist the scoring of the data (Kircher & Raskin, 1988; cited in Honts et al., 1995). The CQT is currently the most widely used technique in criminal investigations. Its purpose is to reduce the number of suspects who require further investigation toreduce the costs of criminal investigations. Obviously, the polygraph is typically used with crime suspects, not witnesses or victims. Indeed, Kircher and Raskin (1992) mentioned that in some jurisdictions prosecutors dismiss charges if a suspect successfully passes a polygraph examination. The CQT is also used in national security programs to test loyalty.

The Guilty Knowledge Test The Guilty Knowledge Test (GKT), also called the Concealed Knowledge Test, assesses whether a suspect has knowledge about the crime that only a guilty person would have (Kircher & Raskin, 1992; Ben-Shakkar, Bar-Hillel, & Kremnitzer, 2002). The test consists of a series of multiple-choice questions where the correct detail is embedded

12.

CREDIBILITY ASSESSMENT

343

among various erroneous details. For example, if a bank robbery involved stealing 1.5 mil~ lion dollars, the control answers could be 1.0, 1.2, and 1.8 million dollars. Usually, the examinee is instructed to answer "no" to all the presented alternatives, including the correct answer. The GKT does not aim to detect guilt, but guilty knowledge. The ratio~ nale behind the GKT is the theory of orienting response and habituation, the former be~ ing evoked by a novel stimulus, a change in stimulation, or a stimulus of signal value. The GKT is used far less than the CQT. This is due to a number of conditions that have to be met for its proper administration. First, the individual must have noticed certain relevant details at the crime scene and, moreover, have formed a memory of them. Second, the details chosen for the GKT examination must not have leaked out to potential examinees (Ben~Shakkar et al., 2002; Elaad, 1998). Crime facts chosen for the GKT should be familiar only to the investigator and the truly guilty suspect of the crime (MacLaren, 2001). Elaad (1998) further mentioned that all alternatives of the multiple~ choice question should be equally likely to point at the examinee's innocence. Mac~ Laren (2001) listed four different applications of the GKT: (1) criminal investigations, to find out whether a suspect has guilty knowledge; (2) cases where a witness knows the perpetrator of a crime but is reluctant to disclose his information; (3) attempts to verify the security of private information (e.g., to test employees if they have access to pass~ words outside the area of their own responsibilities); and (4) the searching GKT, in which an examiner attempts to identify the correct answer (unknown to himself) among several limited alternatives.

countermeasures For both the CQT and the GKT, effective countermeasures have been developed. Coun~ termeasures are "deliberate actions that guilty examinees may use in an effort to distort their physiological responses and yield favorable results" (Elaad, 1998, p. 176). Honts and colleagues (1994) published a study that gave rise to concerns regarding the poly~ graph's validity, given the possible use of countermeasures. The study involved a mock crime, in which guilty subjects were instructed to commit a theft, whereas innocent sub~ jects were only told about the theft. There were four countermeasure conditions for the guilty subjects: subjects were trained to use (1) muscle tension, (2) pain (e.g., pressing a toe against the floor), or (3) both of these physical countermeasures. A mental counter~ measure (4) was to count backward by 7s from 200. The findings of the study suggested that both physical and mental countermeasures diminish the accuracy rate of the CQT as long as training was implemented correctly. The polygraph examiner could detect only 12% of the cases that used physical countermeasures and none of the mental coun~ termeasures. Thus, physical and mental countermeasures seriously jeopardize the poly~ graph's valid use. Drugs, on the other hand, do not seem to work effectively as counter~ measures (Elaad, 1998). The problem with drugs is that they cannot be selectively used to lower the response to relevant answers or to enhance physiological arousal to neutral alternatives. Moreover, drugs may have selective effects on only one of the physiological measures used in polygraph examinations.

GRIESEL AND YUILLE

344

Polygraph Research Research on polygraph examinations is primarily concerned with accuracy, that is, the proportion of correct diagnoses of the examinee as guilty or innocent (Fiedler et al., 2002). Validity deals with the degree to which test results relate to the intention of the assessment-in the case of polygraph examinations the establishment of truth or deception. Generally, two different errors can be distinguished in polygraph tests: false positive and false negative errors. The former refers to the percentage of truly innocent examinees wrongfully diagnosed as guilty; the latter refers to participants who are found innocent but are actually guilty. Data concerning the polygraph's validity can be derived from either laboratory studies (often mock crime simulations) or field situations (evaluation of actual criminal investigations). Laboratory studies have the advantage that the truthful or deceptive status of subjects is known. The major problem with laboratory studies is that subjects differ in their motivations compared to real-life situations. Suspects in real life are obviously more motivated to pass the polygraph than laboratory participants, who are usually promised a cash bonus for successfully passing the test. The major problem of field studies is to find a valid ground truth criterion. The best approximation of ground truth seems to be a complete confession by the suspect (Raskin, 1987), although confessions are usually obtained post-polygraph, so that the truth criterion is confounded with the polygraph outcome. Fiedler and colleagues (2002) pointed out that another problem for field studies is the sampling bias. Only cases for which the validity criterion (truly guilty or innocent) is known can enter a study. Because a confession is the most widely used external criterion to establish ground truth, cases in which the suspect was innocent but failed the test and, consequently, does not confess after his examination would disconfirm the test. Such cases, however, are excluded from research because confessions are only obtained from subjects whose polygraph outcomes are deceptive (Patrick & Iacono, 1991). Another source of error could be false confessions by innocent subjects (cf. Gudjonsson, 2001). Before the research on different polygraph tests is reviewed, two indices that are important in the evaluation of polygraph tests are introduced: specificity and sensitivity. The former is defined as the proportion of correctly identified examinees having knowledge about crime details (i.e., those for whom guilt would be inferred) among all informed participants. Specificity is defined as the proportion of identified noninformed participants among all noninformed participants.

Psychometric Qualities of the CQT The CQT has been in use for over 40 years. The first validity studies were conducted in the early 1970s. Raskin (1987) found accuracy rates of almost 90% and 80% correct decisions for guilty and innocent subjects, respectively. In 1989, Raskin and colleagues reported on four laboratory studies and found very high accuracy rates, 97% for guilty subjects and 93% for innocent subjects. In a review paper on laboratory and field studies of the polygraph, Kircher and Raskin (1992) reported fairly low false negative

12.

CREDIBILITY ASSESSMENT

345

rates, 7% for laboratory and 10% for field studies. The false positive error rate was 13% for laboratory and 20% for field studies. Patrick and Iacono (1991) reported lower rates in a review of 402 CQTs conducted over a 5-year period by federal police examiners in Vancouver. Whenever no confessions were available, other information, such as the judicial outcome, physical evidence, or eyewitness evidence, was collected from police files to establish ground truth. The hit rate was 55% for innocent and 98% for guilty subjects; the latter is questionably high because only a few cases could be verified by independent sources other than confessions. The common pattern is that the CQT provides fairly accurate test results, as long as the examinee is truly guilty. The CQT's major problem is the false positive rate. Elaad (1998) stated that false positive errors should be avoided by any means--even if this means an increase in false negative error rates. However, the false negative rate due to countermeasures is also a serious concern (e.g., Honts & Amato, 2002). More recently, Fiedler and colleagues (2002) looked at the CQT from a psychometric point of view and remarked: "[T]he current CQT practice is simply not the kind of procedure that should be sold in the name of scientific psychology" (p. 323). In detail, they noted that no one has yet produced any research on the CQT's internal consistency, a standard aspect of reliability. As we all know, a test's validity can never exceed its reliability. It remains unclear whether the various physiological measures aggregated in the CQT meet the consistency criterion. What if the findings from different measures do not converge? Also, what if the control questions do not measure one common construct? Moreover, the CQT's objectivity seems to be given only with regard to the interpretation of the results but not the construction of the test, as the control questions are based on the examiner's intuition. In addition, Fiedler and colleagues (2002) disapproved of theunderlying decision model of the CQT, an aspect of construct validity. The assumption that there are psychophysiological correlates of lying is simply wrong. Particular emotions are not correlated with distinguishable physiological patterns (cf., Saxe & Ben-Shakkar, 1999). Consequently, lying cannot be inferred from physiological arousal. It was also doubted that lying in response to control and relevant questions produces different physiological response patterns. The CQT's predictive validity (accuracy) has been extensively researched, but Fiedler and colleagues (2002) emphasized that sampling biases preclude meaningful interpretations of these results. Finally, they remarked, the CQT should not be evaluated against a 50:50 baseline because the a priori likelihood of guilt is usually much higher. To address some of these concerns, Honts and colleagues (1995) proposed the Directed Lie Test (DLT) as an alternative. Subjects are supposed to answer no in response to questions like, "Have you ever told a lie?" That is, they are instructed to lie. They are told the purpose is for the examiner to get an idea of what their physiological response looks like when they lie. The advantage of the DLT, compared to the CQT, is that subjects are not put in a place where they could be embarrassed (causing physiological arousal). Also, as directed lie questions are the same for all participants, the procedure is more standardized than the CQT. The DLT is easier to administer (Gallai, 1999), and Honts and colleagues (1995) found the DLT's accuracy to be at least as good as indices found for the CQT, if not better.

346

GRIESEL AND YUILLE

Psychometric Qualities of the GKT Because of its multiple-choice format, the GKT has the advantage that expected error rates can be mathematically calculated and then compared with the actual outcomes (MacLaren, 2001). Sensitivity and specificity are particularly relevant to evaluation of the GKT. To our knowledge, there are only two published field studies on the GKT (BenShakhar & Furedy, 1990; Elaad, 1990). In one, the specificity was 98% when only conclusive cases were included. For guilty subjects, fewer than 50% were correct outcomes (sensitivity). For laboratory studies, an overall sensitivity of 76% has been reported. These findings exceed the 20% chance level (based on a question with five options) of correct identifications for the GKT. MacLaren (2001) conducted a meta-analysis using binominal effect size displays. The sensitivity of the GKT was estimated to be between 76% and 83%, a rate that promises usefulness of the GKT under real-life conditions. Elaad (1998) summarized findings from 15 mock crime studies involving the GKT. The range of correct detections was between 79% and 100% for innocent participants and between 61 o/o and 100% for guilty subjects. Using multiple measures of physiological arousal and increasing the number of GKT questions further enhanced the rate of correct classifications. Elaad concluded that the GKT is able to protect innocent people from false accusations. A new development is the use of the P300 in event-related brain potentials (ERPs). ERPs are measured via scalp electrodes in the electroencephalogram to detect changes in electrical patterns across the cerebral cortex related to the presentation of a stimulus (Seymour, Seifter, Shafto, & Mosmann, 2000). The P300 is an index of recognition; that is, a positive wave typically occurrs 300 to 1,000 milliseconds after the onset of a meaningful and recognized stimulus (Rosenfeld, 2002). A pattern of electrical activity related to the recognition of a familiar but infrequent stimulus can be reliably detected; the amplitude of the P300 is proportional to the rarity of the stimulus (Rosenfeld, 2002). Examinees are asked to classify a series of stimuli into two categories, one of which occurs less frequently than the other (oddball paradigm; see Farwell & Donchin, 1991). The P300 occurs after an item of the lower-frequency category is presented, reflecting surprise and interest. The design used in these tests is similar to that of the GKT. First, the examinee is introduced to a series of stimuli representing details of a crime (targets). Next, the examinee is tested on the targets as well as other, yet unknown stimuli (irrelevant). Among the new stimuli are also relevant details of the crime under investigation (probes). All examinees will produce a P300 in response to the known targets from the list of details learned before the examination. Innocent subjects will not produce a P300 in response to the irrelevant stimuli and the probes. Guilty examinees will not produce a P300 upon presentation of the irrelevant stimuli either; however, the probes will elicit a P300, as participants with guilty knowledge recognize the details of the crime. Farwell and Donchin's study (1991) confirmed that those who were familiar with a crime scenario could be identified via the P300 in both laboratory and field settings (cf. Rosenfeld, 2002). The test produced neither false positive nor false negative errors; that is, patterns were obtained from all subjects as predicted. For 12.5% of the cases there was inade-

12.

CREDIBILITY ASSESSMENT

347

quate information. Thus, they were classified as indeterminate (Farwell & Donchin, 1991). Applying the test to a different area, Rosenfeld (2002) reported malingerers feigning cognitive deficits after suffering head injury could be identified via the described procedure.' The advantage of this test is that no physiological measures are used to infer lying. The P300 is a direct reflection of a cognitive process. Therefore, countermeasures are useless in this type of test. Furthermore, as in the conventional GKT, crime-relevant details are used to prevent false positive errors, as they are indistinguishable from irrelevant stimuli for innocent examinees (Farwell & Donchin, 1991). The disadvantage of the ERP test is the same as that of the guilty knowledge test, that is, often crime details leak out to the public, for example, via media reports.

Admissibility in Court In North America and Israel, the polygraph is widely used in criminal investigations; however, almost universally, it is not permitted as evidence in federal courts (BenShakkar et al., 2002; Saxe & Ben-Shakkar, 1999). In Germany, the polygraph was recently precluded as evidence from any legal procedures (Fiedler et al., 2002). In the United States, evidence is generally admissible in federal courts if it meets the Daubert criteria (Daubert v. Merrell Dow Pharmaceuticals Inc., 1993; see Gallai, 1999): (1) testability, (2) known error rates, (3) peer review and publication, and (4) general acceptance. The CQT is primarily lacking the last criterion, as it has been subject to intense controversies and fundamental disagreement. Regarding the general acceptance of polygraph tests, Iacono and Lykken (1997) conducted surveys with members of the Society of Psychophysiological Research and the Division 1 of the American Psychological Association (APA). Results indicated that two-thirds of the respondents did not find the CQT to be theoretically well grounded. Furthermore, claims of its validity were not confirmed; for example, it can be beaten by use of countermeasures. Given Fiedler and colleagues' (2002) criticism, one could further conclude the CQT also lacks the first Daubert criterion, testability, because ground truth and a priori base rates are nearly impossible to establish in field studies: "[I] tis very hard, if not impossible, to test the science upon which lie-detection rests" (Gallai, 1999, p. 96). The third Daubert criterion, the number of publications on the polygraph, seems to be fulfilled. However, the second criterion, known error rates, is clearly lacking (Gallai, 1999). In contrast to the CQT, the GKT does meet the Daubert criteria for admissibility in court (Ben-Shakkar et al., 2002). The survey oflacono and Lykken (1997) revealed that the majority of respondents believed in the GKT's sound scientific principles. If one were to use it more often, administrative and procedural changes would be essential to meet the strict preconditions necessary for GKT applications. In Japan, such changes have been made, and the GKT is used as a standard investigative tool (Ben-Shakkar et al., 2002). 1 Laboratory studies indicated that the P300 might also be useful in the distinction of genuine and false memories (cf. Rosenfeld, 2002).

348

GRIESEL AND YUILLE

NONVERBAL CREDIBILITY ASSESSMENT

An important aspect of self-presentation is regulation of nonverbal behaviors (DePaulo, 1992). These aspects of behavior can be clues to deception, although the reader should bear in mind that there are no behavioral indicators oflying (Ekman, 2001). Alternative explanations for the behavior should always be considered. Behavior-based lie detection consists of two steps: first, recognizing a change in behavior and, second, interpreting the reason for the change accurately (Frank & Ekman, 1997). The focus of the following material is on changes in the expression of emotion. This aspect of lie detection has received a great deal of research attention (see Ekman, 1992). Much of this work is based on the Facial Action Coding System developed by Ekman to reliably code the expression of emotions.

Facial Action Coding System DePaulo and colleagues' (2003) recent meta-analysis of deception cues included emotion-related facial clues. Significant effect sizes concerning these clues were d = 1.121 for facial pleasantness, d = 1.161 for pressing lips, and d = 1.251 for chin raises. 2 Thus, facial expressions are valuable cues in the detection of lies. In comparison with other methods, Porter, Campbell, Stapelton, and Birt (2002) found that the reliance on facial cues led to higher detection accuracy in both truthful and deceptive statements than the reliance on verbal cues only. In accordance with this finding, Etcoff, Ekman, Magee, and Frank (2000) demonstrated that aphasics (i.e., people who cannot understand spoken sentences and are bound to rely on facial cues more so than normal people) were more accurate at detecting concealed emotions in videotaped statements than their controls. Ekman, Sorenson, and Friesen (1969) showed people from New Guinea, Borneo, the United States, Japan, and Brazil photographs of facial expressions. Observers from all cultures detected the same emotions when looking at the pictures. These findings contradicted the dominant theory at the time that emotional expressions were socially learned and thus culturally variable. The new findings, in contrast, favored the Darwinian suggestion that facial expressions were evolutionarily shaped and thus universal for all human beings, regardless of their cultural background. A method was developed to measure facial expressions of emotions, the Facial Action Coding System (FACS; Ekman & Friesen, 1978). FACS is the only comprehensive, anatomically based, and objective system to measure visible facial movements (e.g., Rosenberg, 1997). FACS is purely descriptive, which means that no emotional labels are used to code facial expressions. It is based on 44 action units (AUs), as well as positions and movements of the head and the eyes. AUs are responsible for any momentary alterations in the face's appearance, allow-

2

d was defined as the difference between the deceptive and the truthful condition divided by the mean of the standard deviation for truths and lies.

12.

CREDIBILITY ASSESSMENT

349

ing for registration of all facial movements. The emotional dictionary (Friesen & Ekman, 1989; as cited in Ekman, 1997a) assists in interpreting the emotional value offacial expressions. It informs about essential and irrelevant AUs for each emotional category. Emotions are often expressed in only minimal changes of the face. Only FACS has the ability to detect such minor alterations of facial displays (Cohn, Zlochower, Lien, & Kanade, 1999). FACS has a crucial advantage over obtrusive techniques such as electromyography (EMG) because individuals are not necessarily aware of the fact that their face is being assessed (Rosenberg, 1997). Given that FACS has been used since the late 1970s, there is surprisingly little research on its psychometric qualities. Certification as a FACS coder requires a proficiency test; however, proficiency does not inform about the coding of individual AUs. Four different types of FACS interrater reliability were distinguished by Sayette, Cohn, Wertz, Perrott, and Parrott (2001): (a) with regard to individual AUs, (b) concerning the temporal resolution of coding (i.e., can coders detect facial changes on certain time bases?), (c) in terms of the intensity of facial movements, and (d) about emotion-specific expressions. All four issues were addressed in several studies (Sayette et al., 2001), with the use of odor smelling, smoking craving, and speeches about the participant's physical appearance to elicit facial expressions. The results revealed that within certain temporal tolerance windows, all but two AUs had excellent reliability (types a and b). The greater the time precision, the weaker the reliability of the coders. Better kappa results were obtained on a 3-point scale as opposed to a 5-point scale of AU intensity (interrater reliability type c). Finally, good to excellent reliabilities were found for positive as well as negative AU combinations (type d). From these observations, it was concluded that FACS is reliable. Does FACS reveal the emotion actually felt by an individual? The answer lies in a comparison of self-reported emotions and/or physiological indicators of emotions with facial expressions. If the face is a valid indicator of emotions, then momentary congruence between facial expressions and the underlying emotions could be established (Rosenberg & Ekman, 1997 /1994). In the exploration of this question, methodological problems include consciously filtered self-reports, memory reconstruction, and a lack of unambiguous relationships between physical states and emotional nuances. Facing at least some of these challenges, Rosenberg and Ekman ( 1997/1994) had participants watch videos to elicit different emotions while their faces were filmed. In the scope of a cued-review procedure, participants watched the videos a second time and stopped the tape when they remembered having experienced an emotion during their initial viewing of the film. The results revealed that facial expressions were indeed linked to self-reported emotions, not only with regard to time but also the type of emotion. Thus, according to this study, emotions are unified responses corresponding to certain facial expressions, and they can be reliably assessed with the use ofFACS. Another aspect ofFACS is its ability to distinguish deliberate and spontaneous emotional expressions. Rosenberg (1997) noted that most of the research on FACS was based on posed expressions. In order to use FACS as a means to detect deception, it is important to know that true and fake expressions can be distinguished morphologically.

GRIESEL AND YUILLE

350

Hess and Kleck (1990/1997) found that posed expressions were longer and contained more irregularities like pauses or intensity changes than spontaneous expressions. There was no difference in the time of onset and offset. 3 Ekman and Friesen (1981; as cited in Gosslin, Kirouac, & Dore, 1995/1997) further demonstrated that the face displays genuine emotions more symmetrically than simulated emotions (see also DePaulo, 1992). Gosslin and colleagues (1995/1997) used professional actors to explore differences between felt and unfelt emotions. Similarity was only found between AUs on the actors' faces and the facial expression of genuine emotions, particularly for surprise and happiness. Single AUs could be more easily enacted than AU combinations, a possible requirement of genuine emotions. In summary, posed and possibly deceptive emotional expressions differ from spontaneous expressions in terms of morphology, symmetry, total duration, speed of onset, coordination in apexes, and ballistic trajectory (smooth vs. sharp appearance; Ekman, 1997a). FACS as a Lie Detection Tool

FACS seems useful in the context of deception detection because fabricated expressions are distinguishable from genuine emotional expressions. There are certain facial muscular movements that most people cannot enact voluntarily but that are necessary for the expression of genuine emotions (Ekman, 1993). Ekman termed these reliable muscles. They might serve as a distinction between false and genuine emotional expressions (Ekman, 1985). Smiles are highly influenced by social contexts and have multiple meanings. For example, they can stand for genuine happiness, yet also fear or dislike (DePaulo, 1992). Ekman, Friesen, and O'Sullivan (1988/1997) described a study undertaken with nursing students to distinguish happy, genuine smiles from fake smiles masking other emotions. During the experiment some students watched a nature film made to elicit pleasant feelings, while others watched an amputation and burn film made to elicit unpleasant feelings. Both groups were interviewed after they watched the films and instructed to conceal any negative feelings they may have had. FACS measurements were taken from the nurses' facial expressions during the interviews. Both happy and masking smiles were defined as an action of the zygomatic major muscle, although other muscles were involved for each emotion. The FACS raters were blind to the experimental conditions and reached an interscorer reliability of .77. There were more hits (10) than misses (1); however, the majority of subjects could not be classified (20). The results revealed that felt happy expressions occurred more often in the honest interview condition (i.e., for those nurse students who had seen nature films). Masking smiles occurred more frequently during deceptive interviews (i.e., when the students were instructed to lie about the unpleasant film). It was concluded that the face reveals truly felt emotions despite concealment efforts. As long as emotions were involved, facial expressions provided cues to deceit. 3 It was suggested that different neurological pathways accounted for the two types of emotional expressions and provide an explanation for their phenomenological differences. Spontaneous expressions are thought to function in a reflex-like manner (Hess & Kleck, 1990/1997).

12. CREDIBILITY ASSESSMENT

351

Other Behavioral Deception Cues When people lie, they blink more (the so-called Nixon effect; Tecce, 1992; as cited in Mann, Vrij, & Bull, 2002), hold less eye contact, show fewer illustrators (defined as movements to modify or supplement speech), and show more adaptors (i.e., movements in which one's hand makes contact with another body part, such as rubbing; Ekman, Friesen, & Scherer, 1978; DePaulo et al., 2003). It was suggested that the reason for increased body movements is that people tend to become more aroused when they lie, compared with situations in which they tell the truth. Arousal can be established in behaviors such as the ones listed above. Zuckerman, DePaulo, and Rosenthal (1981, as cited in DePaulo et al., 2003) noted that the occurrence of behavioral indicators of fear and guilt (such as fidgeting or sounding unpleasant) are more often experienced by people when they lie than when they tell the truth. However, under a high motivation to succeed in lies, some of these behaviors increase, whereas others decrease or disappear because people attempt to control them (DePaulo, 1992; DePaulo et al., 2003), an effect called motivational impairment (DePaulo & Kirkendol, 1989). Also, when liars have to think hard to be convincing, they usually show fewer illustrators and self-manipulators, such as scratching the head (Mann et al., 2002). Most people try to suppress their nervous behaviors, resulting in overcontrolled movements. A problem in this area of research is the low number of field studies on real suspects of crime. Mann and colleagues (2002) conducted a field study involving 16 crime suspects interviewed by the police. The police files provided reliable independent information for corroboration of truths and lies. The results showed that lying was significantly associated with a decrease in blinking and an increase in speech pauses. In contrast, gaze aversion, head movements, arm and hand movements, and other speech disturbances were not reliably associated with lying. Thus, the attempt to control one's behavior seemed to outweigh nervousness. Large individual differences were found in this study. Similarly, Vrij, Akehurst, and Morris (1997; as cited in Vrij &Graham, 1997) found that the occurrence of hand movements during lies was moderated by high self-consciousness, as measured by a personality inventory resulting in fewer hand movements during lies. Thus, knowledge of the liar's personality is necessary for accurate interpretation of behavioral lying indicators. Without such knowledge, deviations from normal behavior are the most individually reliable behavioral indicators of lying. That is, different people show different patterns of deceptive behavior. Also, deceptive behaviors may differ from one situation to another (cf. Vrij & Graham, 1997). Furthermore, behaviors that are indicative of lying can also be indicative of other states. Again, it should be noted that behavioral indicators are only connected to lying in probabilistic ways, that is, they could reflect psychological states other than lying, a problem of discriminant validity. There is no one-to-one connection between specific nonverbal behaviors and internal states (DePaulo, 1992, 1994). Even when relationships are found between specific body movements and certain internal states, the association is usually not strong-a limitation to keep in mind when behavioral information is incorporated to detect lies. Besides body movements, speech characteristics can also reflect people's emotions. Lying was associated with speaking in a higher pitch (DePaulo, 1994; Ekman et al.,

352

GRIESEL AND YUILLE

1978). Also, a liar's speech was more hesitant and less fluent compared to truth tellers (DePaulo, 1992), especially when people were highly motivated to be successful and their behavior became overcontrolled, that is, they gave shorter responses and spoke more slowly and in higher pitch. Overall, few behavioral cues discriminated between truths and lies in DePaulo and colleagues' (2003) meta-analysis of deception cues (e.g., d = 1.141 for the use of illustrators, d = 1.211 for voice pitch). No significant differences were found for body activity, posture shifts, or head, leg, foot, arm, and hand movements. In short, not behaviors but behavioral changes, which might differ from one individual to the other and must be interpreted in context, are clues to deception. Behaviors per se cannot be used as reliable indicators of lying.

VERBAL APPROACH TO CREDIBILITY ASSESSMENT

Compared with nonverbal behaviors and physiological correlates of lying, the verbal content of lies is the easiest to control for a liar. As with nonverbal behavior, there are no verbal expressions reliably indicative of lying. Verbal deception cues arise from the same reasons as nonverbal lying indicators: emotion, content complexity, and attempted control. These factors lead people to provide overgeneralized or indirect answers (Vrij, 2000). A lie's content complexity causes people to provide implausible or overly short statements. In their literature review, Vrij, Edward, Roberts, and Bull (2000) showed a consistent association of lying and certain verbal behaviors, such as latency period, speech hesitations, quantity of details, reproduction of conversation, and auditory, visual, spatial, and temporal details. The following section is mainly focused on Statement Validity Analysis, the most popular technique in the field of credibility assessment based on verbal cues. An appropriate interview is very important, as it forms the basis of subsequent analyses of verbal contents. Reality Monitoring, which has only recently become the focus of attention, is briefly introduced as an alternative approach.

Statement Validity Analysis Statement Validity Analysis (SVA) was developed in the context of child abuse allegations. Unfortunately, such allegations are often made considerably after the event took place, and thus physical evidence is often scant. Most abuse cases happen in private settings; hence there are rarely any witnesses aside from the victim and the perpetrator. The child victim's statement is often the only evidence (cf. Steller & Koehnken, 1989; Yuille, 1988b). Although false allegations do occur, valid allegations are the majority (Yuille, Tymofievich, & Marxsen, 1995). Nevertheless, assessing the credibility of children's allegations is necessary. Several avenues have been taken in an attempt to assess children's allegations of sexual abuse. One of them is using behavioral checklists, incorporating items such as enuresis, depression, recantations, or sexually inappropriate behavior as signs of abuse

12.

CREDIBILITY ASSESSMENT

353

(Yuille et al., 1995). Unfortunately, these signs are not exclusively caused by abuse. Also, victims of abuse may not show any of these signs. SVA is an alternative and more promising approach. SVA began with the work of Undeutsch in Germany in the 1950s (for a history see Undeutsch, 1989; Steller & Koehnken, 1989). SVA is the label for the overall diagnostic procedure, composed of three components: (1) a properly conducted interview (e.g., the Cognitive Interview, Fisher & Geiselman, 1992; the Step Wise Interview Protoco!, Yuille, Marxsen, & Cooper, 1999), (2) Criteria Based Content Analysis (CBCA), and (3) the Validity Checklist to analyze the CBCA outcome (e.g., Lamb, Sternberg, Esplin, Hershkowitz, & Orbach, 1997; Vrij, 2000). The procedure is based on the Undeutsch hypothesis: Statements based on genuine experiences differ in quality and content from invented or coached statements (Steller, 1989; Vrij, 2000). Underlying this hypothesis are two assumptions, one having to do with the person's motivation, the other with the person's cognition (Koehnken, Schimossek, Aschermann, & Hoefer, 1995; Ruby & Brigham, 1997). A deceptive person's motivation to appear honest will cause the statement to be expressed differently than if the story were actually true (e.g., avoiding stereotypical lying behaviors). Furthermore, everyone's cognitive and verbal abilities have limits when it comes to fabricating and maintaining a lie. In general, it is assumed that someone telling a true story will be able to provide more details, more contextual information, and more subjective experiences.

Criterion-Based Content Analysis CBCA is the qualitative analysis of a statement's contents. Steller (1989) integrated the different criteria used up to that point in the context of child abuse investigations and established 19 criteria contributing to a statement's credibility. It is important that statements, not individuals per se, are analyzed with regard to their credibility. Under certain circumstances, a generally sincere person may have reasons to lie. Alternatively, a rather deceptive person may speak the truth on certain occasions (Steller & Koehnken, 1989). The hypothesis is that criteria are more likely to occur in genuine as opposed to fabricated statements, for example, because the statement provider may not be creative enough or lack the knowledge to make up additional details. SVA is not a verbal lie detector. Rather, the underlying strategy is a search for signs of truthfulness (Koehnken et al., 1995). The presence of each criterion adds to the statement's credibility. Yet their absence does not automatically imply deceptiveness (Lamb, Sternberg, Esplin, Hershkowitz, & Orbach, 1997; Steller, 1989; Yuille, 1989). CBCA ratings are undertaken as follows: 0 if a criterion is absent in a given statement, 1 if present, and 2 if strongly present (Honts, 1994). Despite the scoring, CBCA is not a standardized test with set norms suggesting certain interpretations. There are no commonly agreed-upon rules for deciding what is credible and what is not (e.g., Tye, Amato, Honts, Devitt, & Peters, 1999). Rather, CBCA is a complex qualitative procedure. Credible memory accounts are not necessarily based on actual events. Accounts containing few CBCA criteria are not precluded from being derived from a genuine experience. A composite or total score would be hard to interpret, as certain criteria should be given more weight than others; for example, logical structure and a certain

GRIESEL AND YUILLE

354

amount of details are, according to some practitioners, mandatory for credible accounts, whereas unusual details are optional (cf. Steller & Koehnken, 1989). The same problem would emerge if one were to establish a required minimum total score for credible statements. Especially in cases involving children, even a very low CBCA total score could be meaningless, depending on the child's age, personality, and verbal and intellectual abilities. Thus, CBCA should be viewed as a semistandardized approach (Steller & Koehnken, 1989).

Validity Checklist The Validity Checklist is the third component of SVA and consists of two sets of items referring to the witness's account characteristics as well as the interview and other evidence (for a detailed list see Yuille, 1989). The statement-related factors include psychological characteristics (e.g., appropriateness of language, appropriateness of knowledge, susceptibility to suggestions), interview characteristics (e.g., adequacy of the interview), and motivational characteristics (e.g., pressure to report). Each one of the items, if answered in the affirmative, makes the CBCA outcome more likely to be invalid (Vrij, 2000). The Validity Checklist is used to consider alternative interpretations to prevent premature conclusions in favor of the statement's credibility (Honts, 1994). Yet there has not been any systematic empirical research on the Validity Checklist. Lamb, Sternberg, Esplin, Hershkowitz, and Orbach (1997) have expressed concern about the interrater agreement because of inadequate specification of the criteria.

Research on SVA In spite of SVN.s frequent use in German courts, little research had been conducted on its reliability and validity by the late 1980s (Steller, 1989). However, during the past 15 years, SVA has been the focus of considerable research. Again, there are two common research paradigms: laboratory and field research (cf. Ruby & Brigham, 1997). In the majority of laboratory studies, the participants are asked to provide their statement immediately after witnessing a staged event, watching a film, or viewing slides. In contrast, and this is especially true for allegations of child sexual abuse, real-life witnesses usually report crimes often after a considerable delay. Thus, different storage and retrieval processes are likely to have taken place for accounts elicited in the laboratory and the field, respectively.

Reliability A scale's validity can never exceed its reliability. Although CBCA is not considered a scale, the same reasoning is applied with regard to its psychometric qualities. Traditionally, there are three aspects of reliability: test-retest reliability, item consistency, and interrater reliability. Test-retest reliability is of a lesser concern with CBCA and has rarely been looked at. One study conducted by Horowitz and colleagues (1997) found

12. CREDIBILITY ASSESSMENT

355

that three different raters were highly reliable over a period of three months. That is, they were individually consistent in their scoring. Test~retest correlations ranged be~ tween .85 and .91. Item consistency is not a major concern with CBCA either. The cri~ teria were not collected to be internally consistent. Anson, Golding, and Gully (1993; as cited in Ruby & Brigham, 1997) calculated an index of internal consistency for the 19 CBCA criteria that had been coded from videotapes by trained raters. The intercorrela~ tion turned out to be .15, a finding that suggests that CBCA measures different aspects of truthfulness (i.e., is not internally consistent). Most studies have focused on CBCA's interrater reliability (Honts, 1994). Steller (1989) found that CBCA evaluators achieved a high degree of agreement with relatively little training. Some of the interrater agree~ ment scores reached 100%. In Yuille (1989), the agreement offalse and true classifica~ tions was over 90% between two independent raters. Koehken and Wegener (1982; as cited in Horowitz, 1991) found a correlation of .95 for interrater agreement. Most relia~ bility studies have applied inadequate or inappropriate measures of interrater reliabili~ ties, namely percentages of agreement or correlation coefficients that do not take chance agreement into account (cf. Horowitz, 1991). A better measure for interrater re~ liability is Cohen's kappa (Cohen, 1960). Horowitz and colleagues (1997) analyzed indi~ vidual CBCA criteria with the use of Cohen's kappa and Maxwell random error coeffi~ dents (the latter was to account for the low incidence of some criteria). Accounts of 100 abused children were rated for the 19 CBCA criteria by three judges. The findings suggested the exclusion of five criteria because of insufficient reliability: superfluous details, admitting lack of memory, spontaneous corrections, related external associations, and

unusual detail.

CBCA Studies Involving Child Witnesses SVA was developed in the context of child abuse allegations. Given the high preva~ lence of child abuse and the growth of reported cases in past decades, whether legitimate or false, credibility assessments of child witnesses have become crucial. Experiences like the Scott County case (see Hubert, 1985; Raskin & Yuille, 1989) and the Country Walk case (see Ceci & Bruck, 1993; Cockburn, 1993; Goodman, Aman, & Hirschman, 1987) have taught us how dangerous it is to assume that children cannot make up com~ plex stories or that they do not have the sexual knowledge to invent allegations of sex~ ual abuse. Yuille (1988a) had elementary school children tell a true and a false story. Blind assessors rated their statements according to CBCA. Supporting the Undeutsch hypoth~ esis, over 90% of the true stories and almost 75% of the false stories could be correctly classified based on the presence of the three criteria (spontaneity, appropriate detail, and admission of lack of memory) plus any two other CBCA criteria. The procedure pro~ duced more false positive than false negative errors (a truth bias 4). Steller (1989) reported on a laboratory study with characteristics that were similar to those of sexual abuse cases: specifically, the children's involvement was direct and As stated above, SVA is a search for signs of truthfulness.

4

356

GRIESEL AND YUILLE

personal, the event was of a negative emotional nature, and there was a loss of control. Approximately 100 children were asked to report an event in which they had been bitten by a dog or undergone medical treatment. Interviewing the parents provided an indication of the statements' truthfulness. The children were further asked to tell a fictitious story. Trained graduate students rated the children's accounts for credibility. The Undeutsch hypothesis was confirmed for 11 of 17 criteria, although this was only true for stories referring to medical treatments and tests. It was shown that judgments made by trained students were correct significantly more often than those made by students who had only heard of CBCA but had not been trained. Again, the rate of correct classifications was higher for true stories than for fictitious stories; thus there was a bias toward falsely classifying fictitious stories as true (false positive error). Esplin, Boychuk, and Raskin (1988; as cited in Steller, 1989) conducted the first field study on CBCA validity that applied more than one technique to confirm alleged sexual abuse (confession, medical evidence, polygraph of the perpetrator). The study was based on 20 confirmed and 20 unconfirmed cases of sexual abuse. A comparison of the mean CBCA scores strongly differentiated the two groups: 15 of 19 CBCA criteria were present significantly more often in confirmed cases, and 7 criteria were present in each confirmed case (as reported in Ruby & Brigham, 1997). CBCA was found to be 100% accurate in classifying doubtful and truthful statements (Raskin, Esplin, & Horowitz, 1991; as cited in Tye et al., 1999). A similar finding was made in a field study by Boychuk (1991; as cited in Tye et al., 1999). Another extensive field study was conducted by Lamb et al. (1997). They had access to almost 1,200 transcripts of investigative interviews conducted by trained youth investigators in Israel. Ninety-eight cases were selected according to certain criteria (known assailant, enough independent pieces of evidence, age range of child victims between 4 and 13, and physical contact during alleged abuse). With the use of medical examination evidence, confessions, polygraphic examinations, witness statements, and other evidence, a multifaceted procedure was used to determine the plausibility of the alleged abuse. With the use of only 14 CBCA criteria, a significant mean difference in total scores was found between transcripts of plausible and implausible cases. This was especially true for unstructured production, quantity of details, contextual embedding, interactions, and conversations. The authors concluded that, although CBCA was reliable and valid in terms of distinguishing children's truthful and false statements of sexual abuse, the accuracy of CBCA remained rather poor, as the differences found in this study were far less dramatic than in previously reported field studies. Based on the same transcripts, Hershkowitz (1999) asked whether open-ended interview questions elicit more criteria-rich responses from children than focused utterances. For 12 statements the events seemed likely to have happened; for another 12 the events seemed unlikely to have happened. It was found that CBCA scores were not different for plausible and implausible accounts. However, their strength of presencedefined as the number of times each criterion was present-differed for the two types of statements. Furthermore, open-ended questions led to significantly longer and more criteria-rich accounts than focused utterances. This was only true for plausible accounts, but not for implausible statements.

12.

CREDIBILITY ASSESSMENT

357

In a review of field and laboratory research conducted up to the early 1990s, Horowitz (1991) came to the conclusion the first three criteria (logical structure, amount of detail, and unstructured production) of CBCA are necessary but not sufficient to rate statements as truthful or false. Specific content criteria (e.g., contextual embedding, inter~ actions, complications) were found more often in truthful than in deceptive/fictitious statements, with the exception of accurately reporting details misunderstood, a criterion that relates to a child's age. No age norms considering sexual knowledge have been gath~ ered so far. The criterion attribution of the perpetrator's mental state contributed very little to the distinction of truthful from fictitious accounts. In regard to the motivation~ based CBCA criteria (e.g., admission of lack of memory), Horowitz reported very different results for laboratory~based studies and field research: none of them were helpful in laboratory studies, whereas they seemed to differentiate between unconfirmed and truthful accounts in the field. Vrij, Akehurst, Soukara, and Bull (2002) obtained somewhat scattered results with regard to the witnesses' age. Participants in the study were children of different age groups as well as undergraduate students. A staged event was wiping off a blackboard. Participants in the truthful condition had seen a confederate wiping the blackboard, and those in the untruthful condition were only told about the person wiping the black~ board. A significant age effect was found with young children obtaining lower CBCA and scores than older children and young adults. Only for two age groups (5-6~year~olds undergraduates), CBCA scores were significantly higher for truth tellers than for liars. Thus, the Undeutsch hypothesis was confirmed only for very young children and young adults, but not for children between the ages of 10 and 15.

CBCA Studies Involving Adult Witnesses Koehnken and colleagues' (1995) study on CBCA and the Cognitive Interview (reported above) involved adults. The participants were shown a film (truthful group), versus only being told about the content of the film (untruthful group) and subsequently interviewed about it. Interview transcripts were coded via CBCA. Truthful and fabri~ cated statements could be correctly classified in 93% of the cases. The criteria insecuri~ ties, admitting lack of memory, and reporting style had the best discriminant abilities. CBCA was valid with adult participants, although the generalization of these results is limited because of the use of film stimuli. Landry and Brigham (1992) applied CBCA in a sample of adults' statements re~ ferring to truly experienced versus invented traumatic events. For 10 of the 14 items, ef~ fects were found in the expected direction. CBCA~trained raters performed significantly more accurately and above chance (55%) compared to untrained raters (47%). Better~ than~chance results were obtained when CBCA was combined with watching a video~ taped statement (as opposed to just reading a transcript); thus, information other than purely verbal cues was used. As in the studies involving children, there was a truth bias. The findings suggest that CBCA was somewhat useful for the evaluation of adults' state~ ments, given that visual material was provided besides verbal information. Unfortunately,

GRIESEL AND YUILLE

358

it was not investigated how accurately statements could be evaluated based on visual information only.

CBCA Studies Involving Crime Suspects Motivated by the overreliance on nonverbal deception detection in the suspect context, Porter and Yuille (1996) conducted a simulation study analyzing verbal cues in suspects' statements. The participants committed a theft, under the pretext of testing the effectiveness of a newly introduced security system in the building. The evaluation included deception/credibility cues from several approaches, including SVA and Reality Monitoring (described in more detail in the next section). After witnessing the removal of a folder from an office, participants were assigned to one of four experimental groups: truthful alibi, partial deception, false alibi, and truthful confession. The Step-Wise Interview Protocol was used to interrogate the mock suspects. The results revealed that the accounts differed only in terms of three CBCA criteria: the amount of detail, coherence, and admitting lack of memory. A discriminant analysis correctly classified 78% of both truthful and deceptive accounts. Even though verbal cues were the most helpful of all cues included in the study, the support for SVA in the suspect context is limited, as the majority of CBCA items failed to discriminate between the types of accounts. DePaulo and colleagues' (2003) recent meta-analysis of studies comparing truths and lies told by crime suspects, witnesses, and mock witnesses revealed significant effects sizes ranging from d = 1.141 to d = 1.551 for deception cues that were similar to those of SVA: details, talking time, plausibility, logical structure, discrepancy/ambivalence, verbal and vocal involvement, verbal and vocal immediacy, verbal and vocal uncertainty, word and phrase repetitions, spontaneous corrections, admitted lack of memory, and related external associations. Thus, several SVA cues have proved their validity in the distinction of lies and truths in research conducted over the past 35 years.

Summary and Admissibility in Court German courts have required psychological interviews outside the courtroom in cases of child sexual abuse for 50 years. Expert witnesses frequently give their opinion about whether they find a child is telling the truth or not (Honts, 1994). SVA is also accepted in the Netherlands, where the Supreme Court has ruled that a victim's allegation and an expert's opinion in favor of its credibility are sufficient evidence for civil proceedings (Rassin, 1999). In North America, SVA has been used in the investigation context (Vrij, 2000). Honts (1994) strongly argued for the use of SVA as forensic evidence in North American courts. Generally, Honts saw the four Daubert criteria (Daubert v. Merrell Dow Pharmaceuticals, 1993), the set of guidelines for admitting expert evidence in U.S. federal courts, as being met by SVA: (1) It is testable, has been tested; (2) the error rate is known; (3) the procedure has been subject to peer review, and-if a lack of criticism equals acceptance-it is (4) also generally accepted in the scientific community (Honts, 1994). Contradicting Honts' position, Lamb et al. (1997) argued, although reliable and

12. CREDIBILITY ASSESSMENT

359

valid, CBCA is not precise enough to be permitted in the courtroom. Tully (1998) saw SVA as a "protocol through what it transparently reveals" (p. 188). Rather than being a psychometric test, CBCA should be viewed as a protocol (Rassin, 1999). Several issues remain unresolved and should be taken into account when SVA is submitted in court.

Unresolved Issues of CBCA Rassin (1999) pointed out a number of problems that frequently arise in the context of CBCA. For example, there are concerns that witnesses, especially children, could be coached in SVA Similar to the countermeasure problem in polygraph testing, this knowledge could be used to beat the procedures and lead the expert to make a false positive judgment. The study by Vrij, Kellner, and Mann (2000; as cited in Vrij et al., 2002) confirmed this concern because (adult) liars who were knowledgeable about CBCA scored significantly higher than uninformed participants. In fact, the informed liars' scores could not be statistically differentiated from scores obtained from truth tellers. However, Vrij and colleagues (2002) found only a marginally significant coaching effect for children above the age of 10 as well as adults. For younger children (age 5 and 6), the coaching did not increase CBCA scores at all, indicating that young children did not comprehend or could not apply what they had been taught. Lamb, Sternberg, Esplin, Hershkowitz, and Orbach (1997) noted that CBCA is especially weak if false elements are embedded in truly experienced events. Overall, there is reason to be concerned about the discriminative power of CBCA if liars are informed about the CBCA criteria. Rassin's (1999) most interesting point is his critique of CBCA from a scientifictheoretical point of view. He argued that CBCA uses an inductivistic approach, searching for signs of truthfulness. According to Popper, who emphasized the importance of falsification for theory testing, disconfirming evidence is more valuable than confirming evidence. Deceptive statements can be construed in ways that make them appear truthful. Thus, one single sign of deceit would outweigh several signs of truthfulness. Historically, CBCA was intended to prove that children can indeed be credible witnesses; thus the focus was on the search for truthfulness. "CBCA is a truth-verifying rather than a lie-detection tool" (Vrij, 2000, p. 149). CBCA is problematic if the purpose is to prove someone was lying. This also explains the high rate of truth biases in many applications of CBCA. This adds to the point that SVA is meant as a guideline, not as a diagnostic measure. Rassin (1999) further complained about the absence of decision rules, that is, a lack of clear or operational definitions of what is required for a criterion to be met and the number of criteria necessary to support a statement's credibility. Yuille (1990) described a decision rule, namely the presence of criteria 1 to 5 plus any two other criteria. Similarly, Raskin and Steller (as cited in Yuille, 1990) suggested the presence of criteria 1 to 3 plus any four other criteria. These decision rules originated from the authors' experience with CBCA. Zaparniuk, Yuille, and Taylor (1995) found that a single criterion (spontaneous corrections) had better discriminative power than either one of the decision rules tested. In practice as well as from an empirical point of view, there is a great variation of the number and combinations of criteria used to judge witness credibility. When

360

GRIESEL AND YUILLE

it comes to the single criteria, Ruby and Brigham (1997) found that amount of details was the best-supported CBCA criterion. Another weakness of CBCA could be its inability to distinguish truthfulness and accuracy; that is, unintentional lies (e.g., pseudomemories) and truthful accounts. Unintentionally reported errors cannot be distinguished from true memories, because an unintentionally wrong statement would not have the cognitive and motivational characteristics of a deceptive statement. Huffman and Ceci (1997; as cited in Ruby & Brigham, 1997) used such unintentionally wrong memories. However, unexpectedly, with the help of four CBCA criteria, it was possible to distinguish true accounts from pseudomemories. The major shortcoming of SVA is its use by so-called experts who lack sufficient training. Steller (1989), as well of Landry and Brigham (1992), showed that short training sessions of 45 to 90 minutes are sufficient to improve SVA accuracy. For instance, 45-minute-long training sessions in the study of Landry and Brigham resulted in 75% correct identification of truthful targets, whereas untrained subjects succeeded in only 59% of the cases. However, they also recommend more in-depth training in order to obtain even better results. A challenge in child sexual abuse allegations is the fact that inaccuracy5 in a child's account does not necessarily mean the statement is false. The second author's experience with SVA shows that two to three day-long workshops are clearly insufficient to become an adequately trained and reliable SVA assessor. Prolonged practice including the description of memory processes, children's linguistic capacities, and a discussion of suggestibility are necessary to meet minimum standards (cf. Koehnken, 1999; as cited in Rassin, 1999; Lamb, Sternberg, Esplin, Hershkowitz, & Orbach, 1997). Training needs to incorporate exercises to give the interviewers an opportunity to become experienced with the techniques (cf. Bruck, Ceci, & Hembrooke, 1998). This point does not speak to the usefulness of CBCA per se, but rather to its improper use. It would certainly help if courts had a way of distinguishing between those who have received CBCA training and those who have not (Rassin, 1999). It is regrettable that no standard of practice has been developed related to the use of SVA. Other weaknesses ofSVA have been pointed out, as Horowitz and colleagues (1995) mentioned. CBCA itself does not have scoring guidelines or norms in regard to a child's age, although such considerations are incorporated into the interview preceding CBCA. Ruby and Brigham (1997) commented on cultural differences in the use of CBCA; for example, in some studies in the United States certain criteria were seen significantly more often in whites' than in blacks' true statements. In 1989, Steller and Koehnken concluded that, until further research on the reliability and validity of SVA becomes available, it should be used with great caution. Ten years later, Rassin (1999) still supported this viewpoint. It should be noted that most studies focused exclusively on CBCA without considering the other components of SVA, providing hope that future studies will lead to more stable outcomes. 5

Inaccuracy, in this case, would result from a child's restricted cognitive and linguistic abilities, not the scoring procedure.

12. CREDIBILITY ASSESSMENT

361

REALITY MONITORING

SVA is the most widely used and known technique of verbal credibility assessment (Vrij, Akehurst, Soukara, & Bull, 2004). Reality Monitoring (RM) is an alternative ap~ proach that researchers have recently investigated (Vrij, 2000). A critique of CBCA has been its atheoretical nature; that is, the criteria were selected based on their utility in distinguishing true and false statements. No reasons were provided for why the CBCA criteria have the ability to detect differences between truly experienced and fab~ ricated statements. RM is the mental process by which people discriminate whether their memory for an event is real or imagined. The approach was originally developed to distinguish true and confabulated memories (Johnson, 1988). The idea can be applied to deception de~ tection, because the information people use to determine the veracity of their own memories can also be used to evaluate other people's accounts (Johnson, 1988). John~ son and Raye (1998) suggested that memories based on truly experienced events differ from imagination insofar as the former contain more perceptual, contextual, and affec~ tive information. For example, memories for external (i.e., perceived) events contain sensory details such as smells and colors or information on spatial arrangement. How~ ever, fabricated memories are derived from internal sources and thus contain more cog~ nitive operations (e.g., intention, planning, drawing conclusions) than memories for external events. RM is an internal process of remembering, whereas deception is an in~ terpersonal process of relating. Truthful accounts should be more clear, realistic, and '-:ivid, whereas deceptive accounts should contain more cognitive processes and infer~ ences (DePaulo et al., 2003; Vrij, 2000). Johnson's work is primarily concerned with the way people remember an event, not how they describe it (Vrij, 2000). However, RM provides a theoretical and extensive empirical basis that might be useful in deception detection. It seems intuitively correct that people-when lying-attempt to make their stories more consistent, using cognitive operations, that is, operations related to peo~ ple's own thinking, as opposed to recalling factual information. Sporer (1997) published the first formalized list of criteria based on RM. experiment to compare CBCA and Sporer (1997) conducted a laboratory~based RM. Two blind raters evaluated transcripts of students' truthful and fabricated state~ ments. The results showed that the statements did not differ in terms of average sen~ tence length, number of words, or number of sentences. Overall, the interrater reliabil~ ity was a bit higher for the RM criteria (assessed with the Memory Characteristics Questionnaire; Johnson, Foley, Suengas, & Raye, 1988) than for the CBCA criteria. Of the CBCA criteria, logical consistency and contextual embedding significantly differentiated the two types of statements, although this was true only for statements that were pro~ vided after a short preparation time (as opposed to those provided immediately after the instruction). The correct classification rate for CBCA was 65%. A significant main ef~ feet distinguishing true from false accounts was found for the RM criteria: 71% of all statements were correctly classified. A discriminant analysis including both CBCA and RM criteria correctly classified 79% of the statements.

362

GRIESEL AND YUILLE

In conclusion, it remains unclear whether CBCA or RM is the preferable approach to verbal deception detection. Vrij (2000) suggested that RM might be more effective in adults because children do not distinguish as much between fact and fantasy. If it turned out that CBCA and RM are equally useful, Vrij (2000) would prefer RM, simply because it has a more profound theoretical underpinning and is easier to apply. Also, RM contains criteria indicative of lying, in contrast to CBCA, which is exclusively focused on detection of truth. It was further suggested that RM scoring is easier to teach than CBCA scoring (Vrij et al., 2004). In Vrij's (2000) opinion, the RM criterion cognitive operations could be useful in differentiating between narrative truths but historical wrongs and truly experienced statements, a distinction CBCA cannot perform. Sporer (1997) found five factor-analyzed dimensions underlying both RM and CBCA, suggesting that an integrative approach would be possible, if not desirable, especially because the study demonstrated a higher accuracy rate for the joint application of CBCA and RM criteria. A recent study involving a staged event (rubbing the blackboard) by Vrij and colleagues (2004) revealed correct classification rates of 60% for CBCA and 74% for RM. When both methods were included in a discriminant analysis, only RM contributed significantly. Thus, the inclusion of lie-telling criteria (cognitive operations) improved accuracy.

MODERATORS OF THE STRENGTH OF DECEPTION CUES

This last section is not about specific lie detection tools. Instead, it informs about various factors that potentially moderate the accuracy of detecting deception. One focus is on professionals. One might assume that they are better at detecting lies in comparison to laypeople. Furthermore, training of lie detectors as well as personality traits of liars and their impact on lie detection are discussed. Due to their experience on the job, certain occupational groups such as law enforcement officers, judges, secret service agents, or psychiatrists are expected to be better at detecting lies than the average person. Unfortunately, research has consistently demonstrated that their performance at detecting lies does not differ from control groups' performance (e.g., students). Professionals' mean accuracy at detecting lies was often found to be around chance level or lower (e.g., DePaulo & Pfeifer, 1986; Porter, Woodworth, & Birt, 2000). The only professional group that stood out were Secret Service agents, who spotted lies with above-chance accuracy (Ekman & O'Sullivan, 1991). Their accuracy was predicted by their ability to recognize emotions. The agents used nonverbal cues instead of or in addition to speech-related cues to deception. They were also more flexible in the way they used the cues. Interestingly, age and experience on the job were negatively correlated with accuracy for some professionals, suggesting that experience reduces accuracy. In a survey, Akehurst, Koehnken, Vrij, and Bull (1996) found that neither laypeople's nor police officers' beliefs about valid deception cues were supported by research. In fact, individuals who perform at chance level in assessing deceit ignore the

12.

CREDIBILITY ASSESSMENT

363

most relevant cues to deception, that is, facial, bodily, voice, and text cues (Ekman, 1997b, 2001; Porter et al., 2002; Stromwall & Granhag, 2003; Vrij & Mann, 2001). For example, most police officers, prosecutors, and judges thought liars were more gaze-aversive than truth tellers. However, to date, research has not found a relationship between deception and gaze aversion. Accurate lie detection is mostly impeded by poor cue usage because of wrongfully held convictions of behavioral lie indicators (e.g., Ekman, O'Sullivan, & Frank, 1999; Porter et al., 2000). Furthermore, when looking at the consistency of statements, one study found that the participants defined consistency in differently ways (Granhag & Stromwall, 2000). The relationship between accuracy and confidence in one's own judgment was often very weak (Anderson, DePaulo, & Ansfield, 2002; DePaulo, Charlton, Cooper, Lindsey, & Muhlenbruck, 1997; Vrij, 2001a). The number of years of experience of interviewing suspects was shown to correlate positively with confidence, but not, however, with accuracy (e.g., Vrij & Mann, 2001). Vrij (2001b) suggested that a greater emphasis be placed on indirect measures of deception detection (e.g., rating whether someone is thinking hard as opposed to judging whether someone is lying), because police officers could not distinguish truths and lies when they were directly asked for a judgment (cf. De Paulo, 1994; Vrij, Edward, & Bull, 2001). Granhag and Stromwall (1999) found that the lie catchers' confidence in the veracity of their judgment increased with repeated interrogations, whereas their accuracy was not different from that of participants who viewed a suspect statement only once. Granhag and Stromwall (2001) further demonstrated that active interrogators had a greater truth bias than passive observers of interrogations. The question arises whether professionals' (e.g., law enforcement investigators') lie detection skills could be trained. An early study by Koehnken (1987) showed that various trainings in deception cues did not improve police officers' deception detection skills. Similar results were obtained in a recent effect size analysis of previous deception detection research (Meissner & Kassin, 2002). What is necessary to make professional deception detection trainings more successful? It was suggested that professionals are rather poor lie catchers because they do not obtain corrective feedback after spotting a potential lie (Ekman, 1997b, 1996). Indeed, several studies have shown that feedback training enhanced the accuracy of deception detection (Porter et al., 2000; Vrij, 1994; Zuckerman, Koestner, & Colella, 1985). After training, accuracy rates were as high as 77%, whereas before they were around chance level. O'Sullivan and Ekman's most recent work (2005) demonstrated that very few people have exceptional deception detection skills. Statistically, these people are so rare that they called them wizards, people with amazing skills or accomplishments. To date, only 29 wizards have been identified from a pool of participants of workshops offered by these researchers. In order to be labeled a wizard, one had to score above 80% on two lie detection measures, respectively. These people are currently being studied as individuals. The research is ongoing; therefore only observations (as opposed to thorough empirical findings) have been reported to date. It seems as if all wizards primarily use nonverbal cues to identify lying behavior. They avoid difficulties that most other people have using nonverbal behaviors as clues to deception. For example, they did not assume that there is a single clue that can betray a lie. Furthermore, most wizards had unusual childhoods

364

GRIESEL AND YUILLE

(e.g., not speaking English up to a certain age, being children of alcoholics), which made them aware of changes in emotional levels at a very young age. Most of them valued excellent performance and were characterized by thorough and complex cognitive styles. Most of them were introverts. Thus, although talent seems to be necessary, the ability to detect lies can be acquired by much practice, feedback, and motivation (see also O'Sullivan, 2005). Characteristics of the liar that may influence lie detection accuracy include attractiveness and personality. For example, Porter and colleagues' (2002) study revealed that attractive liars were rated less accurately than unattractive liars. Thus, attractiveness supported the truth bias. The liar's personality (e.g., social adroitness, self-monitoring, expressiveness) was associated with credibility (e.g., DePaulo, 1992; Vrij et al., 2002). Certain situational circumstances can also drive rather sincere people to tell lies. Whereas the relationship between behavioral cues and deception is generally weak, it becomes more pronounced when people are motivated to succeed at lying (DePaulo et al., 2003). Paradoxically, the more motivated the liars, the easier they are to see through (see motivational impairment effect; DePaulo, 1992; DePaulo & Kirkendol, 1989). Liars try to avoid making too many movements and maintaining eye contact in order to avoid behaviors commonly believed to be associated with lying. Thus, they appear overcontrolled. One personality trait, self-consciousness, was shown to be especially counterproductive during lying under high motivation (DePaulo, 1992).

SUMMARY AND CONCLUSION

Throughout this chapter it becomes clear that lie detection has its limitations. Although certain cues-whether physiological, behavioral, or verbal-can help to uncover lies, they can never be unambiguously related to lying behavior. An increased heart rate, avoiding eye contact, or telling an inconsistent story could always be caused by internal states other than lying. From an evolutionary perspective this makes great sense, as overly obvious cues to deception would not have survived: "Evolution favors more flexible deceivers" (DePaulo et al., 2003; p. 81). For a number of reasons, it has been impossible to develop the available tools to the quality level of psychometric scales with norms that inform about lying thresholds. Given the moderate validity of individual lie detection approaches, most tools ought to be applied only during the investigative stage of a case, but not when evidence is presented in court. A combination of the approaches introduced in this chapter is desirable. Recently, DePaulo and colleagues (2003) published an extensive paper presenting their self-presentational perspective of deception research, integrating many of the cues and approaches reviewed above, for example, decreased forthcoming and less compelling tales (as in SVA) and nervousness resulting in higher pitch and arousal (variables presented in the section on physiological markers of deceit). Each one of the techniques introduced in this chapter has proved to be valid in detecting lies to a certain point (i.e., better than chance level of performance), but unfortunately their accuracy is only moderate, thus suggesting that their combined use may be beneficial (e.g., DePaulo et al., 2003; Rassin, 1999; Sporer, 1997). The combination

12. CREDIBILITY ASSESSMENT

365

of cues could potentially improve the deception cues' clarity and strength, thereby increasing the degree to which lies could be discriminated from truths. However, the practicality of combining physiological, behavioral, and content clues to deception remains unexplored. Perhaps the single clear point to emerge from a review of credibility assessment techniques is the need for multiple hypotheses: anyone charged with assessing credibility should entertain alternative explanations for the change in physiology, behavior, or speech. Detecting a change should be the start of exploring credibility, not the end. The reason(s) for the change must still be determined, and lying is only one of many possibilities.

REFERENCES Akehurst, L., Koehnken, G., Vrij, A., & Bull, R. (1996). Lay persons' and police officers' beliefs regarding deceptive behavior. Applied Cognitive Psychology, 10,461-471. Anderson, D. E., DePaulo, B. M., & Ansfield, M. E. (2002). The development of deception detection skill: a longitudinal study of same-sex friends. Personality and Social Psychology Review, 28(4), 536-545. Ben-Shakkar, G., Bar-Hillel, M., & Kremnitzer, M. (2002). Trial by polygraph: reconsidering the use of the guilty knowledge technique in court. Law and Human Behavior, 26(5), 527-541. Ben-Shakhar, G., & Furedy, J. J. (1990). Theories and applications in the detection of deception: A psychophysiological and international perspective. New York: Springer-Verlag. Bruck, M., Ceci, S. J., & Hembrooke, H. (1998). Reliablity and credibility of young children's reports. American Psychologist, 53(2), 136-151. Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113 (3), 403-439. Cockburn, A. (1993, March). Janet Reno's coerced confession. The Nation, pp. 296-297. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. Cohn, J. F., Zlochower, A. J., Lien, J., & Kanade, T. (1999). Automated face analysis by feature point tracking has high concurrent validity with manual FACS coding. Psychophysiology, 36, 35-43. DePaulo, B. M. (1992). Nonverbal behavior and self-presentation. Psychological Bulletin, 111 (2), 203-243. DePaulo, B. M. (1994). Spotting lies: Can humans learn to do better? Current Directions in Psychological Science, 3(3), 83-86. DePaulo, B. M., Charlton, K., Cooper, H., Lindsay, J. J., & Muhlenbruck, L. (1997). The accuracyconfidence correlation in the detection of deception. Personality and Social Psychology Review, 1(4), 346-357. DePaulo, B. M., & Kirkendol, S. E. (1989). The motivational impairment effect in the communication of deception. In J. C. Yuille (Ed.), Credibility assessment (pp. 51-70). Dordrecht, the Netherlands: Kluwer Academic. DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129(1), 74-118. DePaulo, B. M., & Pfeifer, R. L. (1986). On-the-job experience and skill at detecting deception. Journal of Applied Social Psychology, 16(3), 249-267. Ekman, P. (1985). Telling lies: Clues to deceit in the marketplace, marriage and politics. New York: Norton.

366

GRIESEL AND YUILLE

Ekman, P. (1992). Facial expressions of emotion: New findings, new questions. Psychological Science, 3(1), 34-38. Ekman, P. (1993). Facial expression and emotion. American Psychologist, 48(4), 376-379. Ekman, P. (1996). Why don't we catch liars? Social Research, 63(3), 801-817. Ekman, P. (1997a). Conclusion: What we have learned from measuring facial behavior. In P. Ekman, P., & E. Rosenberg (Eds.), What the face reveals (pp. 469-485). New York: Oxford University Press. Ekman, P. (1997b). Lying and deception. InN. L. Stein, P. A. Ornstein, B. Tversky, & C. Brainerd (Eds.), Memory for everyday and emotional events. Mahwah, NJ: Lawrence Erlbaum Associates. Ekman, P. (2001). Telling lies. New York: W. W. Norton & Company. Ekman, P., & Friesen, W. V. (1978). Facial action coding system. Palo Alto, CA: Consulting Psychologists Press. Ekman, P., Friesen, W. V., & O'Sullivan, M. (1997/1988). Smiles when lying. In P. Ekman & E. Rosenberg (Eds.), What the face reveals (pp. 201-214). New York: Oxford University Press. Ekman, P., Friesen, W. V., & Scherer, K. R. (1978). Body movement and voice pitch in deceptive interaction. Semiotica, 16(1), 23-27. Ekman, P., & O'Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46(9), 913-920. Ekman, P., O'Sullivan, D. M., & Frank, M.G. (1999). A few can catch a liar. Psychological Science, 10(3), 263-266. Ekman, P., Sorenson, E. R., & Friesen, W. V. (1969). Pan-cultural elements in facial displays of emotions. Science, 164(3875), 86-88. Elaad, E. (1990). Detection of guilty knowledge in real-life criminal investigations. Journal of Applied Psychology, 75(5), 521-529. Elaad, E. (1998). The challenge of the concealed knowledge polygraph test. Expert Evidence, 6, 161-187. Etcoff, N. L., Ekman, P., Magee, J. J., & Frank, M.G. (2000). Lie detection and language comprehension. Nature, 405, 139-141. Farwell, L. A., & Donchin, E. (1991). The truth will come out: Interrogative polygraphy ("lie detection") with event-related brain potentials. Psychophysiology, 28(5), 531-547. Fiedler, K., Schmidt, J., & Stahl, T. (2002). What is the current truth about polygraph lie detection? Basic and Applied Social Psychology, 24(4), 313-324. Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques for investigative interviewing: The Cognitive Interview. Springfield, IL: Charles C Thomas. Frank, M.G., & Ekman, P. (1997). The ability to detect deceit generalizes across different types of high-stake lies. Journal of Personality and Social Psychology, 72 (6), 1429-1439. Gallai, D. (1999). Polygraph evidence in federal court: should it be admissible? American Criminal Law Review, 36, 87-108. Goodman, G. S., Aman, C., & Hirschman, J. (1987). Child sexual and physical abuse: Children's testimony. InS. J. Ceci, Ross, D., & Toglia, M. (Eds.), Children's eyewitness memory (pp. 123). New York: Springer-Verlag. Gosslin, P., Kirouac, G., & Dore, F. (1997/1995). Components and recognition offacial expressions in the communication of emotion by actors. In P. Ekman & E. Rosenberg (Eds.), What the face reveals (pp. 243-270). New York: Oxford University Press. Granhag, P. A., & Stromwall, L.A. (1999). Repeated interrogations-stretching the deception detection paradigm. Expert Evidence, 7, 163-174. Granhag, P. A., & Strom wall, L. A. (2000). Effects of perceptions on deception detection and new answers to why lie-catchers often fail. Psychology, Crime & Law, 6, 197-218. Granhag, P. A., & Stromwall, L. A. (2001). Deception detection based on repeated interrogations. Legal and Criminological Psychology, 6, 85-101. Gudjonsson, G. (2001). False confession. Psychologist, 14(11), 588-591.

12.

CREDIBILITY ASSESSMENT

367

Hershkowitz, I. (1999). The dynamics of interviews involving plausible and implausible allegations of child sexual abuse. Applied Developmental Science, 3(2), 86-91. Hess, U., & Kleck, R. E. (1997 /1990). Differentiating emotions elicited and deliberate emotional facial expressions. In P. Ekman & E. Rosenberg (Eds.), What the face reveals (pp. 271-288). New York: Oxford University Press. Honts, C. R. (1994). Assessing children's credibility: Scientific and legal issues in 1994. North Dakota Law Review, 70, 879-903. Honts, C. R., & Amato, S. L. (2002). Countermeasures. In M. Kleiner (Ed.), Handbook of polygraph testing (pp. 251-264). San Diego: Academic Press. Honts, C. R., Kircher, J. C., & Raskin, D. C. (1995). Polygrapher's dilemma or psychologist's chimaera: A reply to Furedy's logico-ethical considerations for psychophysiological practitioners and researchers. International Journal of Psychophysiology, 20, 199-207. Honts, C. R., Raskin, D. C., & Kircher, J. C. (1994). Mental and physical countermeasures reduce the accuracy of polygraph tests. Journal of Applied Psychology, 79(2), 252-259. Horowitz, S. W. (1991). Empirical support for Statement Validity Assessment. Behavioral Assessment, 13, 293-313. Horowitz, S. W., Lamb, M. E., Esplin, P. W., Boychuk, T. D., Krispin, 0., & Reiter-Lavery, L. (1997). Reliability of criteria-based content analysis of child witness statements. Legal and Criminological Psychology, 2, 11-21. Horowitz, S. W., Lamb, M. E., Esplin, P. W., Boychuk, T. D., Reiter-Lavery, L., & Krispin, 0. (1995). Establishing ground truth in studies of child sexual abuse. Expert Evidence, 4, 42-51. Hubert, H. H. (1985). The Scott County investigation. Retrieved February 14, 2004 from www .a-team.org. Iacono, W. G., & Lykken, D. T. (1997). The validity of the lie detector: Two surveys of scientific opinion. Journal of Applied Psychology, 82 (3), 426-433. Johnson, M. K. (1988). Reality monitoring: An experimental phenomenological approach. Journal of Experimental Psychology: General, 117(4), 390-394. Johnson, M. K., Foley, M., Suengas, A., & Raye, C. (1988). Phenomenal characteristics of memories for perceived and imagined autobiographical memories. Journal of Experimental Psychology: General, 117(4), 371-376. Johnson, M. K., & Raye, C. L. (1998). False memory and confabulation. Trends in Cognitive Sciences, 2(4), 137-145. Kircher, J. C., & Raskin, D. C. (1992). Polygraph techniques: history, controversies, and prospects. In P. Suedfeld & P. E. Tetlock (Eds.), Psychology and social policy (pp. 295-308). Washington, DC: Hemisphere. Koehnken, G. (1987). Training police officers to detect deceptive eyewitness statements: Does it work? Social Behaviour, 2, 1-17. Koehnken, G., Schimossek, E., Aschermann, E., & Hoefer, E. (1995). The Cognitive Interview and the assessment of the credibility of adults' statements. Journal of Applied Psychology, 80(6), 671-684. Lamb, M. E., Sternberg, K. J., Esplin, P. W., Hershkowitz, I., & Orbach, Y. (1997). Assessing the credibility of children's allegations of sexual abuse: A survey of recent research. Learning and Individual Differences, 9(2), 175-194. Lamb, M. E., Sternberg, K. J., Esplin, P. W., Hershkowitz, I., Orbach, Y., & Hovav, M. (1997). Criterion-Based Content Analysis: A field validation study. Child Abuse & Neglect, 21 (3), 255-264. Landry, K. L., & Brigham, J. C. (1992). The effect of training in Criteria-Based Content Analysis on the ability to detect deception in adults. Law and Human Behavior, 16(6), 663-676. Lindsay, D. S., & Read, J.D. (1995). "Memory work" and recovered memories of childhood sexual abuse: Scientific evidence and public, professional, and personal issues. Psychology, Public Policy, and Law, 1(4), 846-908.

368

GRIESEL AND YUILLE

Loftus, E. F. (1993). The reality of repressed memories. American Psychologist, 48(5), 518-537. MacLaren, V. V. (2001). A quantitative review of the guilty knowledge test. Journal of Applied Psychology, 86(4), 674-683. Mann, S., Vrij, A., & Bull, R. (2002). Suspects, lies and videotape: An analysis of authentic highstake liars. Law and Human Behavior, 26(3), 365-376. Meissner, C. A., & Kassin, S.M. (2002). "He's guilty!": Investigator bias in judgments of truth and deception. Law and Human Behavior, 26(5), 469-480. O'Sullivan, M. (2005). Emotional intelligence and deception detection: Why most people can't "read" others, but a few can. R. E. Riggio & R. S. Feldman (Eds.), Applications of nonverbal communication. Mahwah, NJ: Lawrence Erlbaum Associates. O'Sullivan, M., & Ekman, P. (Eds.). (2005). The wizards of deception detection. In P. A. Granhag & L. A. Stromwell (Eds.), Deception detection in forensic contexts. Cambridge: Cambridge University Press. Patrick, C. J., & Iacono, W. G. (1991). Validity of the control question polygraph test: The problem of sampling bias. Journal of Applied Psychology, 76(2), 229-238. Porter, S., Birt, A. R., Yuille, J. C., & Herve, H. F. (2001). Memory for murder: A psychological perspective on dissociative amnesia in legal contexts. International Journal of Law and Psychiatry, 24, 23-41. Porter, S., & Campbell, M.A. (2002). Pinpointing reality: how well can people judge true and mistaken emotional childhood memories? Canadian Journal of Behavioural Science, 34(4), 217-229. Porter, S., Campbell, M.A., Stapelton, J., & Birt, A. R. (2002). The influence of judge, target, and stimulus characteristics on the accuracy of detecting deceit. Canadian Journal of Behavioural Science, 34(3), 172-196. Porter, S., Woodworth, M., & Birt, A. R. (2000). Truth, lies, and videotape: An investigation of the ability of federal parole officers to detect deception. Law and Human Behavior, 24(6), 643-658. Porter, S., & Yuille, J. (1996). The language of deceit: An investigation of the verbal clues to deception in the interrogation context. Law and Human Behavior, 20(4), 443-458. Raskin, D. C. (1987). Methodological issues in estimating polygraph accuracy in field applications. Canadian Journal of Behavioural Science, 19(4), 389-404. Raskin, D. C., Kircher, J. C., Horowitz, S. W., & Honts, C. R. (1989). Recent laboratory and field research on polygraph techniques. InJ. C. Yuille (Ed.), Credibility assessment. Dordrecht, the Netherlands: Kluwer Academic. Raskin, D. C., & Yuille, J. C. (1989). Problems in evaluating interviews of children in sexual abuse cases. InS. J. Ceci, D. F. Ross, & M.P. Toglia (Eds.), Perspectives on children's testimony (pp. 184-207). New York: Springer. Rassin, E. (1999). Criteria Based Content Analysis: The less scientific road to truth. Expert Evidence, 7(4), 265-278. Rosenberg, E. (1997). Introduction: The study of spontaneous facial expressions in psychology. In P. Ekman & E. Rosenberg (Eds.), What the face reveals (pp. 3-17). New York: Oxford University Press. Rosenberg, E., & Ekman, P. (1997/1994). Coherence between expressive and experiential systems in emotion. In P. Ekman & E. Rosenberg (Eds.), What the face reveals (pp. 63-88). New York: Oxford University Press. Rosenfeld, P. (2002). Event-related potentials in the detection of deception, malingering, and false memories. In M. Kleiner (Ed.), Handbook of polygraph testing (pp. 265-286). San Diego: Academic Press. Ruby, C. L., & Brigham, J. C. (1997). The usefulness of the Criteria-Based Content Analysis technique in distinguishing between truthful and fabricated allegations: A critical review. Psychology, Public Policy, and Law, 3(4), 705-737.

12.

CREDIBILITY ASSESSMENT

369

Saxe, L., & Ben-Shakkar, G. (1999). Admissibility of polygraph tests: The application of scientific standards post-Daubert. Psychology, Public Policy, and Law, 5(1), 203-223. Sayette, M.A., Cohn, J. F., Wertz, J. M., Perrott, M.A., & Parrott, D. J. (2001). A psychometric evaluation of the Facial Action Coding System for assessing spontaneous expressions. Journal ofNonverbalBehavior, 25(3), 167-185. Seymour, T. L., Seifert, C. M., Shafto, M.G., & Mosmann, A. L. (2000). Using response time measures to assess "Guilty Knowledge." Journal of Applied Psychology, 85(1), 30-37. Sporer, S. L. (1997). The less traveled road to truth: Verbal cues in deception detection in accounts of fabricated and self-experienced events. Applied Cognitive Psychology, 11, 3 73397. Steller, M. (1989). Recent developments in Statement Analysis. In J. C. Yuille (Ed.), Credibility assessment. Dordrecht, the Netherlands: Kluwer Academic. Steller, M., & Koehnken, G. (1989). Statement Analysis: Credibility assessment of children's testimonies in sexual abuse cases. In D. C. Raskin (Ed.), Psychological methods in criminal investigation and evidence. New York: Springer. Stromwall, L.A., & Granhag, P. A. (2003). How to detect deception? Arresting the beliefs of police officers, prosecutors and judges. Psychology, Crime & Law, 9, 19-36. Tully, B. (1998). Reliability of criteria-based content analysis of child witness statements: Cohen's kappa doesn't matter. Legal and Criminological Psychology, 3, 183-188. Tye, M. C., Amato, S. L., Honts, C. R., Devitt, M. K., & Peters, D. (1999). The willingness of children to lie and the assessment of credibility in an ecologically relevant laboratory setting. Applied Developmental Science, 3(2), 92-110. Undeutsch, U. (1989). The development of Statement Reality Analysis. In J. C. Yuille (Ed.), Credibility assessment (pp. 101-120). Dordrecht, the Netherlands: Kluwer Academic. Vrij, A. (1994). The impact of information and setting on detection of deception by police detectives. Journal of Nonverbal Behavior, 18 (2), 117-136. Vrij, A. (2000). Detecting lies and deceit: The psychology oflying and the implications for professional practice. Chichester: Wiley. Vrij, A. (2001a). Credibility judgment of detectives: The impact of nonverbal behavior, social skills, and physical characteristics on impression formation. The Journal of Social Psychology, 133(5), 601-610. Vrij, A. (2001b). Detecting the liars. Forensic Psychology, 14(1), 596-598. Vrij, A., Akehurst, L., Soukara, S., & Bull, R. (2002). Will the truth come out? The effect of deception, age, status, coaching, and social skills on CBCA scores. Law and Human Behavior, 26(3), 261-283. Vrij, A., Akehurst, L., Soukara, S., & Bull, R. (2004). Let me inform you how to tell a convincing story: CBCA and Reality Monitoring scores as a function of age, coaching, and deception. Canadian Journal of Behavioural Science, 36 (2), 113-126. Vrij, A., Edward, K., & Bull, R. (2001). Police officers' ability to detect deceit: The benefit of indirect deception detection measures. Legal and Criminological Psychology, 6, 185-196. Vrij, A., Edward, K., Roberts, K. P., & Bull, R. (2000). Detecting deceit via analysis of verbal and nonverbal behavior. Journal of Nonverbal Behavior, 24 (4), 239-263. Vrij, A., & Graham, S. (1997). Individual differences between liars and the ability to detect lies. Expert Evidence, 5(4), 144-148. Vrij, A., & Mann, S. (2001). Who killed my relative? Police officers' ability to detect real-life high-stake lies. Psychology, Crime & Law, 7, 119-132. Yuille, J. C. (1988a). A simulation study of criterion based content analysis. Paper presented at the NATO Advanced Study Institute on Credibility Assessment, Maratea, Italy. Yuille, J. C. (1988b). The systematic assessment of children's testimony. Canadian Psychology, 29(3), 247-262. Yuille, J. C. (Ed.). (1989). Credibility assessment. Dordrecht, Boston, London: Kluwer Academic.

370

GRIESEL AND YUILLE

Yuille,]. C. (1990). Use of the criteria-based content analysis. Unpublished manuscript, University of British Columbia, Vancouver. Yuille, J. C., Marxsen, D., & Cooper, B.S. (1999). Training investigative interviewers: Adherence to the spirit, as well as the letter. International journal of Law and Psychiatry, 22 (3-4), 323-336. Yuille,]. C., Tymofievich, M., & Marxsen, D. (1995). The nature of allegations of child sexual abuse. InT. Ney (Ed.), True and false allegations of child sexual abuse. Assessment and case management. New York: Brunner/Mazel. Zaparniuk, J., Yuille, J. C., & Taylor, S. (1995). Assessing the credibility of true and false statements. Intemational]oumal of Law and Psychiatry, 18(3), 343-352. Zuckerman, M., & Driver, R. E. (1985). Telling lies: verbal and nonverbal correlates of deception. In A. W. Siegman & S. Feldstein (Eds.), Multichannel integration of nonverbal behavior (pp. 129-147). Hillsdale, NJ: Lawrence Erlbaum Associates. Zuckerman, M., Koestner, R., & Colella, M. ]. (1985). Learning to detect deception from three communication channels. Journal of Nonverbal Behavior, 9 (3), 188-194.

13 Eyewitness Confidence from the Witnessed Event Through Trial Johns. Shaw, III Lafayette College

Kimberley A. McClure Western Illinois University

Josie A. Dykstra Lafayette College

It seems to be a matter of common sense that eyewitness confidence should be strongly related to eyewitness accuracy. People should be a good judge of what they know and what they don't know. Indeed, surveys conducted over the past two decades have shown that laypeople (e.g., Brigham & Bothwell, 1983; Deffenbacher & Loftus, 1982; Yarmey &Jones, 1983) and even members of the legal profession (Brigham & WolfsKeil, 1983) believe that eyewitness confidence is a reliable indicator of eyewitness accuracy. Even the United States Supreme Court (Manson v. Braithwaite, 1977; Neil v. Biggers, 1972/1979) has held that jurors should be instructed to consider witness confidence in evaluating identification testimony. As the court stated in Neil v. Biggers and reaffirmed in Manson v. Braithwaite, "the factors to be considered in evaluating the likelihood of misidentification include the opportunity of the witness to view the criminal at the time of the crime, the witness's degree of attention, the accuracy of the witness's prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation" (emphasis added). Yet despite the generally accepted belief that eyewitness confidence and eyewitness accuracy should be related, the bulk of the empirical evidence collected in laboratory and field experiments over the past 25 years indicates that eyewitness confidence is generally 371

372

SHAW, MCCLURE, DYKSTRA

not a good predictor of eyewitness accuracy (e.g., Bothwell, Deffenbacher, & Brigham, 1987; Deffenbacher, 1980, 1991; Leippe, 1980; Luus & Wells, 1994a; Penrod & Cutler, 1995; Shaw, 1996; Shaw & McClure, 1996; Wells, 1993; Wells & Bradfield, 1998; cf. Lindsay, Read, & Sharma, 1998; Read, Lindsay, & Nicholls, 1998; Sporer, Penrod, Read, & Cutler, 1995). The erroneous belief that eyewitness confidence is a good predictor of eyewitness accuracy is troublesome for the legal system, because trial-simulation studies have shown that jurors accord a great deal of weight to eyewitness confidence when evaluating eyewitness testimony (e.g., Cutler, Penrod, & Dexter, 1990; Cutler, Penrod, & Stuve, 1988; Lindsay, Wells, & O'Connor, 1989; Wells, Lindsay, & Ferguson, 1979). These findings imply, of course, that jurors rely heavily on a factor that may have little, if any, probative value in the evaluation of the accuracy of eyewitness testimony. In his seminal review of the eyewitness confidence literature, Leippe (1980) suggested that there are many factors that may affect eyewitness confidence independently of eyewitness accuracy. The primary focus of this chapter is to summarize the current state of our knowledge about the factors that influence eyewitness confidence throughout the witnessing experience. The chapter is structured around the three stages of the typical eyewitness experience. The first stage involves viewing the target event, when the encoding of the original memory occurs. During this stage, variables that are not under the control of the legal system, which Wells (1978) has called estimator variables, such as whether the culprit wore a disguise, may shape witness confidence. The second stage-the time between the witnessed event and trial-can last for months or even years. During this stage, a wide range of memory-related events, including witness questioning, attending lineups and showups, and exposure to information from other witnesses, the police, and the media can alter eyewitness confidence independently from eyewitness memory. Most of these events involve system variables (Wells, 1978), which are factors that are under the control of the criminal justice system. In the final stagetestimony at trial-there are numerous cognitive and social pressures that may influence the publicly stated confidence of witnesses who testify in court. Finally, we conclude with a brief discussion of the forensic utility of eyewitness confidence as a predictor of eyewitness accuracy. Although much of the research on eyewitness confidence involves the identification of faces and people, with a particular emphasis on the processes involved in lineups and showups, the focus of this chapter is on eyewitness confidence more generally, particularly involving memories for events. For a detailed exploration of the identification of people, as well as an in-depth examination of the statistical relationship between confidence and accuracy and the measurement issues arising from the various methods of describing the relationship between eyewitness confidence and eyewitness accuracy, the reader is referred to Leippe and Eisenstadt in Volume II of this handbook.

INITIAL WITNESSED EVENT

Many aspects of a witnessed event can influence information processing and can determine the information that is attended to, encoded, and subsequently available for retrieval

13.

EYEWITNESS CONFIDENCE

373

during police investigations and court proceedings. These factors, which include the characteristics of the event, the eyewitness, and the culprit, may directly or indirectly affect eyewitness accuracy, eyewitness confidence, or both (Leippe, 1980).

Event Characteristics It is well documented that context provides memory cues for subsequent retrieval (e.g., Bjork & Richardson-Klavehn, 1989). If the witnessing context is "optimal"-for example, the target event takes place in broad daylight for a lengthy duration-the encoding conditions are better than if the event occurs at midnight in a dark alley for a few seconds. Subsequent retrieval will be more accurate in the former scenario than in the latter (Brown, Deffenbacher, & Sturgill, 1977; Deffenbacher, 1980). As proposed by Deffenbacher (1980), optimal viewing conditions theoretically lead to well-calibrated confidence (i.e., the witness is neither underconfident nor overconfident). Needless to say, however, most witnessed events do not occur under optimal viewing conditions. Viewing conditions vary greatly, depending upon the circumstances and type of event experienced. Factors that may possibly affect eyewitness confidence include the time of day, whether the location is isolated or well populated, and whether the setting is familiar or unfamiliar. However, research investigating event circumstances has typically focused on how event factors affect witness accuracy instead of witness confidence (e.g., Behrman & Davey, 2001). If confidence is assessed, it is typically a secondary consideration. Researchers investigating the effects of event characteristics on witness confidence have examined (a) the influence of violence (Clifford & Hollin, 1981; Granhag, 1997); (b) emotionally charged events, such as those conducive to "flashbulb" memories (Talarico & Rubin, 2003); (c) whether the witness was a bystander or a victim (Hosch & Bothwell, 1990; Hosch & Cooper, 1982); (d) whether the witness was sleep-deprived (Blagrove & Akehurst, 2000); and (e) whether the witness saw or only heard the event (Olsson, Juslin, & Winman, 1998; Yarmey, 1986).

Violence. Clifford and his associates have investigated the effects of event violence on eyewitness accuracy and confidence. In the work of Clifford and Hollin (1981), participants viewed either a film depicting a violent mugging or a scenario where someone was asking for directions. The number of perpetrators involved in the scenarios was also manipulated. Neither the violence of the incident nor the number of perpetrators changed witness confidence. Although participant confidence did not differ by condition, in the nonviolent condition there were more correct identifications than incorrect, and participants were more confident in their correct identifications. More recently, Granhag (1997) also manipulated the violence depicted in a filmed scenario, and there were no effects of film violence on witness confidence or witness accuracy. Taken collectively, these studies do not provide much information about whether violence during the witnessed event affects eyewitness confidence. Emotionally Charged Events. Although the violence of an event, by itself, does not appear to change witness confidence, events with high emotional content (which can

374

SHAW, MCCLURE, DYKSTRA

generate flashbulb memories) may be reported with consistently high witness confidence. Talarico and Rubin (2003) assessed participants' memories for the terrorist attacks on the United States 1 day after the September 11, 2001 assault. The participants' memories were tested again 7, 42, and 224 days after the event. Participants completed open-ended questions about the terrorist attacks and about an everyday event. For each event, they also reported various other characteristics, such as the emotional content and their confidence. Talarico and Rubin found that the two types of events differed in terms of participants' confidence in the accuracy of their memory reports. While recollection declined across time for both the flashbulb memory of the terrorist attacks and for the everyday memory, confidence in the flashbulb memory remained high even after 224 days.

Victims and Bystanders. During an initial witnessed event, a witness may be in one of two roles-victim or bystander. Hosch and his associates (Hosch & Bothwell, 1990; Hosch & Cooper, 1982) investigated the effects of victimization on eyewitness identification accuracy and confidence. Hosch and Cooper found that the confidence ratings of bystanders, victims, and control participants did not differ, and the correlation between confidence and accuracy was not statistically significant for any of the three types of witnesses. Similarly, Hosch and Bothwell found no differences between bystanders and victims as to either identification accuracy or confidence, and there was no significant confidence-accuracy correlation for either group. Based on this very limited evidence from research laboratories, it appears that there are no significant differences in the confidence of bystanders and victims.

Sleep Deprivation. Blagrove and Akehurst (2000) compared participants who were sleep-deprived for 29-50 hours with controls on logical reasoning, suggestibility, and confidence. Participants responded to questions about a story immediately and then after one- and two-day delays. Sleep-deprived participants were more suggestible, but remained just as confident as controls; that is, the sleep loss had no discernible effect on witness confidence. Earwitnesses. Visual information may not always be available to a witness during a witnessed event. In those circumstances, the witness must rely on auditory information. In a study of eyewitness and earwitness information, Yarmey (1986) evaluated participants' responses to increasingly lower levels of illumination from daylight to nighttime. A nonsignificant trend in the confidence ratings for lineups containing the target suggested that as visual conditions worsened (i.e., moved closer to darkness) the participants were less confident in their visual identification responses. There was a statistically significant effect of illumination on confidence for target absent lineups. Participant confidence was higher for correct rejections during daylight conditions than for twilight. Generally, participants were less confident in their choices from the target-absent voice lineup than from the target-present voice lineup. Olsson, Juslin, and their associates (Olsson et al., 1998; Juslin, Olsson, & Winman, 1996) found that participants were less accurate and more overconfident for voice stimuli than for face stimuli. In a direct comparison of earwitness and eyewitness perfor-

13.

EYEWITNESS CONFIDENCE

375

mance, participants in the sudy by Olsson et al. (1998, Experiment 2) engaged in a standard recognition paradigm in which they viewed target faces, listened to target voices, and subsequently engaged in visual and auditory recognition tasks. Participants' confidence in their voice identifications had no predictive value in terms of accuracy, and participants were more overconfident for auditory than for visual information.

Eyewitness Characteristics Witness characteristics, such as gender and age, may be related to eyewitness confidence. Yarmey (Yarmey, 1986; Yarmey & Yarmey, 1997) reported that male eyewitnesses tend to exhibit more confidence than females (cf. Yarmey, 1993, for an exception). In a field study of eyewitness memory, Yarmey and Yarmey (1997) found that although men and women did not differ on eyewitness performance measures, the men were more confident than the women. Furthermore, Pallier (2003) found that men are generally more confident than women on both general knowledge and visual perception tasks, which implies that men might be more confident in eyewitness settings as well. As for age differences, several studies demonstrated that children are especially vulnerable to factors that can affect eyewitness accuracy and confidence. Sykes, Gruneberg, and Turner (1987) tested 8-10-year-olds' recognition accuracy and confidence. Children were exposed to a confederate who interacted with their teacher in a classroom setting and were subsequently interviewed either immediately after the event or after a 1-week delay. Sykes et al. found that the children's confidence and accuracy were positively correlated. In addition, they found that the children's confidence increased after the 1-week delay. More recently, Erskine, Markham, and Howie (2001) investigated age effects on confidence and accuracy for central and peripheral information. Children (5-6-year-olds and 9-10-year-olds) viewed two versions of a slide event depicting a trip to McDonald's. In one version three central details were omitted from the presentation, and in the other version three peripheral details were omitted. Children were interviewed about the trip after a 90-min or 1-week delay. Although the pattern of results was complex, several general findings are relevant here. For central details, increased confidence was associated with decreased accuracy for both age groups, but there was no association between confidence and accuracy for peripheral details. Moreover, the younger children seemed especially vulnerable to being overconfident. Older adults may exhibit differences in eyewitness confidence compared with younger adults (e.g., O'Rourke, Penrod, Cutler, & Stuve, 1989; Yarmey & Kent, 1980). Yarmey (1993) found that younger and middle-aged adults were more confident in their responses than older adults. Adams-Price (1992) found a modest negative relationship between age and identification accuracy, but no age differences in confidence or the confidenceaccuracy correlation. Scogin, Calhoon, and D'Errico (1994) found that older adults were less adept than younger adults at identifying a young culprit, but they did not differ in reported confidence, which was high for all participants. In addition, consistent with the findings of Adams-Price and Yarmey, there was no significant confidence-accuracy correlation for either age group (young or older adults).

376

SHAW, MCCLURE, DYKSTRA

In an extension of the work of Scogin et al., Wright and Stroud (2002) manipulated the age of the culprit in their study of identification accuracy for young and older adults. For witness accuracy, there was an own-age bias for culprit-present lineups but not for culprit-absent lineups. Older participants were more accurate in identifying the older culprit than the younger culprit, and younger participants were better at identifying the younger culprit than the older culprit in the target present lineups. Overall, accurate decisions were made with more confidence than incorrect choices, and there was a significant confidence-accuracy correlation in all conditions, although the magnitude of the correlation was larger for the younger participants.

Culprit Characteristics Culprit characteristics may also affect what eyewitnesses encode about the initial event and how they subsequently respond during police interviews. Whether the culprit has distinctive characteristics or wears a disguise may possibly affect eyewitness confidence, but the research investigating these factors has focused almost exclusively on identification accuracy instead of confidence. In those few studies that have assessed confidence, it has usually been in the context of calculating a confidence-accuracy correlation. For example, Brigham (1990) found that the confidence-accuracy correlation was higher for culprits who were high in distinctiveness (r = .41) than for culprits who were low in distinctiveness (r = .25). Although this finding is interesting, it does not tell us whether witnesses tend to be more or less confident in their reports when the culprit is distinctive in appearance. Disguise has a deleterious effect on eyewitness accuracy (e.g., Cutler, Penrod, & Martens, 1987a, 1987b), but it is not known whether a culprit's disguise influences confidence independently of accuracy. The presence of a weapon also adversely affects witnesses' memories for descriptive details and identification accuracy (e.g., Pickel, 1999), but it is unclear how the presence of a weapon affects confidence. Items that draw a witness's attention (e.g., a weapon) are likely to be recalled accurately with high levels of confidence because they are easy to remember (Kebbell, Wagstaff, & Covey, 1996), but it is not known what, if any, effects such items have on confidence associated with other details of the event, such as the description of the culprit. Characteristics of the culprit may also interact with characteristics of the witness to influence eyewitness confidence. As described previously, the relative ages of a culprit and an eyewitness may influence what the witness attends to and subsequently reports in his or her statement (Wright & Stroud, 2002). Although there is an extensive literature investigating how the race and ethnicity of culprits and eyewitnesses affect accuracy in cross-race identifications (for a review, see Meissner & Brigham, 2001), it is less clear how race and ethnicity influence witness confidence. In a field study involving black, white, and Mexican-American targets interacting with convenience store clerks from a large metropolitan area, Platz and Hosch (1987) found higher confidence-accuracy correlations for own-race identifications than other-race identifications, but they did not analyze confidence levels per se. Alley and Schultheis (2001) found no differences in confidence for own- and other-race faces in their study involving computer-generated

13.

EYEWITNESS CONFIDENCE

377

faces. Participants engaged in a standard recognition paradigm, first viewing the target faces and rating them on how aggressive and honest they appeared. Subsequently, they were given a recognition task involving target-absent and target-present lineups. Participants' confidence did not differ between own- and other-race faces. In addition to the research on cross-race identifications, there is some suggestion in the social psychology literature that target race and perceiver race may interact in their effects on perceiver confidence. For example, Ryan, Judd, and Park (1996) investigated participants' perceived variability for own-group versus out-group members. In a series of three studies involving sorority and fraternity members, they found that Asian, black, and white American participants perceiving high variability among out-group members on specified traits were less confident in their judgments about specific out-group members. Although this study did not involve a memory task, one can infer that the same sort of decision-making bias could influence eyewitness confidence.

Summary We know relatively little about how eyewitness confidence is shaped during the first stage of the eyewitness experience. Perhaps the dearth of research on eyewitness confidence during the initial event is due to the fact that these issues are in the "estimator variable universe" (Wells, 1978, 1993) and beyond our control. However, even these estimator variables can influence the outcome of a case. Although we know that eyewitness confidence has a large impact on trial jurors (e.g., Cutler, Penrod, & Dexter, 1990; Cutler, Penrod, & Stuve, 1988; Wells, Lindsay, & Ferguson, 1979), we do not know how initial witness confidence affects the course of a criminal investigation. Are police investigators more likely to follow up witness statements that are made with high levels of confidence? Dahl (2003) found that an eyewitness's choice from a photo-lineup had a large impact on participant-investigators' subsequent decisions to arrest a suspect, but to our knowledge there is no research investigating how eyewitness confidence may shift the focus of an investigation. Clearly additional research needs to be conducted on eyewitness confidence during this first stage of the eyewitness experience.

BETWEEN THE WITNESSED EVENT AND TRIAL

In most criminal and civil cases, witnesses wait months or even years before they have a chance to recount their story in court. This retention period is hardly benign in terms of the eyewitnesses' memories. During this stage, a wide range of memory-related events can alter eyewitness confidence. Questioning techniques such as the cognitive interview, the use of hypnosis, and misleading, suggestive, or repeated questioning can influence eyewitness confidence, as can information from co-witnesses and feedback about the accuracy of a witness's memory. As the focus of this chapter is on eyewitness confidence for events, the reader is referred to the chapter by Leippe and Eisenstadt in Volume II of this handbook for an extensive treatment of eyewitness confidence for person identification tasks such as lineups and showups.

378

SHAW, MCCLURE, DYKSTRA

Questioning Techniques Several questioning techniques that are commonly employed by the police in eyewitness interviews have been the subject of considerable experimental research. Three broad categories of questioning techniques are discussed in this section: (a) the cognitive interview and other forms of context reinstatement, (b) hypnosis, and (c) misleading and suggestive questioning.

Cognitive Interview and Context Reinstatement. Originally developed by Fisher and Geiselman (Geiselman, Fisher, MacKinnon, & Holland, 1985), the cognitive interview incorporates many aspects of effective retrieval strategies. A cognitive interview typically involves mentally reinstating the context of the target event, recalling the event in reverse order, recounting the event from different perspectives, and encouraging the witness to recall every detail of an event, no matter how small. Recent improvements to the procedure have resulted in the enhanced cognitive interview, which takes into account communication styles of the witnesses and the social dynamics of the interviewing situation, among other factors. Research on the efficacy of both versions of the cognitive interview has generally shown that if properly administered, the cognitive interview can yield increases in the total amount of information recalled without decreasing overall accuracy (Fisher, McCauley, & Geiselman, 1994; Memon & Bull, 1991; cf. Kebell & Wagstaff, 1997; McMahon, 2000). For further information, the reader should see Chapter 2 by Fisher and Schreiber on interviewing in this volume of the handbook. Several studies have examined the effect of the cognitive interview on eyewitness confidence. In a comparison of the enhanced cognitive interview with a standard structured interview, McMahon (2000) found no differences in overall confidence for the two types of interviews. Similarly, Kebbell and Wagstaff (1997) found that average witness confidence was not different for forced-choice responses given in cognitive interview, hypnosis, and control conditions. The results of both of these studies, however, should be interpreted in light of the fact that the cognitive interview did not yield any increases in either the total amount of information or the accuracy of information as compared with standard types of questioning. Other studies have examined the effects of more general context reinstatement techniques on confidence, and the results have been mixed. In an early field study of store clerks who were asked to identify previous customers, mentally reinstating the context of the event did not affect eyewitness confidence, but it did yield substantially higher identification rates for target-present photographic lineups (Krafka & Penrod, 1985). Smith and Vela (1992) found that mentally reinstating the context of a target event did not affect either eyewitness accuracy or confidence, as compared with a control condition, for witnesses who were asked to choose a face from a photospread. However, actual physical reinstatement of the context (i.e., returning to the scene of the target event) did lead to higher levels of both eyewitness accuracy and eyewitness confidence. Hypnosis. During the past several decades, police investigators have occasionally used hypnosis in an attempt to aid the memory of eyewitnesses (Kebbell & Wagstaff,

13.

EYEWITNESS CONFIDENCE

379

1998). In general, the application of hypnosis in forensic settings involves two steps. First, a hypnotist induces a hypnotic state in the witness. Second, the hypnotist or an investigator uses one or more memory-enhancing techniques, such as context reinstatement, to encourage the witness to recall as much as she or he can about the target event. For more detail about the methods and techniques of hypnosis, the reader is referred to Chapter 11 of this volume (Mazzoni & Lynn). Much of the questioning of hypnotized witnesses involves some sort of context reinstatement process and is similar in many respects to the cognitive interview and other related interviewing techniques. As with other context reinstatement techniques, hypnotized witnesses sometimes recall more total information than nonhypnotized witnesses (Dywan & Bowers, 1983), but the memories of hypnotized witnesses are not usually more accurate than witnesses who have not been hypnotized (for a review see Kebbell & Wagstaff, 1998). Experimental studies have consistently found that hypnotized witnesses tend to have high levels of confidence in both correct and incorrect memory reports (e.g., Fligstein, Barabasz, Barabasz, Trevisan, & Warner, 1998). Spanos, Quigley, Gwynn, Glatt, and Perlini (1991) found that hypnotized participants were much more confident in a mug shot identification task, particularly concerning misidentifications, than were control participants, even though the recognition performance of both groups was no better than chance. In their meta-analysis of experimental studies of hypnotically refreshed testimony, Steblay and Bothwell (1994) found that hypnotized witnesses consistently report greater confidence than nonhypnotized witnesses (cf. Kebbell & Wagstaff, 1997). What do observers think about the apparent confidence exhibited by hypnotized witnesses? Sanders, Gansler, and Reisman (1989) examined how actual police detectives, real attorneys, and mock jurors perceived hypnotized and control participants on a variety of dimensions, including witness confidence. For all three groups of observers, there was no difference in perceived confidence between the hypnotized and control witnesses. Hypnotized witnesses tend to be more susceptible to the effects of misleading questioning because they are more suggestible (e.g., Zelig & Beidleman, 1981). In addition, hypnotized witnesses consistently exhibit higher confidence ratings for both correct and incorrect memory reports after being exposed to misleading information (e.g., Dywan & Bowers, 1983; Sheehan & Tilden, 1983). As in most misinformation studies, Sheehan and Tilden (1983) exposed half of the participants to post-event misleading information about a slide sequence of a theft, and the other half were exposed to neutral information. After reading the post-event information, half of the participants in each information condition were hypnotized and half were not. Consistent with previous research, memory accuracy for the critical items was worse in the misinformation condition than in the neutral-information condition. In addition, hypnosis had no impact on this misinformation effect. More to the point here, however, the hypnotized witnesses were significantly more confident than the nonhypnotized witnesses, despite the fact that they were not more accurate. Even in the absence of hypnosis, misleading information can have a substantial impact on eyewitness confidence. That is the topic of the next section.

380

SHAW, MCCLURE, DYKSTRA

Misleading and Suggestive Questioning. A long line of research, initiated in the 1970s by Elizabeth Loftus (e.g., Loftus, Miller, & Burns, 1978), has demonstrated that misleading or suggestive questioning can affect the accuracy of eyewitness memory. In the typical misinformation paradigm, participant,witnesses view some sort of target event, are then exposed to post,event information that is either consistent or inconsistent with the target event, and are finally questioned about details of the event, includ, ing elements that were and were not subject to the misleading post,event information. Generally speaking, witnesses are less accurate for details that were the subject of mis, leading post, event information than for control items. Additional information about the current state of the research on the misinformation effect can be found in Chapters 7, 8 9, and 10 of this volume. There is only a limited amount of research concerning the effects of post,event mis, information on witness confidence, but most of these studies demonstrate that mislead, ing post,event information can indeed alter witness confidence. After viewing a film, the participant,witnesses in the study by Ryan and Geiselman (1991) were given two memory tests, one before they read a narrative passage about the event in the film and one after they read the passage. The passage contained either correct information (lead, ing), incorrect information (misleading), or no information (neutral). Prior to the pas, sage, eyewitness confidence was higher for correct than for incorrect answers to test questions about the film. After the neutral or leading passages, confidence was still higher for correct than for incorrect answers. However, after the misleading passage, eyewitness confidence was much higher for incorrect than for correct responses. An al, most identical pattern of results was obtained by Tversky and Tuchin (1989). In their control condition, confidence was significantly higher for correct than for incorrect re, sponses to yes-no questions about the critical items. However, in the misled condition, confidence was higher for incorrect responses than for correct responses to questions about items that had appeared in the post,event narrative passage but not in the origi, nal stimulus slides. In the study by Shaw, Garcia, and Robles (1997), participants viewed a videotape of a simulated crime and then read a narrative that contained details that were consistent, inconsistent, or neutral with respect to the details in the videotape. In all three experi, ments the confidence for the responses in the final memory test was significantly higher in the consistent,information condition than in the control condition, and in Experi, ment 3 confidence was also significantly higher in the inconsistent,information condi, tion than in the control condition. Although the witnesses were overconfident in all conditions in all three experiments, the magnitude of the overconfidence was the largest in the inconsistent, informationcondition, which demonstrates the distorting effect misleading information may have on confidence in forensic settings. Some research even suggests that multiple exposures to misleading information may magnify its effects on confidence. For example, Zaragoza and Mitchell (1996) exposed some participants to misleading information three times and others only once. Those in the three,exposure condition were significantly more likely to misremember with high confidence that the suggested information was in the original target event than were

13.

EYEWITNESS CONFIDENCE

381

participants in the single-exposure condition. Moreover, Blagrove and Akehurst (2000) found that the repetition of misleading questions (collapsed across all other conditions) led to a significant increase in confidence for the participants' repeated incorrect responses to the misleading questions. Yet the repetition had no effect on the participants' confidence for correct responses to those questions. When confidence findings in misinformation studies are being interpreted, it is important to consider whether there is an appropriate control group. For example, although Pickel (1999) and Schooler, Gerhard, and Loftus (1986) both reported lower levels of confidence for "false" or "suggested" memories, neither study included a suitable control group in which there was no post-event information. Therefore, it is not possible to know whether the witnesses' confidence ratings for the false memories in the post-event misinformation condition would have been different from the ratings for false memories by witnesses who were not exposed to any postevent information. The studies reviewed in this section thus far have concerned witness confidence for responses to memory questions given sometime after the introduction of the misleading postevent information. Only a few studies have examined confidence for immediate responses to misleading or suggestive questions. Roebers (2002) found that the effect of biased (misleading) questions on witnesses' immediate confidence may be moderated by the age of the witnesses. The confidence for the unbiased questions was higher for correct than for incorrect responses for all of the participants, and this same pattern held for the misleading questions for the adults. However, for the 8- and 10-year-old participants, there was no significant difference in confidence between correct and incorrect responses to misleading questions. In a questionnaire study about details from the televised trial of 0. J. Simpson (Shaw, Garven, & Wood, 1997), participants were exposed to three types of questions: leading questions that suggested a correct response, misleading questions that suggested an incorrect response, and neutral questions. In their immediate responses to the questions that suggested an incorrect response, the participants were less likely to be correct and exhibited less confidence than they did for the other two question types. Although it is difficult to generalize from these last two studies, these results suggest that, at least for adults, the effects of misleading information on confidence are more likely to appear during a later memory test than immediately after the exposure to the misinformation. From a forensic perspective, this implies that post-event misinformation that occurs sometime before trial may ultimately alter a witness's confidence many months later when that witness finally testifies in court. This possibility is addressed in the last section of this chapter.

Repeated Questioning In most criminal and civil cases, witnesses are questioned many times between the witnessed event and trial by a variety of people, including arresting officers, police detectives, district attorneys, defense attorneys, other witnesses, and even friends and family. This questioning can take many forms and occur in many forums. In his early review of

382

SHAW, MCCLURE, DYKSTRA

the eyewitness confidence literature, Leippe (1980) suggested that repeated police interrogation may affect eyewitness confidence, and subsequent research has supported this proposition. In a series of studies involving eyewitness memory for a variety of target events (e.g., common household items in a slide presentation [Shaw, 1996], details about a realistic staged classroom interruption [Shaw & McClure, 1996], and faces presented in a timed sequence by a computer [Shaw, McClure, & Wilkens, 2001]), Shaw and his colleagues have shown that repeated questioning of eyewitnesses consistently leads to increases in witness confidence without corresponding changes in witness accuracy. It is also interesting to note that in all three of these studies (with the exception of Experiment 3 in Shaw et al., 2001) the repeated questioning did not affect the accuracy of the eyewitness reports. This indicates that simply repeating the same questions over and over to witnesses, particularly if the questions call for recognition-type judgments, does not necessarily lead to better witness memory. Shaw and McClure (1996) demonstrated that the increases in witness confidence due to repeated questioning can occur on both a between-participant and withinparticipant basis. Witnesses who were repeatedly questioned about a target event reported higher final confidence ratings than witnesses who were questioned only once. Moreover, witnesses who were repeatedly questioned about only a portion of an event (e.g., one of two perpetrators) showed higher levels of confidence for details about that practiced portion in a final memory test than they did for other previously unquestioned details. In addition, Shaw has demonstrated that just asking witnesses to think repeatedly about their responses to forced-choice questions can lead to later increases in witness confidence. In Experiment 1 of Shaw (1996), half of the participants were asked toreflect on their previous answers to forced-choice questions about a target event three times over a 20-min retention interval. When the participants were given the same forced-choice memory test at the end of the study, those participants who had engaged in the three sessions of reflective thought were significantly more confident about their responses in the final memory test than were participants in the control condition (who did not engage in any reflective thought). This occurred despite the fact that the accuracy of the participants was the same in the two sets of conditions. Wells and Bradfield (1999) found similar results in a study in which some of the participants were asked to think about their previous choice from a lineup (which in all cases was incorrect). Those participants who thought about their lineup choice were more confident in their incorrect choice later on than were participants who did not engage in any reflective thought. Shaw's results concerning the effect of repeated questioning on confidence are also consistent with a study by Turtle and Yuille (1994) in which the participants repeatedly answered questions about details of a videotaped burglary. Turtle and Yuille collected confidence ratings for both open-ended and cued-recall questions. They found that the final confidence ratings for both types of questions, obtained 3 weeks after the videotape was watched, were higher for the participants who engaged in repeated recall than for participants who only recalled the event once 3 weeks later. However, in contrast to

13.

EYEWITNESS CONFIDENCE

383

Shaw's studies, Turtle and Yuille found that the accuracy of the cued~recall questions was greater for the repeated~recall participants. This result was likely due to the fact that the participants were instructed in the basic elements of the cognitive interview. Repeated questioning can also affect confidence when it includes suggestive or mis~ leading information. As mentioned previously, Zaragoza and Mitchell (1996) found that participants who were exposed to misleading information three times were much more likely to make attribution errors for the suggested details (which had not appeared in the original target event) with high confidence than were participants who were only ex~ posed to the misinformation once. In contrast to the research discussed so far, Granhag and his colleagues (Granhag, 1997; Granhag, Stromwall, &Allwood, 2000) have found that repeating confidence rat~ ings may have no effect on confidence or may even lead to a decrease in witness confi~ dence. In a study by Granhag (1997), the participants provided confidence judgments about their responses to forced~choice questions about a videotaped event immediately after answering the questions and then again 1 week later. Compared with participants who gave only one confidence rating, Granhag found that participants who repeated their confidence ratings were less overconfident in the final session (although there was no statistically significant difference between the two groups as to the confidence ratings themselves). This basic finding was replicated in a later study by Granhag et al. (2000). In addition, Granhag et al. found that witness confidence was significantly lower in the repeated~ confidence condition for both correct and incorrect responses in Experiment 1 and for correct responses in Experiment 2. Two other aspects of Granhag's studies bear mentioning here. First, when Granhag et al. asked the witnesses in Experiment 2 to simply recall their original confidence rat~ ings during the second questioning session (instead of providing new ratings), the confi~ dence ratings increased significantly for both correct and incorrect responses. Second, the witnesses were overconfident in almost all conditions in both studies, although the magnitude of the overconfidence was reduced when the participants were asked to give a second confidence rating in the repeated~confidence condition. An important procedural distinction makes it easy to reconcile the observed decrease in confidence after repeated questioning in the Granhag studies with the con~ sistent increase in confidence in Shaw's experiments. In the repeated~questioning condi~ tion in Shaw's studies (Shaw, 1996; Shaw & McClure, 1996; Shaw, McClure, & Wilkens, 2001), the participants repeatedly answered the same questions about the target event. In Granhag's experiments, the participants answered the forced~choice questions only once, but they were asked to provide their confidence ratings twice. In the second questioning session they were shown their earlier response and then asked how confident they were that their earlier response was correct. This distinction suggests that it is the repeated answering of the memory questions (or repeated reflective thought) that has led to the con~ sistent increases in confidence in Shaw's studies. Shaw has proposed that witnesses base their confidence judgments in part on the ease with which they are able to access the target memory (i.e., the response fluency of the memory report). Furthermore, he posits that the repeated questioning of a witness

384

SHAW, MCCLURE, DYKSTRA

can lead to an increase in response fluency, which in turn leads to an increase in confidence. Several empirical studies are consistent with the response-fluency hypothesis (Robinson, Johnson, & Herndon, 1997; Shaw et al., 2001; cf., Robinson, Johnson, & Robertson, 2000). Whatever the mechanism, it seems clear that repeatedly questioning a witness can lead to increases in witness confidence, even when the questioning does not involve any misleading or inconsistent information. From a forensic point of view this is an important finding, because most witnesses are questioned repeatedly about the target event before trial. Such repeated questioning may cause witnesses to be more confident when they are testifying at trial than if they had not been questioned previously, even though they are no more accurate.

Co-witness Information and Feedback During the long wait between the witnessed event and trial, witnesses are often exposed to other information about the target event. This information may come from a variety of sources, including media reports, other witnesses, and even police investigators. One source of additional information, in the form of suggestive or misleading questioning, has already been reviewed. Two other types of post-event information-co-witness information and witness feedback-are discussed in this section.

Co-witness Information. There are often multiple witnesses to a crime. A witness can be exposed to information from other witnesses in a variety of settings-during group interviews, lineup presentations, informal discussions with other witnesses, or even through co-witness information supplied by the police. Despite the many opportunities for witnesses to be affected by co-witness information, only a few studies have examined the effect of such information on eyewitness confidence. In Luus and Wells's (1994b) seminal study on co-witness information, two participants witnessed a staged theft and were then asked to identify the thief from a photographic lineup. After each participant made her or his lineup choice privately, the experimenter provided bogus information concerning the lineup choice of the other witness. When the co-witness information was consistent with the lineup choice of the witness, it served to inflate the witness's confidence; when the co-witness information was inconsistent (e.g., the co-witness picked someone else from the photographic lineup), the witness's confidence decreased significantly. This initial effect of co-witness information generally persevered, even when the co-witness information was later changed or discredited. All of these effects occurred independently of any differences in witness accuracy, because all participants viewed a perpetrator-absent lineup, ensuring that all of their identification decisions would be incorrect. A few researchers have assessed the effect of co-witness information by having pairs of witnesses collaborate in answering questions about a target event. For example, participants in Stephenson and Wagner's (1989) study who worked together in pairs on a written recall task exhibited more confidence in both correct and incorrect responses than did participants who worked alone. This result implies that witnesses should be

13.

EYEWITNESS CONFIDENCE

385

questioned individually to minimize the possibility of increased confidence caused by the effects of collaboration. In a different sort of joint-recall procedure, pairs of participants in Experiment 2 of a study by Wright, Self, and Justice (2000) answered questions about a theft presented via pictures in a storybook format. Although they thought that they had viewed the exact same theft, the two participants saw two versions that differed in a critical detail. After viewing the presentation, they individually answered true-false questions about the theft. Then the two witnesses discussed the target event with each other. Finally, they individually answered the same true-false questions as before. The mean confidence ratings for the final set of individual questions tended to be different from the first set, presumably as a result of the group discussion. Although the direction of the change in confidence was dependent on certain artifacts of the stimulus materialsfor some groups mean confidence increased after discussion and for other groups it decreased-the general message is clear. The group discussion altered the confidence of the individual witnesses.

Witness Feedback. Because there are no legal constraints on whether the police can give feedback, witnesses are frequently told whether they correctly chose the suspect from a lineup (Wells, 1993). In a series of studies, Wells and his colleagues have examined the effect of feedback on witness confidence (Bradfield, Wells, & Olson, 2002; Wells & Bradfield, 1998, 1999; Wells, Olson, & Charman, 2003). All of these studies have involved the same general paradigm. After viewing a videotape of a crime, participants are asked to choose the culprit from a photospread. After making the identification, the participants are given one of several types of bogus feedback about the correctness of their choice, typically either confirming feedback ("good, you identified the suspect"), disconfirming feedback ("actually, the suspect is number _"), or no feedback. Then the participants are asked, either immediately or after some delay, to indicate their confidence in their lineup choices. A consistent finding across all of these studies is that confirming feedback leads to inflated witness confidence as compared with disconfirming feedback and no feedback. However, disconfirming feedback generally does not reduce confidence relative to the no-feedback condition (Wells & Bradfield, 1998; Wells, Olson, & Charman, 2003). The inflating effect of confirming feedback is generally robust, as delaying either the feedback or the confidence measure for 48 hours does not diminish the effect (Wells et al., 2003). However, it is possible to reduce or eliminate the feedback effect by asking the witnesses about their confidence prior to the feedback manipulation (Wells & Bradfield, 1998) or by asking the witnesses to reflect on their lineup choice before they are given feedback (Wells & Bradfield, 1999). In a study in which the participants were asked to confabulate answers about a scene from a Disney movie, Zaragoza, Payment, Ackil, Drivdahl, and Beck (2001) also found that confirming feedback can inflate witness confidence. After the participants made up their responses about the movie, an interviewer gave confirmatory feedback about the confabulated answers to some participants and neutral feedback to others. Not only did the confirmatory feedback increase the proportion of false memories later on, but it also

386

SHAW, MCCLURE, DYKSTRA

led to significantly higher witness confidence for the false memories than in the neutral feedback condition. It appears that the effect of confirmatory feedback is greater for inaccurate memory reports than for accurate ones (Bradfield, Wells, & Olson, 2002). In other words, confirmatory feedback leads to bigger increases in confidence for incorrect responses in a memory task (such as picking the wrong person from a lineup) than for correct responses. This finding is particularly problematic for the justice system, because jurors may place an inordinate amount of weight on inaccurate memory reports simply because those memories are reported with a high degree of confidence.

Summary From the considerable body of experimental evidence collected over the past 25 years, it is clear that there are many opportunities for eyewitness confidence to be altered during the period of time between the witnessed event and trial. This is why researchers consistently suggest that evidence should be collected from witnesses as early as possible in the legal process. Based upon recommendations of a team of psychologists, lawyers, and police officers, the Department of Justice published "Eyewitness Evidence: A Guide for Law Enforcement" in 1999, which established the first national guidelines concerning the collection and preservation of eyewitness evidence. One of the central recommendations of this report is that eyewitness confidence should be assessed as early as possible, for example, when a witness makes an identification from a lineup (for a discussion of the genesis of these guidelines, see Wells et al., 2000). In addition, it would not be surprising to find that eyewitness confidence affects many police decisions during the investigative process, including what investigative leads to follow, whether to rule out alternative suspects, and how to determine the credibility and reliability of witnesses during an interview. Given that there is no empirical research bearing directly on this issue, future studies should address this potentially important consequence of the malleability of eyewitness confidence.

FINAL TESTIMONY AT TRIAL

Testifying in court seems simple enough. In their oath, witnesses typically promise to "tell the truth, the whole truth, and nothing but the truth." Of course, in reality a witness's task is much more complex. Witnesses must attend to a host of cognitive and social demands while testifying about matters of great importance, often with people's lives or livelihoods on the line. They must speak in public, which is a novel experience for many lay witnesses. Questions from the attorneys can be complex and even misleading, and testimony is often interrupted by legal objections and bench conferences. During all of this activity, jurors and judges closely scrutinize every word and movement of the testifying witnesses. Despite the potential for courtroom factors to affect witness confidence, relatively few studies have examined eyewitness confidence in this final stage of the eyewitness ex-

13.

EYEWITNESS CONFIDENCE

387

perience. Of course, much of the research reviewed in the section on memory-related events that occur between the witnessed event and trial (e.g., misleading post-event information, hypnosis, the cognitive interview, repeated questioning, co-witness information, and witness feedback) is relevant here. In many of those studies, the ultimate effects on eyewitness confidence do not appear until a final memory test that is analogous to a witness's testimony at trial. For example, in the repeated-questioning studies, the highest confidence ratings occur during the final memory test (Shaw, 1996; Shaw & McClure, 1996; Shaw, McClure, & Wilkens, 2001). A witness who is questioned many times by the police during an investigation is likely to express significantly higher levels of confidence at trial than a witness who is not repeatedly questioned. Similarly, when witnesses are exposed to misleading information after viewing a target event, the effects on eyewitness confidence usually do not appear until a final memory test that occurs sometime after the exposure to the misleading information (e.g., Ryan & Geiselman, 1991; Tversky & Tuchin, 1989). There are also specific aspects of the courtroom experience that can potentially affect eyewitness confidence while the witness is on the witness stand. Some of these factors are examined in the following two sections.

Social Demands on Eyewitnesses in the Courtroom When testifying in court, eyewitnesses are subjected to numerous social pressures that may affect their confidence levels. Two sources of social influence are (a) the public nature of courtroom testimony and (b) the pressure to work hard to "get it right" on the stand. Testifying in Public: Self-presentation Concerns. In court, witnesses testify in the presence of jurors, judges, attorneys, family, friends, members of the general public, and sometimes even representatives of the press. Despite the public nature of courtroom testimony, few eyewitness studies have collected public confidence ratings. In most experiments, researchers have utilized anonymous, written data collection methods. Of those studies that have collected public measures of witness confidence, most of them (e.g., Brigham, Maass, Snyder, & Spaulding, 1982; Luus & Wells, 1994b) have not compared public and private confidence ratings directly. In a recent series of studies, however, Shaw and his colleagues have demonstrated that confidence ratings shared openly with others can differ from ratings that are given privately and anonymously (Shaw, Appio, Zerr, & Pontoski, 2006; Shaw, Woythaler, & Zerr, 2001; Shaw, Zerr, & Woythaler, 2001). Shaw has proposed that eyewitnesses may use self-presentation strategies in order to manage the impressions they make when testifying in court and that they may adjust their confidence judgments on the stand as an impression-management tool. Thus, when contradiction by another witness is unlikely, a witness may inflate her or his public confidence ratings to impress others. On the other hand, if there is a possibility that her or his testimony will be inconsistent with other evidence (e.g., the accounts of other eyewitnesses), a witness may be more cautious and adjust her or his confidence ratings downward in order to avoid the possibility of being overconfident about inaccurate testimony.

388

SHAW, MCCLURE, DYKSTRA

In studies by Shaw, Woythaler, and Zerr (2001, Experiment 1) and Shaw, Zerr, and Woythaler (2001), the participants were placed in a multiple-witness situation in which all of the participant-witnesses in each experimental session knew whether each of the participants' public memory reports were accurate. These two studies examined the prediction that the participants' need to play it safe would result in lower confidence ratings in public than in private. In Shaw, Woythaler, and Zerr, the memory task involved recognizing faces as old or new, whereas in Shaw, Zerr, and Woythaler, the memory questions were about the details of a simulated robbery. As expected from the self-presentation hypothesis, in both studies the confidence ratings shared publicly with the other participants were lower than those ratings that remained private, yet there was no difference in the accuracy of the responses. One can infer from these results that the participants in the public condition wanted to avoid appearing overconfident in front of their fellow participant-witnesses. In contrast, Shaw, Appio, Zerr, and Pontoski (2006) created a situation in which the participants did not have to worry about whether the other witnesses could assess the accuracy of their responses. When each participant testified aloud during a face recognition test, the other participants had no way of knowing whether the responses were accurate. The results were again consistent with a self-presentation explanation. Because the participants were not worried about being contradicted by other witnesses, they gave higher confidence ratings for their publicly shared responses than for their private responses. Presumably, this was done in an effort to influence the other participants' evaluations of their testimony. Once again, response accuracy was the same in the private and public conditions.

Working Hard to "Get It Right" on the Witness Stand. It is safe to assume that when testifying in court, eyewitnesses generally strive to do their best to testify truthfully, completely, and accurately because they are aware of the importance of their testimony. In addition, they may also be motivated to try hard for more selfish reasons, such as impressing others with their witnessing and testifying skills. Although witnesses certainly want to do their best during all phases of the witnessing experience, nowhere are the social pressures to perform as great as they are on the witness stand. How might such pressures affect eyewitness confidence? Shaw, Woythaler, and Zerr (2001) found that extra effort during a final memory test may lead to higher confidence ratings. The participant-witnesses were asked during a face-recognition test to try harder on half of the test items (because they were the ones that counted in the experiment) than on the other half of the items (which were supposedly included as part of a pretest for a future study). In two experiments, the participant-witnesses were more confident in their responses for items that counted than for items that did not count, yet response accuracy was the same for the two sets of items. In a study that examined the effects of a more selfish reason to try hard, Shaw and Zerr (2003) found that extra effort can be associated with increases in witness confidence. The participants answered several multiple-choice questions about a classroom visitor in an effort to win a prize for the most correct responses. In addition to providing confidence ratings, the participants also reported how much effort they expended in answering the questions. The confidence ratings for those participants who reported ex-

13.

EYEWITNESS CONFIDENCE

389

pending high levels of effort were significantly higher than the confidence ratings for the low-effort participants, despite the fact that response accuracy did not differ as a function of reported effort. Whether the motivation to "get it right" on the witness stand stems from concerns about others (Shaw, Woythaler, & Zerr, 2001) or more selfish considerations (Shaw & Zerr, 2003), the extra effort may be associated with increases in witness confidence without any accompanying changes in witness accuracy. Shaw has proposed that this effect may be due to changes in a witness's self-perception caused by the extra effort-"I'm trying extra hard to testify accurately on the witness stand. Therefore, I'm likely to 'get it right,' so I'm going to say that I'm very confident." However, this explanation has not yet been formally tested.

cognitive Demands on Eyewitnesses in the Courtroom Direct examination and cross-examination of witnesses can be very demanding on their cognitive resources. On the stand, witnesses must answer rapid-fire questioning that can be confusing and even hostile. They also have to attend to complicated legal maneuvering and frequent interruptions, including objections from counsel and legal rulings from the judge. These high cognitive demands leave witnesses little time to engage in careful, reflective thought concerning their in-court confidence judgments. Testifying witnesses typically face complex, confusing questions on the witness stand. Kebbell and his colleagues (Kebbell & Giles, 2000; Kebbell & Johnson, 2000) have examined whether the format of the questions posed to witnesses can affect their reported confidence levels. In both studies, half of the participants were exposed to a "confusing" set of questions, which included leading questions, questions with complex vocabulary or syntax, questions with double negatives, and questions that contained multiple parts. The other participants were exposed to "simplified versions" of the same questions. In Kebbell and Giles (2000), the witnesses reported significantly higher confidence ratings for the confusing questions than for the simple questions, despite the fact that their responses to the confusing questions were actually less accurate. Moreover, although the confidence ratings were higher for correct responses than for incorrect responses for the simple versions of the questions, for the confusing questions the level of confidence was the same for the correct and incorrect responses. The pattern of results in Kebbell and Johnson (2000) was very similar. Though the participants were less accurate in their responses to the confusing questions, overall confidence trended higher for the confusing questions (although not significantly so, as was the case in Kebbell and Giles). In addition, confidence was significantly higher for incorrect responses to the confusing questions as compared with the simplified questions, but there was no such difference for the correct responses. Taken together, the results of these two studies imply that the confusing, complex questions typically asked of witnesses on the stand may lead to higher confidence ratings, particularly for incorrect responses, as compared with the simple types of questions usually asked in other settings. Another cognitive demand on testifying witnesses stems from the fact that witnesses must often answer questions very quickly in court, particularly during cross-examination.

SHAW, MCCLURE, DYKSTRA

390

In such situations, witnesses are not able to spend much time or effort on controlled, reflective thought concerning their confidence ratings, but instead must rely on automatic processes. One such process is the assessment of response fluency, which has been described previously. Whether limiting processing resources results in higher or lower confidence ratings depends on the nature of the automatic and controlled processes. For example, in a situation where controlled processing (such as reflective thought) would cause a witness to second-guess herself and lower her confidence, limiting processing resources might prevent this decrease in confidence. Conversely, if reflective thought would yield confirming evidence that would cause a witness to increase her confidence, limiting that witness's processing resources might prevent that witness's confidence from rising. In two studies, Shaw and his colleagues limited the time that participants had to respond to questions in a face-recognition task (Shaw, McClure, & Wilkens, 2001; Shaw, Zerr, & Woythaler, 2001 [Experiment 2]). In both studies, the confidence ratings were generally lower in the fast or speeded condition (as compared with the "normal" control condition) for questions that had a correct response. Yet the confidence ratings were higher (Shaw, Zerr, & Woythaler, 2001) or the same (Shaw, McClure, & Wilkens, 2001) in the fast or speeded condition for questions that did not have a correct response (i.e., all choices were incorrect).

Summary Based on the few studies that have examined eyewitness confidence in the courtroom, it seems that the social and cognitive pressures inherent in testifying may result in complex effects on the confidence levels reported by testifying witnesses on the stand. In addition, there are likely many other sources of social or cognitive pressures that have not yet been studied. One example of this is that witnesses usually testify in court under conditions of divided attention, because they must listen to objections from the attorneys and rulings by the judge, as well as engage in impression management, all while trying to answer questions on direct examination and cross-examination. Virtually all of the studies on courtroom pressures suffer from poor external validity, so it is impossible at this point to know how well these findings generalize to real courtrooms. Clearly this final stage of the eyewitness experience has not received adequate attention in the literature, and it is an area ripe for further empirical investigation.

CONCLUDING COMMENTS

Although one must be very cautious about drawing sweeping conclusions based upon a body of evidence that is constantly growing and evolving, several general principles emerge from the research reviewed in this chapter: • •

Jurors place a great deal of weight on eyewitness confidence. It is generally accepted that eyewitness confidence is not by itself a reliable predictor of eyewitness accuracy.

13.

• • • • • • • • •

EYEWITNESS CONFIDENCE

391

Many factors can affect eyewitness confidence independently of eyewitness accuracy. Eyewitnesses are often overconfident about the accuracy of their memory reports. Although there may be differences in witness confidence due to witness, culprit, and event characteristics, there is currently little empirical evidence of such differences. Misleading post-event information can alter eyewitness confidence, often resulting in inflated confidence for the suggested information. Repeated questioning of witnesses, even in the absence of any misleading information, can lead to increases in eyewitness confidence. Hypnotized witnesses consistently report greater confidence than nonhypnotized witnesses. Co-witness information can alter eyewitness confidence independently of any changes in eyewitness accuracy. Feedback that confirms a witness's account leads to inflated witness confidence, and the effect is particularly large for inaccurate memory reports. Social and cognitive pressures in the courtroom may affect eyewitness confidence as witnesses testify, but the effects are complex and not yet well understood.

Based upon these empirically supported findings, we propose three general recommendations concerning the collection and use of eyewitness confidence: 1. 2.

3.

Eyewitness confidence should be assessed as early as possible in the legal process. All members of the legal and judicial systems, particularly police officers, investigators, and attorneys, should be sensitive to (a) the fact that eyewitness confidence is often not a good indicator of accuracy and (b) the many factors that can alter eyewitness confidence from the initial event through trial. Judges should instruct jurors about the factors that can affect witness confidence, particularly those relevant to the case being tried.

Despite the accumulated evidence that eyewitness confidence is malleable across the witnessing experience, some may still argue, consistent with the rulings of the United States Supreme Court in Neil v. Biggers (1972/1979) and Manson v. Braithwaite (1977), that eyewitness confidence should be considered by jurors and judges as a reliable indicator of eyewitness accuracy. Such an argument would likely be predicated on the many studies that have reported higher confidence levels for correct memory reports than for incorrect memory reports (e.g., Granhag, Stromwell, & Allwood, 2000; Kebbell & Giles, 2000; Kebbell & Johnson, 2000; Kebbell & Wagstaff, 1997; Kebbell, Wagstaff, & Covey, 1996; Roebers, 2002; Robinson, Johnson, & Herndon, 1997; Shaw, 1996; Shaw & McClure, 1996; Stephenson & Wagner, 1989). Based on these findings, one might be tempted to conclude that eyewitness confidence can be a good predictor of eyewitness accuracy. However, when one examines the evidence more closely, there are four compelling reasons why eyewitness confidence is not a forensically useful predictor of eyewitness accuracy. First, the overall difference between confidence for correct and incorrect responses may mask important differences in confidence that occur in various conditions. For

392

SHAW, MCCLURE, DYKSTRA

example, Kebbell and Giles (2000) found that although the confidence ratings were higher for correct responses than for incorrect responses for simple (control) types of questions, the levels of confidence for the correct and incorrect responses were the same for the confusing questions. In Shaw and McClure (1996), the final confidence ratings were higher for correct responses than incorrect responses when the participants engaged in post-event questioning, but that difference disappeared for the participants in the no-post-event-questioning group. Furthermore, as demonstrated by Ryan and Geiselman (1991), misleading post-event information can actually lead to higher eyewitness confidence for incorrect responses than for correct responses. Further complicating the matter, some studies have found that some confidence-altering factors have a larger effect on confidence for inaccurate memory reports than for accurate reports. For example, Bradfield, Wells, and Olson (2002) demonstrated that the inflating effect of confirmatory feedback on witness confidence might be greater for inaccurate memories than for accurate ones. As is evident from all of these results, the statement that confidence is generally higher for correct memory reports than for incorrect reports masks the more complex patterns of results that are often obtained. Second, in real-life cases one does not usually have comparative information about "correct" and "incorrect" responses. If a witness says she is "pretty sure" that the defendant is the man who robbed her at gunpoint, it is impossible to assess that confidence judgment in relation to confidence judgments for other identifications. Because jurors and judges do not have access to a witness's confidence judgments for correct and incorrect memory reports, the comparative levels of confidence commonly reported in laboratory studies do not yield useful information about real-world eyewitness situations. In the future, it may be useful for researchers to attempt to develop a standardized measure of confidence that could be used in all types of settings. Third, a consistent finding in experimental studies is that eyewitnesses are generally overconfident (e.g., Granhag, 1997; Granhag, Stromwall, & Allwood, 2000; Roebers, 2002; Shaw, Garcia, & Robles, 1997; Shaw & McClure, 1996). Thus, even when confidence is higher for correct than for incorrect memories, witness confidence may still be too high for the inaccurate memory reports. This finding implies, of course, that if jurors rely on witness confidence as an indicator of witness accuracy, they will give undue weight to all eyewitness memory reports, even those that are inaccurate. This is particularly problematic because eyewitness overconfidence can be enhanced even further by factors such as misleading or inconsistent post-event information (e.g., Shaw, Garcia, & Robles, 1997). Finally, even if eyewitness confidence is sometimes a "good" predictor of eyewitness accuracy, the magnitude of the confidence-accuracy correlation almost always falls well short of being forensically useful, particularly in criminal cases where the standard of proof is beyond a reasonable doubt. Although higher correlations have sometimes been reported, most researchers agree that the magnitude of the confidence-accuracy correlation is quite small, probably about r = .25 (e.g., Bothwell, Deffenbacher, & Brigham, 1987), which means that only about 6% of the variance in eyewitness accuracy is accounted for by variations in eyewitness confidence. Even if we take one of the largest confidence-accuracy correlations reported in the literature (r = .59; Lindsay, Read, &

13.

EYEWITNESS CONFIDENCE

393

Sharma, 1998), that still means that confidence accounts for less than 35% of the vari~ ance in eyewitness accuracy. Given the high standard of proof in criminal cases, even moderately large confidence-accuracy correlations confer limited forensic utility. Based on all of these considerations, the weight of the evidence still indicates that eyewitness confidence is not by itself a reliable predictor of eyewitness accuracy. Two other important conclusions emerge from the research reviewed in this chapter. First, it is clear that eyewitness confidence is malleable and can be affected independently from eyewitness accuracy across the entire eyewitness experience. Second, we still have much to learn about the factors that can alter eyewitness confidence and about how eyewitness confidence can affect participants in the legal system, such as police investigators, other witnesses, judges, and jurors.

REFERENCES Adams-Price, C. (1992). Eyewitness memory and aging: Predictions of accuracy in recall and person recognition. Psychology & Aging, 7, 602-608. Alley, T. R., & Schultheis,]. A. (2001). Is facial skin tone sufficient to produce a cross-racial identification effect? Perceptual & Motor Skills, 92, 1191-1198. Behrman, B. W., & Davey, S. L. (2001). Eyewitness identification in actual criminal cases: An archival analysis. Law & Human Behavior, 25,475-491. Bjork, R. A., & Richardson-Klavehn, A. (1989). On the puzzling relationship between environmental context and human memory. In C. lzawa (Ed.), Current issues in cognitive processes. Hillsdale, NJ: Lawrence Erlbaum Associates. Blagrove, M., & Akehurst, L. (2000). Effects of sleep loss on confidence-accuracy relationships for reasoning and eyewitness memory. Journal of Experimental Psychology: Applied, 6, 59-73. Bothwell, R. K., Deffenbacher, K. A., & Brigham, J. C. (1987). Correlation of eyewitness accuracy and confidence: Optimality hypothesis revisited. Journal of Applied Psychology, 72, 691-695. Bradfield, A. L., Wells, G. L., & Olson, E. A. (2002). The damaging effect of confirming feedback on the relation between eyewitness certainty and identification accuracy. Journal of Applied Psychology, 87, 112-120. Brigham,]. C. (1988). Is witness confidence helpful in judging eyewitness accuracy? In M. M. Gruneberg & P. E. Morris (Eds.), Practical aspects of memory: Current research and issues (Vol. 1, pp. 77-82). Oxford, England: John Wiley & Sons. Brigham, J. C. (1990). Target person distinctiveness and attractiveness as moderator variables in the confidence accuracy relationship in eyewitness identifications. Basic & Applied Social Psychology, 11, 101-115. Brigham, J. C., & Bothwell (1983). The ability of prospective jurors to estimate the accuracy of eyewitness identifications. Law & Human Behavior, 7, 19-30. Brigham, J. C., Maass, A., Martinez, D., & Whittenberger, G. (1983). The effect of arousal on facial recognition. Basic & Applied Social Psychology, 4, 279-293. Brigham, J. C., Maass, A., Snyder, L. D., &Spaulding, K. (1982). Accuracy of eyewitness identification in a field setting. Journal of Personality & Social Psychology, 42, 673-681. Brigham, J. C., & Wolfskeil, M.P. (1983). Opinions of attorneys and law enforcement personnel on the accuracy of eyewitness identification. Law and Human Behavior, 7, 337-349. Brown, E., Deffenbacher, K., & Sturgill, W. (1977). Memory for faces and the circumstances of encounter. Journal of Applied Psychology, 62, 311-318. Clifford, B. R., & Hollin, C. R. (1981). Effects of the type of incident and the number of perpetrators on eyewitness memory. Journal of Applied Psychology, 66, 364-370.

394

SHAW, MCCLURE, DYKSTRA

Cutler, B. L., Penrod, S.D., & Dexter, H. R. (1990). Juror sensitivity to eyewitness identification evidence. Law and Human Behavior, 14, 185-191. Cutler, B. L., Penrod, S. D., & Martens, T. K. (1987a). Improving the reliability of eyewitness identification: Putting context into context. Journal of Applied Psychology, 72, 629-637. Cutler, B. L., Penrod, S. D., & Martens, T. K. (1987b). The reliability of eyewitness identifications: The role of system and estimator variables. Law & Human Behavior, 11, 233-258. Cutler, B. L., Penrod, S.D., & Stuve, T. E. (1988). Juror decision making in eyewitness identification cases. Law & Human Behavior, 12, 41-55. Dahl, L. C. (2003). Investigating investigators: How witness identifications and other evidence influence investigators. Unpublished master's thesis. University of Victoria, Victoria, British Columbia, Canada. Deffenbacher, K. A. (1980). Eyewitness accuracy and confidence: Can we infer anything about their relationship? Law & Human Behavior, 4, 243-260. Deffenbacher, K. A. (1991). A maturing of research on the behaviour of eyewitnesses. Applied Cognitive Psychology, 5, 377-402. Deffenbacher, K. A., & Loftus, E. F. (1982). Do jurors share a common understanding concerning eyewitness behavior? Law and Human Behavior, 6, 15-30. Dywan, J., & Bowers, K. (1983). The use of hypnosis to enhance recall. Science, 222, 184-185. Erskine, A., Markham, R., & Howie, P. (2001). Children's script-based inferences: Implications for eyewitness testimony. Cognitive Development, 16, 871-887. Fisher, R. P., McCauley, M. R., & Geiselman, R. E. (1994). Improving eyewitness testimony with the cognitive interview. In D. F. Ross, J.D. Read, & M.P. Toglia (Eds.), Adult eyewitness testimony (pp. 245-269). New York: Cambridge University Press. Fligstein, D., Barabasz, A, Barabasz, M., Trevisan, M. S., & Warner, D. (1998). Hypnosis enhances recall memory: A test of forced and non-forced conditions. American Journal of Clinical Hypnosis, 40, 297-305. Geiselman, R. E., Fisher, R. P., MacKinnon, D.P., & Holland, H. L. (1985). Eyewitness memory enhancement in the police interview: Cognitive retrieval mnemonics versus hypnosis. Journal of Applied Psychology, 70,401-412. Granhag, P. A. (1997). Realism in eyewitness confidence as a function of type of event witnessed and repeated recall. Journal of Applied Psychology, 82, 599-613. Granhag, P. A., Stromwall, L. A., & Allwood, C. M. (2000). Effects of reiteration, hindsight bias, and memory on realism in eyewitness confidence. Applied Cognitive Psychology, 14,397-420. Hosch, H. M., & Bothwell, R. K. (1990). Arousal, description and identification accuracy of victims and bystanders. Journal of Social Behavior & Personality, 5, 481-488. Hosch, H. M., & Cooper, D. S. (1982). Victimization as a determinant of eyewitness accuracy. Journal of Applied Psychology, 67, 649-652. Juslin, P., Olsson, N., & Winman, A. (1996). Calibration and diagnosticity of confidence in eyewitness identification: Comments on what can be inferred from the low confidence-accuracy correlation. Journal of Experimental Psychology: Learning, Memory, & Cognition, 22, 1304-1316. Kebbell, M. R., & Giles, D. C. (2000). Some experimental influences of lawyers' complicated questions on eyewitness confidence and accuracy. Journal of Psychology, 134, 129-139. Kebbell, M. R., & Johnson, S. D. (2000). Lawyers' questioning: The effect of confusing questions on witness confidence and accuracy. Law & Human Behavior, 24, 629-641. Kebbell, M. R., & Wagstaff, G. F. (1997). An investigation into the influence of hypnosis on the confidence and accuracy of eyewitness recall. Contemporary Hypnosis, 14, 157-166. Kebbell, M. R., & Wagstaff, G. F. (1998). Hypnotic interviewing: The best way to interview eyewitnesses? Behavioral Sciences & the Law, 16, 115-129. Kebbell, M. R., Wagstaff, G. F., & Covey, J. A. (1996). The influence of item difficulty on the relationship between eyewitness confidence and accuracy. British Journal of Psychology, 87, 653-662.

13.

EYEWITNESS CONFIDENCE

395

Krafka, C., & Penrod, S. (1985). Reinstatement of context in a field experiment on eyewitness identification. ]oumal of Personality & Social Psychology, 49, 58-69. Leippe, M. R. (1980). Effects of integrative memorial and cognitive processes on the correspondence of eyewitness accuracy and confidence. Law & Human Behavior, 4, 261-274. Lindsay, D. S., Read, J.D., & Sharma, K. (1998). Accuracy and confidence in person identification: The relationship is strong when witnessing conditions vary widely. Psychological Science, 9, 215-218. Lindsay, R. C., Wells, G. L., & O'Connor, F.]. (1989). Mock-juror belief of accurate and inaccurate eyewitnesses: A replication and extension. Law & Human Behavior, 13, 333-339. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. ]oumal of Experimental Psychology: Human Learning and Memory, 4, 19-31. Luus, C. A. E., & Wells, G. L. (1994a). Eyewitness identification confidence. In D. F. Ross, J.D. Read, & M. P. Toglia (Eds.), Adult eyewitness testimony (pp. 348-361). New York: Cambridge University Press. Luus, C. A. E., & Wells, G. L. (1994b). The malleability of eyewitness confidence: Co-witness and perseverance effects. ]oumal of Applied Psychology, 79, 714-723. Mansonv. Braithwaite, 432 U.S. 98, 112,97 S.Ct. 2243,2252,53 L.Ed.2d 140 (1977). McMahon, M. (2000). The effect of the enhanced cognitive interview on recall and confidence in elderly adults. Psychiatry, Psychology & Law, 7, 9-32. Meissner, C. A., & Brigham, J. C. (2001). Thirty years of investigating the own-race bias in memory for faces: A meta-analytic review. Psychology, Public Policy, and Law, 7(1), 3-35. Memon, A., & Bull, R. (1991). The cognitive interview: Its origins, empirical support, evaluation, and practical implications. ]oumal of Community and Applied Social Psychology, 1, 291-307. Neil v. Biggers, 409 U.S. 188,34 L.Ed.2d 401 (1972), cert. denied, 444 U.S. 909, 100 S.Ct. 221,62 L.Ed.2d 144 (1979). Olsson, N., Juslin, P., & Winman, A. (1998). Realism of confidence in earwitness versus eyewitness identification. ]oumal of Experimental Psychology: Applied, 4, 101-118. O'Rourke, T. E., Penrod, S. D., Cutler, B. L., & Stuve, T. E. (1989). The external validity of eyewitness identification research: Generalizing across subject populations. Law & Human Behavior, 13, 385-395. Pallier, G. (2003). Gender differences in the self-assessment of accuracy on cognitive tasks. Sex Roles, 48, 265-276 Penrod, S., & Cutler, B. (1995). Witness confidence and witness accuracy: Assessing their forensic relation. Psychology, Public Policy, & Law Special Issue: Witness Memory and Law, 1, 817-845. Pickel, K. L. (1999). Distinguishing eyewitness descriptions of perceived objects from descriptions of imagined objects. Applied Cognitive Psychology, 13, 399-413. Platz, S. ]., & Hosch, H. M. (1988). Cross-racial/ethnic eyewitness identification: A field study. ]oumal of Applied Social Psychology, 18, 972-984. Read, J.D., Lindsay, D. S., & Nicholls, T. (1998). The relation between confidence and accuracy in eyewitness identification studies: Is the conclusion changing? In C. P. Thompson & D.]. Herrmann (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 107-130). Mahwah, NJ: Lawrence Erlbaum Associates. Robinson, M.D., Johnson, J. T., & Herndon, F. (1997). Reaction time and assessments of cognitive effort as predictors of eyewitness memory accuracy and confidence. ]oumal of Applied Psychology, 82, 416-425. Robinson, M. D., Johnson, J. T., & Robertson, D. A. (2000). Process versus content in eyewitness metamemory monitoring. Journal of Experimental Psychology: Applied, 6, 207-221. Roebers, C. M. (2002). Confidence judgments in children's and adult's event recall and suggestibility. Developmental Psychology, 38, 1052-1067.

396

SHAW, MCCLURE, DYKSTRA

Ryan, R. H., & Geiselman, R. E. (1991). Effects of biased information on the relationship between eyewitness confidence and accuracy. Bulletin of the Psychonomic Society, 29, 7-9. Ryan, C. S., Judd, C. M., & Park, B. (1996). Effects of racial stereotypes on judgments of individuals: The moderating role of perceived group variability. Journal of Experimental Social Psychology, 32, 71-103. Sanders, G. S., Gansler, D. A., & Reisman, S. (1989). The effects of hypnosis on eyewitness testimony and reactions to cross-examination. American Journal of Forensic Psychology, 7, 33-60. Schooler, J. W., Gerhard, D., & Loftus, E. F. (1986). Qualities of the unreal. Journal of Experimental Psychology: Learning, Memory, and Cognition, 12, 171-181. Scogin, F., Calhoon, S. K., & D'Errico, M. (1994). Eyewitness confidence and accuracy among three age cohorts. Journal of Applied Gerontology, 13, 172-184. Shaw, J. S., III. (1996). Increases in eyewitness confidence resulting from postevent questioning. Journal of Experimental Psychology: Applied, 2, 126-146. Shaw, J. S., III, Appio, L. M., Zerr, T. K., & Pontoski, K. E. (2006). Social pressures can alter witness confidence. Manuscript submitted for publication. Shaw, J. S., III, Garcia, L.A., & Robles, B. (1997). Cross-language postevent misinformation effects in Spanish-English bilinguals. Journal of Applied Psychology, 82, 889-899. Shaw, J. S., III, Garven, S., & Wood, J. M. (1997). Co-witness information can have immediate effects on eyewitness memory reports. Law & Human Behavior, 21,503-523. Shaw, J. S., & McClure, K. A. (1996). Repeated postevent questioning can lead to elevated levels of eyewitness confidence. Law & Human Behavior, 20, 629-653. Shaw, J. S., III, McClure, K. A., & Wilkens, C. E. (2001). Recognition instructions and recognition practice can alter the confidence-response time relationship. Journal of Applied Psychology, 86, 93-103. Shaw, J. S., III, Woythaler, K. A., & Zerr, T. K. (2001). Courtroom pressures can alter eyewitness confidence. In R. Roesch, R. R. Corrado, & R. J. Dempster (Eds.), Psychology in the courts: International advances in knowledge (pp. 223-240). London: Routledge. Shaw, J. S., III, & Zerr, T. K. (2003). Extra effort during memory retrieval may be associated with increases in eyewitness confidence. Law and Human Behavior, 27, 315-329. Shaw, J. S., III, Zerr, T. K., & Woythaler, K. A. (2001). Public eyewitness confidence ratings can differ from those held privately. Law & Human Behavior, 25, 141-154. Sheehan, P. W., & Tilden, J. (1983). Effects of suggestibility and hypnosis on accurate and distorted retrieval from memory. Journal of Experimental Psychology: Learning, Memory, & Cognition, 9, 283-293. Smith, S. M., & Vela, E. (1992). Environmental context-dependent eyewitness recognition. Applied Cognitive Psychology, 6, 125-139. Spanos, N. P., Quigley, C. A., Gwynn, M. 1., Glatt, R. L., & Perlini, A. H. (1991). Hypnotic interrogation, pretrial preparation, and witness testimony during direct and cross-examination. Law & Human Behavior, 15, 639-653. Sporer, S. L., Penrod, S., Read, D., & Cutler, B. (1995). Choosing, confidence, and accuracy: A meta-analysis of the confidence-accuracy relation in eyewitness identification studies. Psychological Bulletin, 118, 315-3 27. Steblay, N. M., & Bothwell, R. K. (1994). Evidence for hypnotically refreshed testimony: The view from the laboratory. Law & Human Behavior, 18, 635-651. Stephenson, G. M., & Wagner, W. (1989). Origins of the misplaced confidence effect in collaborative recall. Applied Cognitive Psychology, 3, 227-236. Sykes, R.N., Gruneberg, M. M., & Turner, T. (1988). Feelings of certainty in child witnesses. In M. M. Gruneberg & P. E. Morris (Eds.), Practical aspects of memory: Current research and issues (Vol. 1, pp. 128-134). Oxford, England: John Wiley & Sons. Talarico, J. M., & Rubin, D. C. (2003). Confidence, not consistency, characterizes flashbulb memories. Psychological Science, 14, 455-461.

13.

EYEWITNESS CONFIDENCE

397

Turtle, J. W., & Yuille, J. C. (1994). Lost but not forgotten details: Repeated eyewitness recall leads to reminiscence but not hypermnesia. Journal of Applied Psychology, 79, 260-271. Tversky, B., & Tuchin, M. (1989). A reconciliation of the evidence on eyewitness testimony: Comments on McCloskey and Zaragoza. Journal of Experimental Psychology: General, 118, 86-91. Wells, G. L. (1978). Applied eyewitness testimony research: System variables and estimator variables. Journal of Personality and Social Psychology, 36, 1546-1557. Wells, G. L. (1993). What do we know about eyewitness identification? American Psychologist, 48, 553-571. Wells, G. L., & Bradfield, A. L. (1998). "Good, you identified the suspect": Feedback to eyewitnesses distorts their reports of the witnessing experience. Journal of Applied Psychology, 83, 360-376. Wells, G. L., & Bradfield, A. L. (1999). Distortions in eyewitnesses' recollections: Can the postidentification-feedback effect be moderated? Psychological Science, 10, 138-144. Wells, G. L., Lindsay, R. C., & Ferguson, T. J. (1979). Accuracy, confidence, and juror perceptions in eyewitness identification. Journal of Applied Psychology, 64, 440-448. Wells, G. L., Malpass, R. S., Lindsay, R. C. L., Fisher, R. P., Turtle, J. W., & Fulero, S.M. (2000). From the lab to the police station: A successful application of eyewitness research. American Psychologist, 55, 581-598. Wells, G. L., Olson, E. A., & Charman, S.D. (2003). Distorted retrospective eyewitness reports as functions of feedback and delay. Journal of Experimental Psychology: Applied, 9, 42-52. Wright, D. B., Self, G., & Justice, C. (2000). Memory conformity: Exploring misinformation effects when presented by another person. British Journal of Psychology, 91, 189-202. Wright, D. B., & Stroud, J. N. (2002). Age differences in lineup identification accuracy: People are better with their own age. Law & Human Behavior, 26, 641-654. Yarmey, A. D. (1986). Verbal, visual, and voice identification of a rape suspect under different levels of illumination. Journal of Applied Psychology, 71, 363-370. Yarmey, A. D. (1993). Adult age and gender differences in eyewitness recall in field settings. Journal of Applied Social Psychology, 23, 1921-1932. Yarmey, A. D., &Jones, H. P. T. (1983). Is the psychology of eyewitness identification a matter of common sense? InS. Lloyd-Bostock and B. R. Clifford (Eds.), Evaluating witness evidence: Recent psychological research and new perspectives (pp. 13-40). Chichester, England: Wiley. Yarmey, A. D., & Kent, J. (1980). Eyewitness identification by elderly and young adults. Law & Human Behavior, 4, 359-371. Yarmey, A. D., & Yarmey, M. J. (1997). Eyewitness recall and duration estimates in field settings. Journal of Applied Social Psychology, 27, 330-344. Zaragoza, M.S., & Mitchell, K. J. (1996). Repeated exposure to suggestion and the creation of false memories. Psychological Science, 7, 294-300. Zaragoza, M.S., Payment, K. E., Ackil, J. K., Drivdahl, S. B., & Beck, M. (2001). Interviewing witnesses: Forced confabulation and confirmatory feedback increase false memories. Psychological Science, 12, 473-477. Zelig, M., & Beidleman, W. B. (1981). The investigative use of hypnosis: A word of caution. International]ournal of Clinical & Experimental Hypnosis, 29, 401-412.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

III LIFESPAN EYEWITNESS ISSUES: CHILDREN

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

14 The Suggestibility of Children's Memory Laura Melnyk King's University College at the University of Western Ontario

Angela M. Crossman John Jay College of Criminal Justice, The City University of New York

Matthew H. Scullin West Virginia University

Suggestibility refers to the reporting of false information that is explicitly provided by, or inferred from, interviews or conversations with another person. For example, false information may be explicitly presented in the form of misleading questions in which the interviewer's question includes or presupposes a false detail, such as Was his hat blue? when there was no hat or the hat was in fact yellow. False information may also be more subtly presented through the interviewer's reaction to the child's report when it does not coincide with her preexisting ideas. Although individuals of all ages may show suggestibility (e.g., Hyman & Pentland, 1996; Poole & Lindsay, 2001), preschool children (i.e., 3- to 5-year-olds) display heightened levels of suggestibility relative to older children and adults (e.g., Ceci & Bruck, 1993, 1995). Prior to the late 1980s, little empirical research had examined young children's memory for experienced events. Because of the increased involvement of preschool children in high-profile sexual abuse cases in the 1980s and 1990s, a surge of research focused on the accuracy of preschoolers' autobiographical memory and factors affecting their suggestibility. In this chapter, we provide a brief historical review of early studies of children's suggestibility and discuss the increased attention that this field has received in the last 20 years. We present a list of factors that influence children's suggestibility and discuss the methodology used in these studies. Finally, we offer some policy recommendations and directions for future research. 401

402

MELNYK, CROSSMAN, SCULLIN

EARLY STUDIES OF CHILDREN'S SUGGESTIBILITY

Interest in children's suggestibility dates to the early twentieth century and the pioneering work of Alfred Binet in France, Julian Varendonck in Belgium, and Clara and William Stern in Germany. Binet's (1900) early research in suggestibility was especially groundbreaking because many of his findings were repeatedly replicated over the course of the twentieth century. Binet conceived of suggestibility as being influenced by characteristics within the child and by social influences created by the way in which the questioner posed questions to the child. Binet referred to individual differences in children's suggestibility as autosuggestion and devised a number of tasks designed to elicit selfgenerated suggestibility. The best known of these tasks consisted of asking children to reproduce lines that they had been shown one by one. Each of the first several lines was progressively longer than the one that preceded it, and then children were presented with several target lines that were all the same length. Children tended to be influenced by the earlier presentation of successively longer lines and reproduced the target lines as getting progressively longer. When children were queried about their thought processes while they were drawing the lines, many expressed awareness that they were completing the series incorrectly but were influenced by the interviewing situation. This inspired Binet to examine the influence of different questioning techniques upon children's recall of objects they had been presented with (e.g., a button on a board). His findings foreshadowed many later findings: children were most accurate when asked free recall questions and were swayed by interviewer suggestions when they were presented with misinformation embedded in questions. Like later researchers, Binet distinguished between suggestibility due to cognitive factors (in which suggestibility occurs because of poor encoding or misremembered information) and suggestibility due to social factors (such as desire to please an interviewer), which may or may not result in a change in children's memory. Varendonck (1911; as cited in Ceci & Bruck, 1995) developed an interest in children's suggestibility as a result of a murder trial in which the testimony against the accused was primarily from two young girls who had been repeatedly interviewed about the murder and had shown inconsistency across interviews in their testimony. In preparation for his trial testimony, Varendonck conducted a number of demonstrations in which he queried children about the presence of a beard on a beardless man they had seen earlier, about whether a man wearing a hat on his head was in fact carrying it in his hand, and about whether the children could describe a nonexistent stranger from the playground. On the basis of children's high rate of assents for these incorrect claims and the two girls' lack of consistency across interviews, Varendonck argued that the testimony of the children in the trial had been compromised by the suggestive interviewing techniques of the investigators. Clara and William Stern (1908/1999) conducted a number of studies of children's memory and suggestibility, often using their own children as participants. The Sterns (1908/1999) compiled detailed records of their three young children's recollections and tested their suggestibility about pictures and items in the house. Their analysis of the

14. CHILDREN'S SUGGESTIBILITY

403

development of children's memory and suggestibility is remarkably prescient because it describes many issues relating to memory, recall, and suggestibility that would be ad, dressed again in later research. These included the distinction between lies and false beliefs, the importance of cognitive and motivational factors in children's testimony, and the presence of striking individual differences between their children in their suggestibil, ity. Like later researchers, they found that their children's free narratives were more ac, curate than their responses to direct questions. They also addressed issues such as how to increase the accuracy of children's testimony through mental exercises and the im, portance of allowing children to testify in court. William Stern (1910) later replicated and extended these studies with college students in his classroom, interviewing them about pictures and staged events in a suggestive manner. Like later researchers, Stern found that students had a high rate of accuracy during free recall (around 90-95% accu, racy), but accuracy dropped off sharply in response to interrogative questions (to around 70-75% accuracy). Following this outpouring of activity in the early 1900s, the middle years of the twen, tieth century were a relatively lean time for the study of children's suggestibility. Although the early researchers touched upon many of the themes that would interest researchers later in the century, children did not constitute a large enough presence in the court, rooms to interest very many researchers.

REVITALIZATION OF THE FIELD OF CHILDREN'S SUGGESTIBILITY

During the 1980s, there was a confluence of factors that sparked a dramatic increase in the number of children testifying in court (Lyon, 1999). These factors included the lift, ing of restrictions upon children's testimony in courtrooms by easing of competency judgment requirements, removal of hearsay requirements in cases of child sexual abuse, and removal of requirements of corroboration for physical and sexual abuse. Concurrent with the increased presence of children in court, there was a dramatic rise in the number of preschool,age children who testified about being abused by adults in bizarre, ritualistic, and sadistic ways. Often, even though dozens of children would testify about the abuse, there was little or no physical corroborating evidence that a crime had occurred. For example, in the McMartin Preschool case in Manhattan Beach, California, children accused Raymond Buckey, his sister, Peggy Ann Buckey, his mother, Peggy McMartin Buckey, and grandmother, Virginia McMartin, with a variety of forms of ritualistic physical and sexual abuse, including sacrificing babies, taking children on helicopter rides to the desert, and having a secret network of tunnels underneath their day care in which children were forced to participate in sexual activities (Eberle & Eberle, 1993). Charges against Peggy Ann Buckey and Virginia McMartin were dropped before trial because of a lack of evidence, but Raymond Buckey and his mother were tried. After the first trial, the jury acquitted the pair on some charges and deadlocked on others. Raymond Buckey was tried again, but this time the jury deadlocked on all

404

MELNYK, CROSSMAN, SCULLIN

charges and eventually all charges were dropped. After two long and costly trials, the case ended with no convictions but was a harbinger of a series of ritualistic child abuse prosecutions that continued through the mid-1990s. Prosecutors were (at least initially) more successful in other cases. In the Wee Care Nursery School case in New Jersey (State v. Michaels, 1988), children accused Kelly Michaels of participating in a wide range of sexual activities with them over a period of 7 months, including inserting utensils into their anuses and vaginas, smearing peanut butter on them and licking it off, and forcing them to drink her urine. Michaels was convicted of 115 counts of child abuse and sentenced to 4 7 years in prison, although her conviction was eventually overturned (State v. Michaels, 1993), and Michaels was released. In the Little Rascals Day Care case in North Carolina, 90 children accused Robert and Betsy Kelly and several of their employees of physically and sexually abusing them. Robert Kelly was convicted and sentenced to life in prison (State v. Robert Fulton Kelly, ]r., 1991-1992), although this was overturned in 1995 and all charges were eventually dropped in 1999. The appeals of defendants serving lengthy prison terms who were caught up in the wave of child sexual abuse cases continues to this day. Gerald Amirault, the last of the imprisoned defendants in the Fells Acres Day School case in Malden, Massachusetts, was released on parole in 2004 (Ranalli, 2004). John Stoll of the Bakersfield, California mass allegation case was released from prison after his conviction was overturned in 2004. He had served more than 20 years of a 40-year sentence (Jones, 2004). These trials are credited with sparking the resurgence of interest in children's memory and suggestibility during the 1980s and 1990s. Researchers who were called to testify in these trials or who reviewed transcripts or videotapes of interviews with children raised concerns about whether the children had been suggestively led by interviewers (e.g., Ceci & Bruck, 1995). For example, Sena Garven and her colleagues examined videotapes and transcripts of children in the McMartin trial and found that the interviewers used a number of highly problematic techniques (Garven, Wood, Malpass, & Shaw, 1998). These included repeatedly asking suggestive questions, using peer pressure against children by trying to get them to make their accounts conform to what other children were saying, providing positive and negative reinforcement for particular responses, repeatedly asking the same question over and over in hopes of getting children to change their answers, and inviting children to use pretense when responding to questions. As can be seen in the next section, many of these techniques were the focus of research on what Ceci and Bruck (1995) referred to as the architecture of suggestive interviews.

FACTORS INFLUENCING CHILDREN'S SUGGESTIBILITY

The primary factors influencing children's suggestibility can be classified as interviewing factors or child-based characteristics, such as the child's cognitive and psychosocial traits. An individual child's suggestibility results from a complex interaction of these factors (Bruck, Ceci, & Melnyk, 1997). In this section, we highlight those factors that are related to increased suggestibility.

14. CHILDREN'S SUGGESTIBILITY

405

Interviewing Factors There is a great deal of evidence showing that interviewing factors strongly influence children's suggestibility (Ceci & Bruck, 1995). Reviews of the interviews with the children involved in high-profile cases such as the Wee Care Nursery case and the McMartin Preschool case prompted researchers to examine how the techniques employed by the child interviewers empirically related to children's suggestibility in controlled conditions (e.g., Bruck, Ceci, Francoeur, & Renick, 1995; Garven et al., 1998; Leichtman & Ceci, 1995). The central theme in this research is examining how interviewer bias-an interviewer's preexisting ideas about the event under investigation-shapes the entire interview and thus what the child reports (Ceci & Bruck, 1995). This bias may be manifested in the types of questions asked, the number of times questions are asked within or across interviews, stereotype induction, the types of nonverbal techniques employed, and the atmosphere or emotional tone of the interview. From a theoretical perspective, this work allows researchers to develop a better understanding of the mechanisms underlying children's suggestibility. From a practical perspective, this research can inform investigators about techniques that pose a risk to the accuracy and completeness of children's reports.

Open-Ended Questions us. Direct Questions One of the main difficulties faced by investigators in cases involving child witnesses is obtaining accurate and complete reports. When asked open-ended questions such as "Tell me what happened" or "What was he wearing?" children's free narratives tend to be accurate. However, these accounts also tend to be incomplete (e.g., Baker-Ward, Gordon, Ornstein, Larus, & Clubb, 1993; Peterson & Bell, 1996; Steward & Steward, 1996). As a consequence, to increase the amount of information provided by child witnesses, forensic interviewers often have to rely on direct questions requiring a yes/no answer (e.g., Did he take your picture?) or a one-word answer (e.g., Were you upstairs or downstairs?). These direct questions increase the amount of information provided by child witnesses, but this is accompanied by a reciprocal decrease in the accuracy of the child's report (e.g., Ceci & Bruck, 1995). Although protocols for interviewing child witnesses often warn against it, direct questions are often asked in investigative interviews (Warren, Woodall, Hunt, & Perry, 1996). A biased interviewer's direct questions may be misleading and reflect preconceived notions about the event and thus pose additional risk to the accuracy of the child's report.

Repeated Questioning Biased interviewers often repeat questions if the child does not provide an appropriate or anticipated answer (Poole & Lamb, 1998). Children may change their answer to a repeated question, reasoning that if the question is being asked again, their original

406

MELNYK, CROSSMAN, SCULLIN

response must not have been correct (Ceci & Bruck, 1995). Poole and White (1991) found that repeating open-ended questions within an interview did not affect the responses provided by 4-, 6-, or 8-year-old children or adults. But repeating direct questions had a negative effect on the accuracy of the 4-year-olds' responses: The 4-yearolds' accuracy was compromised when they were repeatedly asked yes/no questions, whether these questions were repeated within a given interview or across different interviews. In addition, when asked direct questions that invited speculation (e.g., "What did the man do for a living?" when this could not be known), the 4-year-olds provided concrete answers, and when the question was repeated, their responses included few qualifiers such as "It might have been," thus making their statements sound confident.

Repeated Interviewing Another indication of interviewer bias is repeated interviewing. A reliable research finding is that a single presentation of misinformation is sufficient to produce suggestibility in children (e.g., Ackil & Zaragoza, 1995; Ceci, Ross, & Toglia, 1987; Garven, Wood, & Malpass, 2000). However, because children involved in legal proceedings may be officially interviewed up to 12 times over the course of an investigation (Whitcomb, 1992), research has examined the effects of repeated questioning on the accuracy of children's reports. Repeated neutral interviews (where no misinformation is presented) may offer some benefits to the accuracy of children's recall. Repeatedly talking about an event can prevent one from forgetting the correct details of the event (Baker-Ward, Hess, & Flannagan, 1990; Dent & Stephenson, 1979), it may allow the child to report previously unmentioned details in later interviews (e.g., Howe, Kelland, Bryant-Brown, & Clark, 1992; Howe, O'Sullivan, & Marche, 1992, but see Cassel & Bjorklund, 1995), and it may increase in the amount of new information recalled over time that exceeds the amount of information forgotten (Memon, Wark, Bull, & Koehnken, 1997). In practice, however, even a well-intentioned interviewer may not be completely neutral, because the interviewer may simply not be aware of all the details of the event in question and thus may inadvertently ask misleading questions. In a recent study assessing whether suggestive interviews had a cumulative effect on children's suggestibility, Melnyk and Bruck (2004, experiment 2) suggestively interviewed kindergarten children about an interactive magic show about 2 days or 3 7 days following the event or at both 2 and 3 7 days following the event. In a memory test about 39 days after the magic show, those children who were interviewed twice reported more of the suggestions than the children who were interviewed only one time. The specific timing of the suggestive interviews may be critical to the augmented effects; studies in which the timing of the interviews is more temporally distant from the event and/or the memory test do not promote a cumulative effect of repeated interviewing (e.g., Melnyk & Bruck, 2004, experiment 1; Warren & Lane, 1995). At any rate, it is clear that even a single suggestive interview poses significant risks to the accuracy of children's recall, and repeated interviews can further increase these risks.

14.

CHILDREN'S SUGGESTIBILITY

407

Stereotype Induction Interviewer bias may be manifested in more subtle ways than misleading questions. In stereotype induction the interviewer describes the accused or the accused's presumed actions in a negative manner. For example, Ceci and Bruck (1995) cite interview transcripts from cases in which children were told things such as the accused was "bad" or "does bad things." In one study (Leichtman & Ceci, 1995), 3- to 6-year-old children were repeatedly told by their teacher about a man named Sam Stone who was very clumsy. When Sam Stone finally visited the classroom, he said "hello" and simply walked around the room for about 2 minutes. When the children were later asked about two things that did not happen during Sam Stone's visit-a book was ripped and a teddy bear was soiled-about a quarter of the children reported that Sam Stone may have been responsible. The children were then asked about Sam Stone performing these acts over the course of four suggestive interviews. When later asked by a new interviewer about the book and the teddy bear, 72% of the 3- to 4-year-olds and about 40% of the 5- to 6-year-olds reported that Sam Stone was responsible for at least one of these transgressions. The children's reports often included details that were not suggested. In addition, many of the children maintained their false reports even when the interviewer gently challenged their claims.

Nonverbal Techniques: Props, Drawing, and Dolls Because young children's responses to open-ended questions tend to be accurate but sparse (e.g., Baker-Ward et al., 1993; Peterson & Bell, 1996) and direct questions may inadvertently include misinformation, researchers have attempted to identify nonverbal interviewing techniques that would maximize accurate reports without increasing errors. For example, forensic interviewers may utilize nonverbal cues such as props and anatomically detailed dolls in an attempt to facilitate children's recall. However, many of these techniques do not fare well under scientific scrutiny. Props, for example, increase not only the number of accurate details recalled, but also the number of inaccurate details recalled (Gee & Pipe, 1995; Salmon, Bidrose, & Pipe, 1995; Salmon & Pipe, 1997; Steward & Steward, 1996; Wilson & Pipe, 1989). Drawing may facilitate children's reports under neutral conditions (Butler, Gross, & Hayne, 1995), but when an interviewer asks children to draw specific details of an experienced event, drawing increases children's suggestibility even more than simply hearing the suggestions (Bruck, Melnyk, & Ceci, 2000). Much attention has been devoted to examining whether anatomically detailed (AD) dolls facilitate children's reporting of genital touching. The use of AD dolls in interviews with children has been endorsed by the American Psychological Association (1991) as sometimes being the best available practical solution for breaching topics related to sexual abuse with children, although it was noted that little normative research had been conducted with AD dolls. In the years since the resolution, the utility of AD dolls with 3- and 4-year-old children has been found to be problematic in several studies because it can lead to false positives in which children indicate genital touching when this had not occurred (Bruck, et al., 1995; Bruck, Ceci, & Francoeur, 2000). This problem

MELNYK, CROSSMAN, SCULLIN

408

may arise because young children have difficulty with symbolic representation, in which they are required to transfer their own knowledge about their own anatomy to a sym, bolic representation, the doll (DeLoache & Marzolf, 1995). This ability seems to de, velop between the ages of 3 and 5, but there appear to be striking differences in this abil, ity at any given age. There is growing consensus that the use of AD dolls with children under the age of 5 is problematic (Hungerford, 2005), and some researchers question whether the dolls add significantly to investigative interviews with children of any age group (Dickinson, Poole, & Bruck, 2005). For example, in a study with 4, to 12,year, olds, a group in which AD dolls were used was found to produce reports that were simi, lar in quality to those by children who were interviewed without AD dolls, but the chil, dren interviewed without the AD dolls provided longer and more detailed reports (Lamb, Hershkowitz, Sternberg, Boat, & Everson, 1996).

Atmosphere Interviewer bias may manifest itself in the atmosphere or emotional tone created by an interviewer. Children interviewed in an accusatory atmosphere may provide false reports that correspond to the interviewer's tone (e.g., Garven et al., 1998; Lepore & Sesco, 1994). For example, in a study by Goodman, Wilson, Hazan, and Reed (1989), children played with an unfamiliar research assistant for 5 minutes and were brought back 4 years later to talk about their experience. The interviewer created an accusatory atmosphere by telling the children that they were going to talk about an important event and providing statements such as ''Are you afraid to tell? You'll feel better once you've told." Although the children had little memory for the original event, several of their reports included details that matched the interviewer's tone, such as agreeing that they were hugged or kissed by the research assistant, the research assistant took their photo, graph in the bathroom, or they were given a bath. In contrast, a supportive interviewing atmosphere may decrease children's suggestibil, ity. For example, J, to 7,year,old children were asked questions about a recent inocula, tion by either a supportive interviewer who smiled and complimented and praised the child's performance or a neutral interviewer who did not provide these niceties. The J, to 4,year,olds who were interviewed by the supportive interviewer answered abuse, related questions more accurately and provided fewer false details compared with the children who were interviewed by the neutral interviewer (Goodman, Rudy, Bottoms, & Aman, 1990; but see Cordon, Saetermoe, & Goodman, 2005, and Imhoff & Baker, Ward, 1999).

Child-Based Characteristics Factors inherent to the child that influence suggestibility include the child's cognitive abilities and psychosocial characteristics. Although many studies have examined "indi, vidual difference" factors in children's suggestibility, only a few have been consistently related to children's suggestibility. Bruck and Melnyk (2004) reviewed 69 studies cover, ing over 500 analyses examining several cognitive and psychosocial factors and their

14.

CHILDREN'S SUGGESTIBILITY

409

relationship to children's suggestibility. At this time, there is no profile for identifying children who are most likely to be highly suggestible, and most of these factors account for relatively little of the variance in children's suggestibility. As research continues, however, a clearer picture of the key factor or interaction of key factors that relate to in~ dividual children's suggestibility may emerge. Factors found to be consistently related to children's suggestibility include the following:

•

• • •

Intelligence. Children with mental retardation are more suggestible than their age~ matched, typically developing peers. However, when typically developing children are examined, intelligence is not reliably related to children's suggestibility. Language ability. Children with more advanced language skills tend to be less sug~ gestible than children with weaker language skills. Creativity. Children who are more imaginative and creative tend to be more sug~ gestible than less imaginative and creative children. Self-concept. High self~concept is related to lower levels of suggestibility.

MECHANISMS UNDERPINNING CHILDREN'S SUGGESTIBILITY

Social Mechanisms Children's suggestibility is rooted in a combination of social and cognitive mechanisms. Social mechanisms include compliance with authority figures and children's understand~ ing of conversational rules. For example, children are more suggestible when interviewed by an interviewer of high status (e.g., Ceci et al., 1987). Tobey and Goodman (1992) made less accurate reports about their experience with a "baby found that 4~year~olds sitter" when they were interviewed by an interviewer dressed as a police officer com~ pared with a more neutral interviewer. Similarly, children may respond to misleading questions with a yes or no answer because they do not think it is appropriate to answer "I don't know" (Hughes & Grieve, 1980; Melnyk, 2002). Children may view providing I don't know answers as uncoopera~ tive in their conversations with high~status adults (Ceci & Bruck, 1995). Indeed, train~ ing children to provide I don't know answers reduces the number of errors in their reports (Mulder & Vrij, 1996).

Cognitive Mechanisms Cognitive factors that may account for children's suggestibility include memory abili~ ties, knowledge, and source~monitoring ability. For example, children may simply fail to encode events, store memories, and retrieve memories as efficiently as adults, and this leaves their memories more susceptible to misleading questions (Ceci & Bruck, 1995). There is evidence that children with weaker memory traces for an event are more sug~ gestible than children with stronger memory traces for the event (Marche, 1999; Marche

410

MELNYK, CROSSMAN, SCULLIN

& Howe, 1995; Pezdek & Roe, 1995). A related cognitive factor is the child's knowledge about the event in question. That is, children who know a fair amount about the event in question may be less susceptible to suggestive questioning than children with less knowledge of the event (e.g., Sutherland, Pipe, Schick, Murray, & Gobbo, 2003, but see Greenhoot, 2000). Source monitoring is the ability to distinguish the sources for information in memory (e.g., Johnson, Hashtroudi, & Lindsay, 1993; Roberts, 2000). Failures of source monitoring involve confusing misinformation with the details of the actual event in question, such as confusing misinformation that was heard in a suggestive interview with what the child actually saw in an original event. Recent work correlates children's suggestibility with failures in source monitoring (e.g., Giles, Gopnik, & Heyman, 2002; Thierry & Spence, 2002; but see Bruck et al., 2000). Unfortunately, attempts to train or warn children to use source monitoring to guide their recall have met with varied and limited success. Sometimes source monitoring training facilitates accurate recall even in younger children (Thierry & Spence, 2002), whereas in other studies, it does not, or it only benefits older children (Poole & Lindsay, 2002).

Consistency of Children's Reports Consistency of a report refers to the witness repeating the same information when questioned repeatedly about an event. Although consistency of a child witness's report influences perception of accuracy (Leippe, Manion, & Romanczyk, 1992), a witness's report can be accurate without being consistent, inasmuch as different correct, noncontradictory information may be reported over several interviews (Ghetti, Goodman, Eisen, Qin, & Davis, 2002). Younger children tend to be less consistent than older children and adults in multiple reports (e.g., Hudson & Fivush, 1991). Brady, Poole, Warren, and Jones (1999) asked 3- to 7-year-old children yes/no questions twice and found that the younger children were less accurate and less consistent than the older children. Similarly, Poole and White (1991) found that 4-year-olds answered repeated yes/no questions less consistently than 6- and 8-year-olds and adults. Preschoolers' inconsistent reporting is not limited to yes/no questions; they are also inconsistent when providing reports in response to open-ended questions, even though their reports to open-ended questions can be highly accurate (e.g., Hamond & Fivush, 1991; Peterson, Moores, & White, 2001; but see Pipe, Gee, Wilson, & Egerton, 1999). Interestingly, Poole and White (1991, 1995) found that children were less consistent in response to misleading direct questions but were more consistent for neutral open-ended questions.

Long-Term Effects of suggestive Interviewing There may be several months between interviews with child witnesses and the start of a trial (Flin, 1995; Goodman et al., 1992). This has prompted research examining the long-term effects of suggestive interviewing on children's later reports. There is some

14.

CHILDREN'S SUGGESTIBILITY

411

indication that children's reporting of completely false autobiographical events may diminish over two years (Huffman, Crossman, & Ceci, 1997), and suggestibility may decrease between 3.5 months and 4.5 months (Poole & Lindsay, 2001); however, some false reporting did persist over these intervals. Melnyk and Bruck (2004, experiment 1) found that children's rate of reporting false suggestions was maintained from a 6-week memory test to a 21-week memory test. Although the results of these few studies examining the long-term effects of suggestive interviewing are inconsistent, they illustrate that false reporting can persist for several months to 2 years.

METHODOLOGICAL ISSUES IN THE STUDY OF CHILDREN'S SUGGESTIBILITY

Researchers have clearly learned a great deal over the past century about the potential suggestibility of child witnesses in a variety of contexts. Nevertheless, one central critique of much of this research has been its lack of ecological validity; that is, the types of events typically used as the "target" events in these studies do not correspond well with situations involving actual child witnesses. For suggestibility research, this entails scrutiny of such factors as salience, meaningfulness, and abuse relevance of stimuli and questioning, characteristics of the children in the study, and children's motivations. The reality is that it is not possible to recreate the exact conditions inherent in an abuse investigation, nor would it be desirable to do so, for ethical reasons. This raises the question of whether the research should be applied to actual cases that rely on children's evidence. To answer this question, Lyon (2002) highlighted the importance of examining the relevant research methodology for its applicability to real-world cases. One common critique of many suggestibility studies is that suggestibility assessed via children's responses to questions about irrelevant, superficially relevant, or peripheral details of witnessed events is not germane to an understanding of actual child witnesses' experiences. For example, one might question whether the initial studies in which children were asked suggestive questions about a picture they observed (Stern, 1910) or about a button glued to a poster board (Binet, 1900) are meaningfully related to the suggestibility of their memories about a potentially traumatic event, such as abuse. Nevertheless, the findings from these historical studies are consistent with more modern results (Ceci & Bruck, 1995). In addition to concerns about irrelevance, criticism can be leveled at the use of superficially relevant or peripheral details of witnessed events as the focus of suggestibility manipulation. Noticing actions or characteristics of nonessential actors in an observed event (e.g., the clothing of a noncentral character; Brady et al., 1999) might actually indicate a lack of attention to central information and perhaps a lack of appreciation for what is important in an event. Moreover, it is important to bear in mind that what is considered to be central event information to an adult might not be so to a child. That is, event salience is not an absolute, and children might attend to aspects of the event other than what experimenters might anticipate. Both factors can lead to an overesti-

412

MELNYK, CROSSMAN, SCULLIN

mation of children's suggestibility if children are queried about events to which they were not attentive, as they were peripheral and lacked salience. Event presentation style varies across suggestibility studies as well. Whereas some researchers have read children stories about false events (Miles, Powell, & Stokes, 2004), others have shown them videotaped events (Scullin & Ceci, 2001), and still others have involved live events (e.g., Leichtman & Ceci, 1995). Clearly, live events are the most ecologically valid of the presentation styles; yet there is variation even in live events. Children have been studied both as observers and as participants in live events (Gobbo, Mega, & Pipe, 2002; Lindberg, Jones, McComas, & Thomas, 2001). For example, in the study described above, Leichtman and Ceci (1995) suggestively interviewed children about a 2-minute visit to their classroom by a stranger named Sam Stone. Predictably, suggestibility was quite high in some conditions of this study, helped in part by the briefness of the stranger's visit, which might have lacked salience for many of the children. In contrast, Rudy and Goodman (1991) questioned 4- and 7-year-old children who had either observed or played a game with an unfamiliar adult. Findings indicated that memory is better and more accurate with participation, as opposed to simple observation. Nevertheless, some children remain suggestible, even about events in which they have participated (e.g., Bruck et al., 2000; Poole & Lindsay, 1995). With respect to event content, researchers have attempted to more closely represent abuse-relevant situations by studying situations involving bodily contact (e.g., children's visits to their pediatricians) and by asking abuse-related questions (Goodman, Batterman-Faunce, Schaaf & Kenney, 2002; Merritt, Ornstein, & Spieker, 1994; Rudy & Goodman, 1991; Saywitz, Goodman, Nicholas, & Moan, 1991; Tobey & Goodman, 1992). These were considered potential domains in which suggestion was not possible. Using this more ecologically valid approach, researchers have found most (but not all) children to be highly resistant to abuse-related suggestions about bodily touch. Yet, when such interviews have been conducted in an accusatory context, children have been more vulnerable to suggestion (Lepore & Sesco, 1994). Similarly, young children's reports of bodily touching may include false details when anatomically detailed dolls are used in the interview (Bruck et al., 1995). This raises the question of the prevalence of such suggestive interviewing techniques, which is not known. Nevertheless, these methodological contributions heighten the relevance of the research to actual cases. Subsequent to the event experience, researchers seek responses from participants in an effort to gauge their suggestibility. However, there is a continuum of responses to suggestibility measures as well. For example, in some experiments, children must simply agree or disagree (i.e., yes or no) with an interviewer to be considered suggestible, whereas other designs require greater elaboration. For example, the Gudjonsson Suggestibility Scale (GSS; Gudjonsson, 1984, 1987) and the Video Suggestibility Scale for Children (VSSC; Scullin & Ceci, 2001) both provide suggestibility summary scores reflecting yes responses with which children acquiesce to misinformation. However, recent work is attempting to validate these simple forced-choice responses as predictors of suggestibility for real-life events (e.g., McFarlane & Powell, 2002; Scullin, Kanaya, & Ceci, 2002). Perhaps somewhat more convincing has been research finding that simply asking children repeatedly to think about false events can lead to elaborate false reporting

14.

CHILDREN'S SUGGESTIBILITY

413

(Ceci, Huffman, Smith, &Loftus, 1994; Ceci, Loftus, Leichtman, &Bruck, 1994), although this methodology is not without its critics as well (Leavitt, 1999). It is worth noting as well that definitional and procedural variations in terms of coding might also influence suggestibility outcomes. Across various studies, what is meant by direct, leading, and misleading questions varies somewhat, making direct comparisons difficult. Indeed, it might be argued that any questions that ask anything more specific than "What happened ... ?" are inherently leading or misleading-it is simply a question of degree (Goodman & Schaaf, 1997). Certainly the degree of leading present in a question will affect the extent of children's suggestibility. Furthermore, how accuracy and suggestibility are defined across studies varies and can be subject to different interpretations (Ornstein, Baker-Ward, Gordon, & Merritt, 1997). At the simplest level, researchers can report the proportion of children acquiescing to suggestion or the proportion resisting suggestion. Another frequent methodological variation that affects children's suggestibility is the length of time between an event, suggestive questioning, and reporting. Children's vulnerability to suggestion is influenced by their memory trace for an event, which will be diminished by the passage of time. Researchers have attempted to vary delays (e.g., Leichtman &Ceci, 1995; Poole & White, 1993; Shapiro, Blackford, &Chen, 2004) to better reflect the forensic reality that children are frequently interviewed long after an event occurs. Goodman et al. (2002), for example, suggestively interviewed children about an event 4 years after it occurred (see also Goodman et al., 1989). However, rehearsal of memory over time may be more likely in forensic settings, as witnesses are prepared for trial. Indeed, inasmuch as it is estimated that child witnesses are officially interviewed up to 12 times during an investigation (Whitcomb, 1992), the importance of understanding repeated questioning and the impact of time on children's reports is critical. Repeated questioning, if nonsuggestive, might in fact reinforce memory (accurate or not), highlighting the importance of initial nonsuggestive interviews with young children (Fivush & Schwarzmueller, 1995, 1998). Moreover, given the reality of child victimization, interviewers are more likely to be retrieving memories involving known individuals, as opposed to strangers, who are the typical targets in experimental research. Further research involving longer delays and repeated interviews will help to improve researchers' understanding of children's suggestibility in increasingly ecologically valid ways. One final critique of typical experimental research has been the characteristics of the children involved in the research sample. The typical research sample does not resemble the typical child victim/witness, which raises the question of the generalizability of the findings. Research indicates that victimization can affect psychological functioning (e.g., Eigsti & Cicchetti, 2004; Pears & Fisher, 2005; Van Voorhees & Scarpa, 2004); hence studies examining suggestibility of those most likely to experience abuse, or those who have been abused, is vitally important (Eisen, Qin, Goodman, & Davis, 2002). It is also important to recognize that a child's motivations in an experiment could vary drastically from those in reality. For instance, manipulations of shame, fear, and/or embarrassment are not present in typical experimental research. Hence, lack of a representative sample is a valid criticism of much of the existing research.

414

MELNYK, CROSSMAN, SCULLIN

Cognizant of the need for more ecologically valid studies, suggestibility researchers have adapted their methods over time to be increasingly relevant to forensic settings, examining all aspects of children's reports, up to and including the credibility of children offering tainted memory reports. Most importantly, researchers have increasingly attempted to examine children's memory for situations that share characteristics with those faced by most child witnesses. That is, researchers have looked in the field to study memory for invasive, personally involving experiences that involve bodily touch and possible embarrassment (e.g., Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994, 1997; Merritt et al., 1994; Quas et al., 1999). Although participants in such research are not suspected to be victims of abuse, their traumatic exposure would seem relevant. Again, although children can be highly resistant to suggestion, this research indicates that not all children are always resistant. Given the numerous criticisms leveled at laboratory research, one might ask why researchers do not rely more heavily on field research, which offers greater ecological validity. However, field research suffers from its own shortcomings. In particular, lack of control over the to-be-recalled experiences and group membership makes causal attributions impossible. One type of event that resembles child sexual abuse in several ways is a medical procedure called a voiding cysto-urethrogram (VCUG; Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994). In this painful procedure, used to image the urinary tract of children suffering from chronic infections, a probe is inserted into the urethra, the bladder is filled with fluid, and the child is asked tQ urinate on the examining table in front of medical personnel. But even in studies in which the VCUG is used as the target event, it is not possible to determine whether characteristics of children exposed to the medical procedure co-vary in some way with their suggestibility or other relevant characteristics. In addition, although one can study the suggestiveness of an interview in field studies, the ground truth is generally not known, making unclear the extent to which a child has or has not succumbed to suggestion. One useful emerging methodology is the combination of laboratory and field research. For example, Zajac and Hayne (2003) have studied the impact of cross-examination on child witness accuracy and credibility. Having examined trial transcripts to glean an understanding of how child witnesses are treated, the researchers then have recreated these methods under the controlled conditions of the laboratory to examine their impact. This combination of real-world knowledge and experimental control may provide the most useful understanding of child witness suggestibility (see also Garven et al., 1998, 2000). Overall, research on child witness suggestibility, accumulated via a broad spectrum of research studies with varying methodologies, has converged on the finding that children, particularly very young children, can be vulnerable to suggestion (Ceci & Bruck, 1995). In addition, children do not readily acquiesce to abuse-related suggestions, particularly regarding personally involving, invasive experiences that include bodily touch. However, some children are more vulnerable than others and some contexts are more persuasive than others. The strength of suggestion is important in this regard, with more extreme forms of interview pressure resulting in greater degrees of suggestibility. This

14.

CHILDREN'S SUGGESTIBILITY

415

again raises the question of how prevalent such practices are in real-world cases (Lyon, 2002). Nevertheless, even moderate forms of suggestion can influence some child witnesses, reinforcing the need for continued caution, as described below.

POLICY RECOMMENDATIONS

Over the course of this chapter we have highlighted many of the factors affecting the reliability of children's testimony and methodological issues related to the study of children's suggestibility. Many of the studies that we have discussed so far have demonstrated that children can incorporate suggestions from interviewers; numerous studies have also been conducted that have assessed techniques for eliciting accurate information from children and reducing their suggestibility. On the basis of this research, we can provide a number of concrete policy recommendations for improving or maintaining the quality of children's testimony.

Use a Structured Interview When Interviewing Children. There has been an increasing level of research on how to enhance the quality of interviews conducted with children. As mentioned earlier, an analysis of transcripts of interviews conducted during the McMartin Preschool case revealed that interviewers used numerous coercive techniques upon children in order to elicit information (Garven et al., 1998), and analyses of trial transcripts found that interviewers often use complicated, developmentally inappropriate language when interviewing children (Perry et al., 1995). In order to provide guidance for interviewers of children, researchers have developed structured interviews for use with children that are based upon empirical research on children's testimony (Home Office, 1992; Kohnken, 1993; Poole & Lamb, 1998; Orbach et al., 2000). Structured interviews help increase the reliability of children's testimony by using techniques that help the interviewer develop a rapport with the child and elicit as much information as possible through free recall. Free recall questioning techniques have been found to elicit more accurate information from children than close-ended questions. The National Institute of Child Health and Human Development (NICHD; Orbach et al., 2000) has developed a structured interview for use with children whose basic elements are similar to those found in other structure interviews (e.g., Kohnken, 1993) and provide a useful framework for successful interviews. Mere familiarity with the concepts behind a structured interview may not be sufficient to ensure that their guidelines are utilized. Researchers have found that simply knowing general principles of interviewing may not be translated into action (Warren, Woodall, Thomas, et al., 1999). The following include many of the empirically based general principles utilized in structured interviews. A more detailed description of the NICHD interview may be found in Orbach et al. (2000). Provide a Supportive Atmosphere for the Child. Researchers have found that children tend to be more accurate in response to both open-ended and close-ended

416

MELNYK, CROSSMAN, SCULLIN

questions when they are interviewed in a supportive atmosphere (Carter, Bottoms, & Levine, 1996; Davis & Bottoms, 2002; Goodman et al., 1990). The elements of a supportive interview include having an open body posture toward the child and smiling and being responsive to the child. The benefits of a supportive atmosphere are thought to occur by helping children regulate their emotions and resist suggestive influences. Consistent with this hypothesis, the benefits of a supportive interview have been found to be especially pronounced with highly reactive children (Quas, Bauer, & Boyce, 2004).

Establish the Child's Understanding of the Interview. Interviewers should introduce themselves to the child witness and explain who they are and what they do in general terms. The NICHD protocol begins with some questions that help determine whether the child understands the distinction between the truth and a lie. The interviewer then sets the ground rules for the interview, such as telling the child that it is acceptable to say "I don't know" in response to questions they do not know the answer to or do not understand. The child is also told that it is acceptable to ask for clarification or to contradict the interviewer when the interviewer says something that is incorrect. Build Rapport. The next part of a well-conducted investigative interview is to build rapport with the child witness. These include asking the child questions about his or her family and things that he or she has done recently. The use of open-ended rapport-building questions has been found to enhance the quality of children's recall of an event later in the interview (Roberts, Lamb, & Sternberg, 2004). Use Open-ended Questions to Prompt Free Narratives. The interviewer then introduces the topic of interest in the interview and asks broad, open-ended questions about the event. The interviewer should follow up responses to these questions with more open-ended probes (e.g., ':A.nd then what happened?"). All interview protocols reviewed by Walker and Hunt (1998) recommended starting the information-gathering part of the interview with free narratives. In spite of the near-uniformity of this recommendation, analysis of interviewer transcripts has found a heavy reliance on direct questions rather than open-ended questions (e.g., Lamb, Hershkowitz, Sternberg, Esplin et al., 1996; Hershkowitz, Lamb, Sternberg, & Esplin, 1997). Direct Probe Questions. If the child appears to have left out crucial details of the event during the free narrative portion of the interview, the NICHD protocol recommends that the interview then progress to the use of direct probe questions, followed by open-ended questions where feasible (e.g., Have you seen [the perpetrator] before? or Where did you see him before?). Interview Closure. The interview then concludes with some general questions to the child about the event (e.g., Is there anything else I should know?) followed by thanking the child for the interview. Then the interviewer can ask a few rapport-building questions before returning the child to the person who brought the child in.

14.

CHILDREN'S SUGGESTIBILITY

417

Use Developmentally Appropriate Practices with Children Anatomically Detailed Dolls Should Be Used with Caution. As described above, anatomically detailed dolls pose a threat to the accuracy of children's reporting, particularly when the child is a preschooler. This is largely due to young children's difficulties with understanding that the doll serves as a representation of their own bodies. Although it may seem that AD dolls could facilitate young children's abilities to report abuse by circumventing their linguistic limitations and potential embarrassment, research demonstrates that AD dolls are problematic when used with young children (e.g., Bruck et al., 1995, 2000) and perhaps children of any age (Dickinson et al., 1995). Developmentally Appropriate Language Improves Children's Testimony. Analyses of interview transcripts have found that interviewers often use complicated syntax and styles of questioning that children have difficulty understanding (Walker & Hunt, 1998). Walker and Hunt's analyses found that styles of questions that young children have difficulty responding to include direct forced-choice questions that limit the possible range of responses (e.g., Was the car red or blue?) and multipart questions (e.g., Would it be a special time when you would go up there with [him], or would there be anybody there at the house?). Children also tend to misinterpret Can you ... ? questions as calling for a yes-no response (e.g., Can you tell me where he hit you?). Walker (1994) has also argued that young children tend to act on the assumption that every question requires a response and will provide responses to questions they do not understand or are nonsensical. Many forensic interviewers assume that young children can understand and discuss time in the same manner as adults. However, as Stern and Stern (1908/1999) noted nearly a century ago, young children can have difficulty placing events in correct sequence or placing events into temporal contexts like yesterday or a week ago, or understanding numerical concepts such as the number of times that an event occurred. Interviewers should exhibit caution in their use of and interpretation of temporal terms in their investigative interviews.

Videotape Investigative Interviews We recommend that all interviews with child witnesses be videotaped, for several reasons. A well-conducted forensic interview with a child in supportive surroundings provides an important early record of the child's testimony that can be evaluated by other professionals, sometimes in lieu of re-interviewing the child (Lamb, Sternberg, & Esplin, 1995). It provides transparency for the interview process by allowing other viewers to see the techniques that are utilized by the children's interviewers. Documentation of children's early accounts of an event can be important because children's memory for events tends to weaken over time, and later interviews can become less accurate (e.g., Flin, Boon, Knox, & Bull, 1994; Pipe et al., 1999). Younger children are more prone to being

418

MELNYK, CROSSMAN, SCULLIN

influenced by the suggestions of an authoritative adult (Ceci & Bruck, 1995), so a record of their interactions with interviewers is especially important. Videotaping is also beneficial for the reconstruction of what occurred during the interview. Researchers have found that interviewers have a great deal of difficulty recalling verbatim statements from a child and will frequently report the gist of a conversation or report that the child said something that was actually said by the interviewer (Lamb, Orbach, Sternberg, Hershkowitz, & Horowitz, 2000). Although interviewers tend to recall the gist of an interview well (Warren & Woodall, 1999), they tend to misattribute responses to direct (close-ended) questions as being responses to open-ended questions (Lamb et al., 2000). It is frequently mentioned anecdotally that when children's testimony is videotaped, this reduces the number of cases that go to trial because they increase the likelihood that a defendant will plead guilty rather than face a jury (McGough, 2002). However, there is little evidence to back up this assertion. In summary, psychological research has provided numerous practical suggestions for improving the quality of information received from children during interviews. Key among these is the attempt to generate as much free narrative reporting as possible from the child by being supportive, waiting for the child to finish responding, and asking open-ended probe questions. Research has also highlighted the pitfalls of developmentally inappropriate language and AD dolls, and as a practical matter we recommend the videotaping of interviews.

FUTURE RESEARCH DIRECTIONS

Professionals charged with the duty of interviewing child witnesses shoulder a difficult burden. They must ascertain the truth of what happened in order to protect child victims, while not influencing children's reports through suggestion. They seek to protect both children and those who might be falsely accused of harming them. Hence, child protection professionals must perform a delicate balancing act between less suggestive interviewing strategies that yield accurate but less complete reports versus more suggestive interviewing strategies that promote more complete, but less accurate, reports. This chapter has provided several policy recommendations for interviewers struggling with this dilemma. These recommendations are based on converging findings from methodologically varied studies on children's memory and suggestibility. Psychological reliance on convergent evidence means that conclusions were drawn from various studies' findings converging on similar results. However, the probabilistic nature of research means there is no litmus test for suggestibility in every situation-only risk and protective factors that might increase or decrease likely suggestibility, respectively. In the end, we are left with the fundamental tension between what psychologists can offerprobabilistic information based on aggregate data from converging studies-and the forensic practitioner's need for guidance in individual cases. One tool that might be of great value to interviewers is an individual measure of suggestibility. With such an instrument, one might be better able to predict or detect which children are most susceptible to influence of varying degrees. The Gudjonsson Sug-

14. CHILDREN'S SUGGESTIBILITY

419

gestibility Scale (Gudjonsson, 1984, 1987) is one such tool designed for use with adults, and the Video Suggestibility Scale for Children (Scullin & Ceci, 2001) is one designed for children. With such tools, researchers could inform practice in individual cases, perhaps highlighting a particular child's robust resistance to typical suggestion or raising the need for exceptional caution with another child who might be unusually susceptible to interviewer influence. Although such information will remain probabilistic and perhaps situationally dependent, it could at least serve as another tool in the toolbox of frontline workers. Research on the prediction of individual suggestibility also can help to clarify the nature of the construct. That is, both the mechanics and underlying nature of the construct of suggestibility remain unclear. Is it a state? A trait? A combination? There is no genetic marker for suggestibility and no developmental absolute. There is no one social context indicative of universal child suggestibility. Yet, using individual assessments, researchers can examine the stability of the phenomenon across contexts to begin to address these questions. Hence, more research is needed examining children's suggestibility across contexts. Is suggestibility stable and predictable for an individual across situations? Is suggestibility stable if the different interviews vary in suggestive intensity? On the other hand, it also seems clear that contextual factors are relevant to suggestibility effects, such as atmosphere or emotional tone of the interview. Although there are some data addressing whether there are particular contextual factors that indicate suggestibility is highly likely and others that are unlikely to be associated with suggestibility, future research needs to examine how these contextual factors relate to the tendencies of individual children as well as to developmental factors. Are there some contexts in which nearly all young children with a poor self-concept, for example, would be vulnerable to suggestion? Moreover, it will be important to examine how the child's characteristics interact with interview factors. Does a child's self-concept have a differential impact on suggestibility, depending on the interviewer's self-concept, emotional tone, and questioning style (e.g., choice of whether use repeated questions)? With respect to the latter example, this type of more complex analysis of suggestibility will be necessary in future research. Following the lead of Gilstrap and Papierno (2004), it will become increasingly important to simultaneously examine the multiple determinants of suggestibility. When examining child characteristics, researchers must consider them in the context of both the interview situation and the dyadic interaction that interviews are. Indeed, a promising line of research examines not only how an interviewer's characteristics influence children's suggestibility, but how the child's behaviors influence the interviewer's behaviors (Gilstrap & Ceci, 2005). As well, it will be important for researchers to further study the longevity of suggestibility effects to better understand the boundary conditions for individual suggestibility. Indeed, it remains unclear what actually occurs to one's memory when it becomes tainted by suggested misinformation. By better understanding the bidirectional interview context and the longterm impact of suggestive interviewing, both the true nature of suggestibility and the mechanism(s) underlying suggestibility can be better elucidated. Finally, researchers must continue to seek creative ways to address concerns about the ecological validity of suggestibility research because, in the end, for the work to be

420

MELNYK, CROSSMAN, SCULLIN

maximally useful, it must be maximally relevant. It seems likely that combined laboratorybased and field-based research, examining suggestibility from different perspectives-for example, the analyses in the laboratory by Garven et al. (1998, 2000) of interviewing techniques observed in the McMartin Preschool case-will be very useful in guiding researchers' recommendations and interviewers' practices. Moreover, research that includes samples of children most often represented among victims/witnesses, in addition to samples of children who actually have been victims or witnesses, will be vitally important (Eisen et al., 2002). In addition, to the extent possible, research should include stimuli most relevant to abuse situations and/or interview contexts most relevant to forensic interviews. Researchers also must be aware of additional factors, such as motivational factors, that might differentiate laboratory-based research findings from reality. Children's reluctance to disclose abuse and their tendency to retract their claims are not typical results found in their memory reports about a videotaped birthday party, for example. Both field research and research in other domains of psychology (e.g., children's admissions of wrongdoing) could help to address some of these factors (Lyon, 2002). Finally, given that, in most cases, children are more likely than not to be making honest and generally accurate allegations, how can psychologists best inform the legal system about risks to accuracy? How does knowledge of suggestibility and the various risk factors present in individual cases affect legal decision-making? The most important application of suggestibility research is in the training of frontline child protection workers. Although many of the theoretical issues regarding the construct of suggestibility are of less relevance here, knowledge of the impact of an interviewer, a questioning style, and context on an individual child's ability to report the details of an event is vital. Thus, for the protection of all, psychological research and training must be available to frontline workers. At the same time, it is important to examine the impact of the training on the effectiveness of the interviewers. Previous findings have suggested that training can be ineffective over time (Warren, Woodall, & Thomas, 1999); thus research is needed to determine why training loses its effectiveness. Is it a matter of training interviewers before they begin to practice, creating habits that are difficult to break? Is it that practice leads individuals to more inappropriate and risky interviewing styles in their attempts to assist increasing numbers of victimized children? Should periodic training occur to help practitioners overcome biases that might develop over time and influence their interviewing strategies? What sort of training will be most effective over time? Finally, what should be the content of such training? Will different interviewing techniques be needed for children, depending on the type and severity of maltreatment they experienced (Goodman & Schaaf, 1997)? How do we maximize disclosure while minimizing error? Answers to these important questions are far from clear. Nevertheless, researchers have come a long way toward an understanding of the suggestibility of children's memory. We know that children can be exceptionally accurate, even at very young ages, in reporting events they have witnessed. We know that universal statements about factors leading to suggestibility are not warranted. We know that interviews can be conducted in a sensitive, supportive environment that is not likely to elicit suggestibility from most children. However, we also know that mistakes

14.

CHILDREN'S SUGGESTIBILITY

421

can be made and it must be the goal of future research to clarify when and how such mis~ takes are made, and by whom. What is most clear, however, is the urgent need for more information on the suggestibility of children's memory.

REFERENCES Ackil, J. K., & Zaragoza, M. S. (1995). Developmental differences in eyewitness suggestibility and memory for source. Journal of Experimental Child Psychology, 60, 57-83. American Psychological Association (1991). Minutes of the Council of Representatives. Ameri~ can Psychologist, 46, 722. Baker~Ward, L., Gordon, B. N., Ornstein, P. A., Larus, D. M., &Clubb, P. A. (1993). Youngchil~ dren's long-term retention of a pediatric examination. Child Development, 64, 1519-1533. Baker~ Ward, L., Hess, T. M., & Flannagan, D. A. (1990). The effects of involvement on children's memory for events. Cognitive Development, 5, 55-69. Binet, A. (1900). La suggestibilite. Paris: Schleicher Freres. Brady, M.S., Poole, D. A., Warren, A. R., &Jones, H. R. (1999). Young children's responses to yes-no questions: Patterns and problems. Applied Developmental Science, 3, 47-57. Bruck, M., Ceci, S. J ., & Francoeur, E. (2000). Children's use of anatomically detailed dolls tore~ port genital touching in a medical examination: Developmental and gender comparisons. Journal of Experimental Psychology: Applied, 6, 74-83. Bruck, M., Ceci, S. J., Francoeur, E., & Renick, A. (1995). Anatomically detailed dolls do not facilitate preschoolers' reports of a pediatric examination involving genital touching. Journal of Experimental Psychology: Applied, 1, 95-109. Bruck, M., Ceci, S. J., & Hembrooke, H. (2002). The nature of children's true and false narra~ tives. Developmental Review, 22, 520-554. Bruck, M., Ceci, S. J., & Melnyk, L. (1997). External and internal sources of variation in the ere~ ation of false reports in children. Learning and Individual Differences, 9 (4), 289-316. Bruck, M., & Melnyk, L. (2004). Individual differences in children's suggestibility: A review and synthesis. Applied Cognitive Psychology, 18, 947-996. Bruck, M., Melnyk, L., & Ceci, S. J. (2000). Draw it again Sam: The effect of drawing on children's suggestibility and source monitoring ability. Journal of Experimental Child Psychology, 77, 169-196. Butler, S., Gross, J., & Hayne, H. (1995). The effect of drawing on memory performance in young children. Developmental Psychology, 31, 597-608. Carter, C., Bottoms, B. L., & Levine, M. (1996). Linguistic and socioemotional influences on the accuracy of children's reports. Law and Human Behavior, 20, 335-358. Cassel, W. S., & Bjorklund, D. F. (1995). Developmental patterns of eyewitness memory and suggestibility: An ecologically based short-term longitudinal study. Law and Human Behavior, 19, 507-532. Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and syn~ thesis. Psychological Bulletin, 113, 403-439. Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific analysis of children's testimony. Washington, DC: American Psychological Association. Ceci, S. J., Huffman, M. L. C., Smith, E., & Loftus, E. F. (1994). Repeatedly thinking about a nonevent: Source misattributions among preschoolers. Consciousness and Cognition, 3, 388-407. Ceci, S. J., Loftus, E. F., Leichtman, M.D., & Bruck, M. (1994). The possible role of source mis~ attributions in the creation of false beliefs among preschoolers. International Journal of Clinical and Experimental Hypnosis, 42, 304-320. Ceci, S. J., Ross, D. F., & Toglia, M.P. (1987). Suggestibility of children's memory: Psycholegal implications. Journal of Experimental Psychology, 116, 38-49.

422

MELNYK, CROSSMAN, SCULLIN

Cord6n, I. M., Saetermoe, C. L., & Goodman, G. S. (2005). Facilitating children's accurate responses: Conversational rules and interview style. Applied Cognitive Psychology, 19, 249266. Davis, S. L., & Bottoms, B. L. (2002). Effects of social support on children's eyewitness reports: A test of the underlying mechanism. Law and Human Behavior, 26, 185-215. DeLoache, J. S., & Marzolf, D.P. (1995). The use of dolls to interview young children: Issues of symbolic representation. Journal of Experimental Child Psychology, 60, 155-173. Dent, H. R., & Stephenson, G. M. (1979). An experimental study of the effectiveness of different techniques of quesrioning child witnesses. British Journal of Social and Clinical Psychology, 18,41-51. Dickinson, J. ]., Poole, D. A., & Bruck, M. (2005). Back to the future: A comment on the use of anatomical dolls in forensic interviews. Journal of Forensic Psychology Practice, 5, 63-74. Eberle, P., & Eberle, S. (1993). Abuse of innocence: The McMartin Preschool trial. Amherst, NY: Prometheus Books. Eigsti, 1.-M., & Cicchetti, D. (2004). The impact of child maltreatment on expressive syntax at 60 months. Developmental Science, 7, 88-102. Eisen, M. L., Qin, J., Goodman, G. S., & Davis, S. L. (2002). Memory and suggestibility in maltreated children: Age, stress arousal, dissociation, and psychopathology. Journal of Experimental Child Psychology, 83, 167-212. Fivush, R., & Schwarzmueller, A. (1995). Say it once again: Effects of repeated questions on children's event recall. Journal of Traumatic Stress, 8, 555-580. Fivush, R., & Schwarzmueller, A. (1998). Children remember childhood: Implications for childhood amnesia. Applied Cognitive Psychology, 12, 455-473. Flin, R. (1995). Children's testimony: Psychology on trial. In M. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 240-254). Thousand Oaks, CA: Sage. Flin, R., Boon, J., Knox, A., & Bull, R. (1992). The effect of a five-month delay on children's and adults' eyewitness memory. British Journal of Psychology, 83, 323-336. Garven, S., Wood, J. M., & Malpass, R. S. (2000). Allegations of wrongdoing: The effects of reinforcement on children's mundane and fantastic claims. Journal of Applied Psychology, 85, 38-49. Garven, S., Wood, J. M., Malpass, R. S., & Shaw, J. S. (1998). More than suggestion: The effect of interviewing techniques from the McMartin Preschool case. Journal of Applied Psychology, 83, 347-359. Gee, S., & Pipe, M.-E. (1995). Helping children to remember: The influence of object cues on children's accounts of a real event. Developmental Psychology, 31, 746-758. Ghetti, S., Goodman, G. S., Eisen, M. L., Qin, J., & Davis, S. L. (2002). Consistency in children's reports of sexual and physical abuse. Child Abuse & Neglect, 26, 977-995. Giles, J. W., Gopnik, A., & Heyman, G. D. (2002). Source monitoring reduces the suggestibility of preschool children. Psychological Science, 13, 288-291. Gilstrap, L. L., & Ceci, S. ]. (2005). Reconceptualizing children's suggestibility: Bidirectional and temporal properties. Child Development, 76, 40-53. Gilstrap, L. L., & Papierno, P. B. (2004). Is the cart pushing the horse? The effects of child characteristics on children's and adults' interview behaviors. Applied Cognitive Psychology, 18, 1059-1078. Gobbo, C., Mega, C., & Pipe, M. E. (2002). Does the nature of the experience influence suggestibility? A study of children's event memory. Journal of Experimental Child Psychology, 81, 502-530. Goodman, G. S., Batterman-Faunce, J. M., Schaaf, J. M., & Kenney, R. (2002). Nearly 4 years after an event: Children's eyewitness memory and adults' perceptions of children's accuracy. Child Abuse & Neglect, 26, 849-884.

14.

CHILDREN'S SUGGESTIBILITY

423

Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1994). Predictors of accurate and inaccurate memories of traumatic events experienced in childhood. Consciousness & Cognition, 3, 269-294. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, G. (1997). Children's reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge, and parental attachment. Applied Developmental Science, 1, 54-7 5. Goodman, G. S., Rudy, L., Bottoms, B., & Aman, C. (1990). Children's concerns and memory: Issues of ecological validity in the study of children's eyewitness testimony. In R. Fivush & ]. Hudson (Eds.), Knowingandrememberinginyoungchildren (pp. 249-284). New York: Cambridge University Press. Goodman, G. S., & Schaaf, J. M. (1997). Over a decade of research on children's eyewitness testimony: What have we learned? Where do we go from here? Applied Cognitive Psychology, 11, S5-S20. Goodman, G. S., Taub, E. P.,Jones, D.P. H., England, P., Port, L. K., Rudy, L., &Prado, L. (1992). Testifying in criminal court: Emotional effects on child sexual assault victims. Monographs of the Society for Research in Child Development, 57 (5, Serial No. 229). Goodman, G. S., Wilson, M. E., Hazan, C., & Reed, R. S. (1989). Children's testimony nearly four years after an event. Paper presented at the Annual Meeting of the Eastern Psychological Association, Boston, MA. Greenhoot, A. F. (2000). Remembering and understanding: The effects of changes in underlying knowledge on children's recollections. Child Development, 71, 1309-1328. Gudjonsson, G. H. (1984). A new scale of interrogative suggestibility. Personality and Individual Differences, 5, 303-314. Gudjonsson, G. H. (1987). A parallel form of the Gudjonsson Suggestibility Scale. British Journal of Clinical Psychology, 26, 215-221. Hamond, N. R., & Fivush, R. (1991). Memories of Mickey Mouse: Young children recount their trip to Disney World. Cognitive Development, 6, 433-448. Hershkowitz, I., Lamb, M. E. Sternberg, K. ]., & Esplin, P. W. (1997). The relationships among interviewer utterance type, CBCA scores, and the richness of children's responses. Legal and Criminological Psychology, 2, 169-176. Home Office (1992). Memorandum of good practice: On video recorded interviews with child witness for criminal proceedings. London: Author. Howe, M. L., Kelland, A., Bryant-Brown, L., & Clark, S. L. (1992). Measuring the development of children's amnesia and hypermnesia. In M. L. Howe, C.]. Brainerd, & V. F. Reyna (Eds.), Development oflong-term retention (pp. 56-102). New York: Springer-Verlag. Howe, M. L., O'Sullivan, J. T., & Marche, T. A. (1992). Toward a theory of the development oflong-term retention. In M. L. Howe, C.]. Brainerd, & V. F. Reyna (Eds.), Development of long-term retention (pp. 245-255). New York: Springer-Verlag. Hudson, J. A., & Fivush, R. (1991). As time goes by: Sixth graders remember a kindergarten experience. Applied Cognitive Psychology, 5, 347-360. Huffman, M. L., Crossman, A.M., & Ceci, S. ]. (1997). "Are false memories permanent?": An investigation of the long-term effects of source misattributions. Consciousness and Cognition, 6, 482-490. Hughes, M., & Grieve, R. (1980). On asking children bizarre questions. First Language, 1, 149-160. Hungerford, A. (2005). The use of anatomically detailed dolls in forensic investigations: Developmental considerations. Journal of Forensic Psychology Practice, 5, 75-87. Hyman, I. E., & Pentland, J. (1996). The role of mental imagery in the creation of false childhood memories. Journal of Memory and Language, 35, 101-117. Imhoff, M. C., & Baker-Ward, L. (1999). Preschoolers' suggestibility: Effects of developmentally appropriate language and interviewer supportiveness. Journal of Applied Developmental Psychology, 20, 407-429.

424

MELNYK, CROSSMAN, SCULLIN

Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28. Jones, M. (2004). Who was abused? New York Times. Retrieved June 6, 2005, from Lexis/Nexis. Kohnken, G. (1993). The cognitive interview: A step-by-step introduction. Unpublished manuscript. Lamb, M. E., Hershkowitz, I., Sternberg, K. J., Boat, B., & Everson, M.D. (1996). Investigative interviews of alleged sexual abuse victims with and without anatomical dolls. Child Abuse & Neglect, 20, 1251-1259. Lamb, M. E., Hershkowitz, I., Sternberg, K. J., Esplin, P. W., Hovav, M., Manor, T., & Yudilevitch, L. (1996). Effects of investigative utterance types on Israeli children's responses. International]ournal of Behavioral Development, 19,627-637. Lamb, M. E., Orbach, Y., Sternberg, K. J., Hershkowitz, I., & Horowitz, D. (2000). Accuracy of investigators' verbatim notes of their forensic interviews with alleged child abuse victims. Law & Human Behavior, 24,699-708. Lamb, M. E., Sternberg, K. J., & Esplin, P. W. (1995). Making children into competent witnesses: Reactions to the amicus brief in re Michaels. Psychology, Public Policy, and the Law, 1, 438-449. Leavitt, F. (1999). Is fantasy dangerous to the health of children? Credibility of memory research validating creation of false realities. Journal of Child Sexual Abuse, 8, 95-105. Leichtman, M.D., & Ceci, S. J. (1995). The effects of stereotypes and suggestions on preschoolers' reports. Developmental Psychology, 31, 568-578. Leippe, M. R., Manion, A. P., & Romanczyk, A. (1992). Eyewitness persuasion: How and how well do fact finders judge the accuracy of adults' and children's memory reports? Journal of Personality and Social Psychology, 63, 181-197. Lepore, S. J., & Sesco, B. (1994). Distorting children's reports and interpretations of events through suggestion. Applied Psychology, 79, 108-120. Lindberg, M.A., Jones, S., McComas Collard, L., & Thomas, S. W. (2001). Similarities and differences in eyewitness testimonies of children who directly versus vicariously experience stress. Journal of Genetic Psychology, 162,314-333. Lyon, T. (1999). The new wave in children's suggestibility research: A critique. Cornell Law Review, 84, 1004-1087. Lyon, T. D. (2002). Applying suggestibility research to the real world: The case of repeated questions. Law and Contemporary Problems, 65, 97-126. Marche, T. A. (1999). Memory strength affects reporting of misinformation. Journal of Experimental Child Psychology, 73, 45-71. Marche, T. A., & Howe, M. L. (1995). Preschoolers report misinformation despite accurate memory. Developmental Psychology, 31,554-567. McFarlane, F., & Powell, M. B. (2002). The Video Suggestibility Scale for Children: How generalizable is children's performance to other measures of suggestibility? Behavioral Sciences & the Law, 20, 699-716. McGough, L. (2002). Good enough for government work: The constitutional duty to preserve forensic interviews of child victims. Law & Contemporary Problems, 65, 179-208. Melnyk, L. (2002). The influence of imagery, timing, and individual differences on the accuracy of children's recall. Unpublished doctoral dissertation, McGill University, Montreal, Canada. Melnyk, L., & Bruck, M. (2004). Timing moderates the effects of repeated suggestive interviewing on children's eyewitness memory. Applied Cognitive Psychology, 18, 613-631. Memon, A., Wark, L., Bull, R., & Koehnken, G. (1997). Isolating the effects of the cognitive interview techniques. British Journal of Psychology, 88, 179-197. Merritt, K. A., Ornstein, P. A., & Spieker, B. (1994). Children's memory for a salient medical procedure: Implications for testimony. Pediatrics, 94, 17-23. Miles, K. L., Powell, M. B., & Stokes, M.A. (2004). A comparison of the effectiveness of two suggestibility paradigms in predicting preschoolers' tendency to report a non-experienced event. Applied Cognitive Psychology, 18, 1021-1036.

14. CHILDREN'S SUGGESTIBILITY

425

Mulder, M. R., & Vrij, A. (1996). Explaining conversation rules to children: An intervention study to facilitate children's accurate responses. Child Abuse & Neglect, 20, 623-631. Orbach, Y., Hershkowitz, I., Lamb, M. E., Sternberg, K. ]., Esplin, P. W., & Horowitz, D. (2000). Assessing the value of scripted protocols for forensic interviews of alleged abuse victims. Child Abuse and Neglect, 24, 733-752. Ornstein, P. A., Baker-Ward, L., Gordon, B. N., & Merritt, K. A. (1997). Children's memory for medical experiences: Implications for testimony. Applied Cognitive Psychology, 11, S87-S104. Pears, K. C., & Fisher, P. A. (2005). Emotion understanding and theory of mind among maltreated children in foster care: Evidence of deficits. Development & Psychopathology, 17, 47-65. Perry, N. W., McAuliff, B. D., Tarn, P., Claycomb, L., Dostal, C., & Flanagan, C. (1995). When lawyers question children: Is justice served? Law & Human Behavior, 19, 609-629. Peterson, C., & Bell, M. (1996). Children's memory for traumatic injury. Child Development, 67, 3045-3070. Peterson, C., Moores, L., & White, G. (2001). Recounting the same event again and again: Consistency across multiple interviews. Applied Cognitive Psychology, 15, 353-371. Pezdek, K., & Roe, C. (1995). The effect of memory trace strength on suggestibility. Journal of Experimental Child Psychology, 60, 116-128. Pipe, M. E., Gee, S., Wilson, J. C., & Egerton, J. M. (1999). Children's recall1 or 2 years after an event. Developmental Psychology, 35, 781-789. Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. Poole, D. A., & Lindsay, D. S. (2001). Children's eyewitness reports after exposure to misinformation from parents. Journal of Experimental Psychology: Applied, 7, 2 7-50. Poole, D. A., & Lindsay, D. S. (2002). Reducing child witnesses' false reports of misinformation from parents. Journal of Experimental Child Psychology, 81, 117-140. Poole, D. A., & White, L. T. (1991). Effects of question repetition on the eyewitness testimony of children and adults. Developmental Psychology, 27, 975-986. Poole, D. A., & White, L. T. (1993). Two years later: Effect of question repetition and retention interval on the eyewitness testimony of children and adults. Developmental Psychology, 29, 844-853. Poole, D. A., & White, L. T. (1995). Tell me again and again: Stability and change in therepeated testimonies of children and adults. In M. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 2243). Thousand Oaks, CA: Sage. Quas, J. A., Bauer, A., & Boyce, W. T. (2004). Physiological reactivity, social support, and memory in early childhood. Child Development, 75, 797-814. Quas, J. A., Goodman, G. S., Bidrose, S., Pipe, M. E., Craw, S., & Ahlin, D. S. (1999). Emotion and memory: Children's long-term remembering, forgetting, and suggestibility. Journal of Experimental Child Psychology, 72, 235-270. Ranalli, R. (2004, May 1). Amirault is freed from prison conviction stands in day-care abuse. The Boston Globe. Retrieved June 6, 2005, from Lexis/Nexis. Roberts, K. P. (2000). An overview of theory and research on children's source monitoring. In K. P. Roberts & M. Blades (Eds.), Children's source monitoring (pp. 11-57). Mahwah, NJ: Lawrence Erlbaum Associates. Roberts, K. P., Lamb, M. E., & Sternberg, K. J. (2004). The effects of rapport-building style on children's reports of a staged event. Applied Cognitive Psychology, 18, 189-202. Rudy, L., & Goodman, G. S. (1991). Effects of participation on children's reports: Implications for children's testimony. Developmental Psychology, 27, 527-538. Salmon, K., Bidrose, S., & Pipe, M.-E. (1995). Providing props to facilitate children's event reports: A comparison of toys and real items. Journal of Experimental Child Psychology, 60, 174-194.

426

MELNYK, CROSSMAN, SCULLIN

Salmon, K., & Pipe, M.-E. (1997). Props and children's event reports: The impact of a one year delay.Joumal of Experimental Child Psychology, 65, 261-292. Saywitz, K., Goodman, G. S., Nicholas, E., & Moan, S. (1991). Children's memories of physical examinations involving genital touch: Implications for reports of child sexual abuse. Journal of Consulting and Clinical Psychology, 59, 682-691. Scullin, M. H., & Ceci, S. J. (2001). A suggestibility scale for children. Personality and Individual Differences, 30, 843-856. Scullin, M. H., Kanaya, T., & Ceci, S. J. (2002). Measurement of individual differences in children's suggestibility across situations. Journal of Experimental Psychology: Applied, 8, 233-246. Shapiro, L. R., Blackford, C., & Chen, C.-F. (2005). Eyewitness memory for a simulated misdemeanor crime: The role of age and temperament in suggestibility. Applied Cognitive Psychology, 19, 267-289. State v. Michaels, Superior Court, Essex County, New Jersey (1988). State v. Michaels, 264 N.J. Super 579, 625 A.D. 2d 489 (N.J. Super Ad, 1993). State v. Michaels, 136 N.J. 299, 642 A.2d 1372 (N.J., 1994). State v. Robert Fulton Kelly, Jr., Superior Criminal Court, Pitt County, North Carolina, #91-CRS4250-4363 (1991-1992). Stem, C., & Stem, W. (1999). Recollection, testimony, and lying in early childhood (J. T. Lamiell, transl.). Washington, DC: American Psychological Association (original work published 1908). Stem, W. (1910). Abstracts oflectures on the psychology of testimony and on the study of individuality. American Journal of Psychology, 21, 270-282. Steward, M.S., & Steward, D. S. (1996). Interviewing young children about body touch and handling. Monographs of the Society for Research in Child Development, 61 (4-5, Serial No. 248). Sutherland, R., Pipe, M.-E., Schick, K., Murray, J., & Gobbo, C. (2003). Knowing in advance: The impact of prior event information on memory and event knowledge. Journal of Experimental Child Psychology, 84, 244-263. Thierry, K. L., & Spence, M. J. (2002). Source-monitoring training facilitates preschoolers' eyewitness memory performance. Developmental Psychology, 38, 428-437. Tobey, A., & Goodman, G. S. (1992). Children's eyewitness memory: Effects of participation and forensic context. Child Abuse & Neglect, 16, 779-796. Van Voorhees, E., & Scarpa, A. (2004). The effects of child maltreatment on the hypothalamicpituitary-adrenal axis. Trauma, Violence, & Abuse, 5, 333-352. Walker, A. G. (1994). Handbook on questioning children: A linguistic perspective. Washington, DC: American Bar Association Center on Children and the Law. Walker, N. E., & Hunt, J. S. (1998). Interviewing child victim-witnesses: How you ask is what you get. In C. P. Thompson, D. J. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 55-88). Mahwah, NJ: Lawrence Erlbaum Associates. Warren, A. R., & Lane, P. (1995). The effects of timing and type of questioning on eyewitness accuracy and suggestibility. In M. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 44-60). Thousand Oaks, CA: Sage. Warren, A. R., & Woodall, C. E. (1999). The reliability of hearsay testimony: How well do interviewers recall their interviews with children? Psychology, Public Policy, & Law, 5, 355-371. Warren, A. R., Woodall, C. E., Hunt, J. S., & Perry, N. W. (1996). Inducing resistance to suggestibility in children. Law and Human Behavior, 15, 273-285. Warren, A. R., Woodall, C. E., & Thomas, M. (1999). Assessing the effectiveness of a training program for interviewing child witnesses. Applied Developmental Science, 3, 128-135. Warren, A. R., Woodall, C. E., Thomas, M., Nunno, M., Keeney, J., Larson, S., et al. (1999). Assessing the effectiveness of a training program for interviewing child witnesses. Applied Developmental Science, 3, 128-135.

14.

CHILDREN'S SUGGESTIBILITY

427

Whitcomb, D. (1992). When the victim is a child (2nd ed.). Washington, DC: National Institute of Justice. Wilson, J. C., & Pipe, M.-E. (1989). The effects of cues on young children's recall of real events. New Zealand Journal of Psychology, 18, 65-70. Zajac, R., & Hayne, H. (2003). I don't think that's what really happened: The effect of crossexamination on the accuracy of children's reports. Journal of Experimental Psychology: Applied, 9, 187-195.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

15 Enhancing Performance: Factors Affecting the Informativeness of Young Witnesses Michael E. Lamb University of Cambridge

Yael Orbach National Institute of Child Health and Human Development

Amye R. Warren University of Tennessee at Chattanooga

Phillip W. Esplin Private Practice, Phoenix, AZ Irit Hershkowitz University of Haifa

Over the past 25 years, many researchers have studied children's testimony with a particular focus on children's capacities to provide reliable and valid information about their past experiences (see reviews by Ceci & Bruck, 1995; Ceci, Leichtman, & Putnick, 1992; Dent & Flin, 1992; McGough, 1994; Memon & Bull, 1999; Perry & Wrightsman, 1991; Poole & Lamb, 1998; Spencer & Flin, 1990). Initially, most researchers conducted controlled studies in the laboratory, but their ecological validity was often questioned (Doris, 1991; Lamb & Thierry, 2005). Since then, studies conducted in both field and laboratory circumstances have focused more narrowly on issues of particular relevance to forensic application-and have helped generate a remarkable consensus about children's competencies and limitations. 429

430

LAMBET AL.

In brief, the research shows that although children clearly can remember incidents they have experienced, the relationship between age and memory is complex; a variety of factors influence the quality of information provided. For our present purposes, perhaps the most important of these factors pertain to the interviewers' ability to elicit information and the child's willingness and ability to express it, rather than the child's ability to remember it. Like adults, children can be informative witnesses, and a variety of professional groups and experts have recognized this, offering recommendations regarding the most effective ways of conducting forensic or investigative interviews (e.g., American Professional Society on the Abuse of Children [APSAC], 1990, 1997; Bull, 1992, 1995, 1996; Fisher & Geiselman, 1992; Jones, 1992; Lamb, Sternberg, & Esplin, 1994, 1995, 1998; Home Office & Department of Health, 1992; Orbach, Hershkowitz, Lamb, Sternberg, Esplin, & Horowitz, 2000; Poole & Lamb, 1998; Raskin & Esplin, 1991; Raskin & Yuille, 1989; Sattler, 1998). As Poole and Lamb (1998) pointed out, these books and articles reveal a substantial degree of consensus regarding the ways in which investigative interviews should be conducted and a remarkable convergence with the conclusions suggested by a close review of the experimental and empirical literature. Clearly, it is often possible to obtain valuable information from children, but doing so requires careful investigative procedures as well as a realistic awareness of their capacities and tendencies. Our goal in this chapter is to summarize current understanding of the factors that influence children's ability to provide accurate information about events they have experienced and to demonstrate how application of this knowledge in forensic settings in fact enhances their informativeness. In the next section, we thus examine the ways that rapport, language and communicative capacity, memory development, and suggestibility shape children's informativeness. In the course of our selective literature review, we also discuss aspects of our own research program, mostly conducted with the use of transcripts of actual forensic interviews with alleged child abuse victims, to illustrate how informative children can be when interviewers build on their capacities and strengths. Most of the interviews had been conducted in Israel by youth investigators who are statutorily mandated to conduct all investigative interviews of children (Sternberg, Lamb, & Hershkowitz, 1996), and in the United States, Britain, and Sweden by social workers, sheriffs, or police officers.

GENERAL RESEARCH PROCEDURE For research purposes, coders review portions of the transcripts concerned with substantive issues and categorize each interviewer utterance. Most utterances are of one of the following types:

1.

Invitations (using questions, statements or imperatives) for an open-ended response from the child. Such utterances do not delimit the child's focus, except in the most general way (for example: And then what happened?)

15.

2.

3.

4.

5.

FACTORS AFFECTING YOUNG WITNESSES

431

Facilitators, which are utterances like o.k., restatements of the child's previous utterance, and nonsuggestive words of encouragement that are designed to prompt continuation of the child's narrative. Directive utterances, which focus the child's attention on presumed details or aspects of the event that the child had previously mentioned, typically in the form of a wh- question (e.g., What color was that shirt?). Option-posing utterances, which focus the child's attention on details or aspects of the account that the child had not previously mentioned by asking the child to choose among options explicitly or implicitly provided by the interviewer (e.g., Did he touch you inside or outside?). Suggestive utterances, stated in such a way that the interviewer strongly communicates what response is expected (e.g., That hurt, didn't it?) or assumes details that have not been revealed by the child (e.g., Did it hurt when he touched you?, when the child had mentioned no touching).

The three types of focused utterances (directive, option-posing, suggestive) thus lie along a continuum of risk, varying with respect to the degree of suggestive influence they exert on children's responses while tapping the child's reconstructive and recognition memory. As we show below, option-posing and suggestive utterances are especially risky because of their tendency to elicit inaccurate information. Raters also count the number of new details conveyed by the child. By definition, details involve the identification and description of individuals, objects, events, or actions relevant to the alleged incident. As indicated below, there are systematic associations between the types of prompts used by investigators and the quality and quantity of information reported by children.

FACTORS INFLUENCING CHILDREN'S COMPETENCE

Much of the debate over the competency of young witnesses has focused on how children distinguish fantasy from reality, express themselves clearly and unambiguously, understand what interviewers say, remember events, overcome their shyness or reticence with intimidating strangers, and resist suggestions that may affect their competence as informants. As we show in this chapter, these are legitimate concerns. The problem lies not only in children's weaknesses, however, but in the failure of communication because more mature adult interlocutors do not recognize or adapt to children's capacities, tendencies, and limitations. Although we have organized this section in relation to mental processes in children, it should be clear that the problem is co-created by interviewers and children. Linguistic, cognitive, and social immaturities do not make children incompetent witnesses, but an understanding of their capacities and limitations should influence the ways in which children are interviewed and the ways in which their accounts are interpreted. In the following sections, we focus our attention on shyness/rapport, the ability to distinguish fantasy from reality, linguistic and communicative capacities and difficulties, memory development, and suggestibility. We do not deal with social factors

432

LAMBET AL.

affecting children's interview responses separately, although we recognize that such factors (e.g., compliance, expectations that adults have greater knowledge and are honest) pervade the interview situation. Instead, we attempt to address some of the related social factors within the other sections.

Shyness and Rapport Most adults are aware that children are often reticent with strangers and recognize the need to establish rapport when initiating interaction with an unfamiliar child, especially when the topics are stressful or embarrassing. Forensic interviewers are thus routinely encouraged to establish rapport with alleged victims before seeking to elicit information about the suspected incidents of abuse. Despite this consensus, many forensic interviewers fail to make more than perfunctory efforts to establish rapport before broaching the substantive issue under investigation (Sternberg, Lamb, Esplin, & Baradaran, 1999; Warren, Woodall, Hunt, & Perry, 1996). Although it seems intuitively obvious that rapport-building is critically important, we actually know very little about how much and what kind of rapport-building is necessary or effective and almost nothing about how rapport-building needs may differ depending on the age and other characteristics of the children interviewed. Surprisingly, few researchers have examined different kinds of rapport-building techniques or compared interviews with and without attempts to build rapport. Field research by Sternberg et al., (1997) found that rapport-building with the use of open-ended questions (invitations) about the child's everyday life and a particular past event (such as a recent birthday or holiday) helped interviewers elicit more abuse-relevant information than closedended specific rapport-building questions did, but as this was a field study, it did not include a no-rapport control group. The complexities of rapport-building were illustrated by Davies, Westcott, and Horan (2000), who found that certain types of abuse-relevant information were more likely to be elicited when rapport-building was shorter (less than 8 minutes). They speculated that longer rapport-building phases may have reduced children's attention to the later abuse questioning, or that interviewers who spent more time building rapport had less time to spend questioning the children about the alleged abuse. Davis et al. also suggested that children who required more rapport-building may have been less comfortable, less prepared, and more reluctant to provide information. Thus, rapport-building may not work for those who need it the most, and lengthy rapport-building sessions may be counterproductive. On a more positive note, Wood, McClure, and Birch (1996) found that children who seemed reluctant and were uncommunicative at the beginning of an interview were more likely to talk or open up to the interviewer later when rapport-building was conducted well. Clearly, further research is needed on the type and amount of rapport-building that best promotes truthful and detailed disclosures. In the meantime, we strongly encourage interviewers to build rapport with children by asking open-ended questions about neutral, everyday events before questioning them about sensitive topics.

15.

FACTORS AFFECTING YOUNG WITNESSES

433

Fantasy Children over 6 years of age appear similar to adults in their ability to discriminate be, tween events of internal (imagined) and external (experienced) origin (Johnson & Foley, 1984; Lindsay &Johnson, 1987; Roberts, 2000; Roberts & Blades, 1995), and the extent to which younger children have difficulty discriminating between fantasy and reality is poorly identified. The presence of fantastic elements in children's accounts of abuse is affected by the presence of props (such as toys or dolls) usually associated with fantasy (Thierry, Lamb, Orbach, & Pipe, 2005) or by interviewers prompting children to imag, ine or pretend. As a result, forensic investigators have been urged to avoid having such props present during investigative interviews and to avoid using such expressions (Lamb et al., 1995, 1998; Poole & Lamb, 1998).

Language and Communication Young children often do not articulate individual sounds consistently even after they seem to have mastered them (Reich, 1986), so it is not uncommon for interviewers to misunderstand children, especially preschoolers. In addition, the vocabularies of young children are much more limited and less descriptive than those of adults (Brown, 1973; Dale, 1976; de Villiers & de Villiers, 1999), and their statements are likely to lack adjec, tival and adverbial modifiers. Misunderstandings also occur because children's rapid vocabulary growth often leads adults to overestimate their linguistic capacities. Despite their apparent maturity, young children-especially preschoolers-frequently use words before they know their conventional adult meaning, they may even use words that they do not understand at all, and they may poorly understand some apparently simple concepts, such as any, some, touch, yesterday, and before (Harner, 1975; Walker, 1994). Unlike adults and older children, furthermore, young children cannot draw upon an array of past experiences to enrich and clarify their descriptive accounts (Johnson & Foley, 1984). In the process of learning words and the rules for combining words into sentences, children are also learning how to participate in conversations. They must learn how to stay on topic, how to adapt their speech appropriately to different audiences (e.g., a strange interviewer who does not know their family members and was not present for the event in question), and how to structure coherent narratives about past events (Warren & McCloskey, 1997). The challenge confronting investigators is to obtain organized accounts that are sufficiently rich in descriptive detail to permit an understanding of the children's testimony. Researchers have shown that the accuracy of children's accounts is greatly influenced by the linguistic style and the complexity of the language addressed to them by investigators (Carter, Bottoms, & Levine, 1996; Imhoff & Baker-Ward, 1999; Perry, McAuliff, Tam, Claycomb, Dostal, & Flanagan, 1995). Children are often asked to negate adult statements or to confirm multifaceted summaries of their accounts (e.g., Is it not true that ... ?) and expected to understand unfamiliar words and syntactically complex or ambiguous compound sentences (Dent, 1982; Pea, 1980; Perry & Wrightsman, 1991;

434

LAMBET AL.

Saywitz, 1988; Walker, 1993; Walker & Hunt, 1998; Warren et al., 1996). For example, Brennan and Brennan (1988) showed that fewer than two-thirds of the questions addressed to 6- to 15-year-olds in court were comprehensible to their peers. Perry and colleagues (1995) similarly showed that kindergarten through college-age students had much more difficulty correctly answering complex questions as opposed to more simply phrased versions about the same witnessed event. More importantly, the kindergartners did not even recognize that they misunderstood the complex questions, responding at chance levels on a task measuring how well they monitored their own comprehension. Unfortunately, systematic analyses show that inappropriate questioning strategies are not limited to the courtroom, but characterize the vast majority of forensic interviews (e.g., Lamb, Hershkowitz, Sternberg, Boat, & Everson, 1996; Lamb, Hershkowitz, Sternberg, Esplin, et al., 1996; Sternberg, Lamb, Hershkowitz, Esplin, et al., 1996; Warren et al., 1996). Years of analyzing forensic interviews have led us, along with forensic linguists such as Walker (1993), to question widespread assumptions about children's linguistic inabilities and to focus more on the competence, perspective-taking abilities, and linguistic styles of investigators. The more impoverished the children's language, the greater the likelihood that their statements will be misinterpreted or that children will misinterpret the interviewers' questions and purposes (King & Yuille, 1987; Perry & Wrightsman, 1991; Walker, 1993). When interviewers misrepresent what children say, furthermore, they are infrequently corrected, and thus the mistakes, rather than the correct information, may be reported later in the interview (Roberts & Lamb, 1999). Following up on this finding in the laboratory, Hunt and Borgida (2001) similarly found that disagreement was uncommon, with adults significantly more likely than children to disagree when interviewers distorted their answers. In subsequent interviews, 4- and 5-year-old children were more likely than older children or adults to incorporate the interviewers' earlier distortions into their later reports about witnessed events, suggesting that their memories of the event might have been distorted. This further underscores the extent to which the interviewers' behavior-particularly their vocabularies, the complexity of their utterances, their suggestiveness, and their success in motivating children to be informative and forthcoming-profoundly influences the course and outcome of their interviews. Children's accounts of abusive experiences are also influenced by social or pragmatic aspects of communication. For example, when asked questions such as Do you remember his name? or Do you know why you are here today?, Can you show me where he touched you?, older children usually read between the lines and provide the desired information, whereas younger children may simply answer yes or no (Walker & Warren, 1995; Warren et al., 1996). Young witnesses are typically unaware of the amount and type of information being sought by forensic investigators and are unaccustomed to being viewed as informants rather than novices being tested about the quality of their knowledge. As a result, interviewers need to communicate their needs and expectations clearly, motivating children to provide as much information as they can. Fortunately, it is clear that interviewers can train children to be informative by making clear how much detail is desired before beginning to ask questions about substantive issues. For example, Sternberg et al. (1997) asked forensic interviewers to follow very

15.

FACTORS AFFECTING YOUNG WITNESSES

435

specific scripts in the rapport~building phases of their interviews. In half of the inter~ views, investigators who were naive with respect to the experimental hypotheses used a script containing many open~ ended utterances to establish rapport, whereas in the other interviews the same investigators used a script involving many focused questions. Both introductory scripts took about 7 minutes to complete, and both ended with the identi~ cal open~ended utterance used to initiate the substantive phase of the interview. Chil~ dren who had been trained in the open~ended condition provided two and one~half times as many details and words in response to the first substantive utterance as did chil~ dren in the direct introduction condition. Children in the open~ended condition con~ tinued to provide more information in response to subsequent invitations, suggesting that the initial training was successful in conveying the interviewers' desire for detailed description of the alleged events. Evidently, children who had the opportunity to prac~ tice providing lengthy narrative responses to open~ended questions in the introductory phase of the interview continued this pattern after the interviewers shifted focus to the alleged incidents of abuse. Unfortunately, similarly dramatic effects were not evident in a later analog study (Roberts, Lamb, & Sternberg, 2004), perhaps because the analog in~ cidents (which involved dressing up and being photographed in costume) were not as meaningful, complex, and emotionally engaging as these incidents of abuse. We were quite impressed that the richness of children's accounts of experienced in~ cidents of abuse could be influenced by the interrogatory style modeled in brief intro~ ductory segments of the interview and that a single carefully worded prompt could elicit so much information from the children in both conditions. In the study by Sternberg et al. (1997), the first substantive question yielded an average of 38 details from children in~ in the direct introduction condition and 91 details from children in the open~ended traduction conditions, whereas in earlier unscripted interviews by the same group of in~ vestigators, the average invitation yielded only 5 details, and the first invitation yielded an average of only 6 details (Lamb, Hershkowitz, Sternberg, Esplin, et al., 1996). The findings reported by Sternberg et al. (1997) are consistent with the results of laboratory analog studies suggesting that motivational and contextual factors play an important role in shaping children's reports of experienced events (Campara, Wagner, & Saywitz, 2001; Paris, 1988; Saywitz, Nathanson, Snyder, & Lamphear, 1993; Saywitz & Snyder, 1996). They also suggest that, even in authentic forensic interviews, it is possible to teach response styles that enhance the richness of information provided by children by affording them an opportunity to practice providing detailed narrative accounts of ex~ perienced events and by reinforcing this style in the presubstantive portion of the inter~ views. Richly detailed accounts of abusive events facilitate the more effective investiga~ tion of crimes and provide child protection workers with more information upon which to base their evaluations. The results of this study also highlight the value of having interviewers clearly communicate their expectations concerning the child's role. training in the study by Sternberg et al. (1997) influenced Although the open~ended the children's responses, it had little effect on the interviewers' style of questioning after the first substantive question was posed. In other words, even when children provided substantive question, interviewers did not lengthy responses to the first open~ended continue to ask open~ended questions, but rather shifted to more focused questions.

436

LAMBET AL.

This unexpected finding suggested that it might be valuable to specify additional openended questions to be asked throughout the substantive phase of the interview. Interviewers clearly have difficulty internalizing recommended interview techniques and may need more explicit guidelines than those typically provided in training sessions or manuals, however intensive. We thus developed increasingly detailed scripts for the entire interview (including substantive and nonsubstantive sections). We discuss this research in the next section.

Memory Research on memory development suggests that, as children grow older, the length, informativeness, and complexity of their recall narratives increase, but the basic structure remains the same (Davies, Tarrant, & Flin, 1989; Flin, Boon, Knox, & Bull, 1992; Nelson & Gruendel, 1981; Saywitz, 1988). In general, young children tend to provide briefer accounts of their experiences than do older children and adults, but their accounts are quite accurate (e.g., Goodman & Reed, 1986; Johnson & Foley, 1984; Marin, Holmes, Guth, & Kovac, 1979; Oates & Shrimpton, 1991). As time passes, some information is forgotten by children, just as it is forgotten by adults (Flin et al., 1992). Errors of omission are much more common than errors of commission among both adults and children (Oates &Shrimpton, 1991; Steward, 1993) but are a special problem where children are concerned because their accounts-especially their recall narratives-are often so brie£ It is also important to distinguish between memory performance and memory capacity. Young children's accounts may be brief not only because their memories are relatively poor or because their limited experiences do not provide a rich network of associations from which to draw analogies or metaphors, but also because their vocabularies are much more limited and less elaborate than those of adults and because they may not be motivated to reveal what they do remember. Whenever events recur with any regularity, both children and adults tend to blur distinctions among incidents and establish script memories (representations of averaged or typical events rather than particular incidents). Accounts based on scripts are likely to contain fewer distinctive details than are memories of discrete incidents (Nelson & Gruendel, 1981), and the passage of time between experience and recall increases the tendency to rely on scripts (Myles-Worsley, Cromer, & Dodd, 1986). Scripts are useful because they help individuals to focus on and remember the important features of repetitive events or sequences while enabling them to ignore less central or repetitious elements (Nelson, 1986; Shank & Abelson, 1977). In addition, scripts may provide a temporal sequence or structure that makes the accounts of specific experiences more comprehensible. But scripts (like stereotypes) have disadvantages too, because they tend to be brief or skeletal and may incorrectly describe particular events when general knowledge about a class of events influences reports of specific events. For example, the 5- and 6-year-olds studied by Martin and Halverson (1983) incorrectly remembered the gender of a character who played a nontraditional gender role, and 5- and 7-year-olds studied by McCartney and Nelson (1981) embellished restatements of stories with items and events

15.

FACTORS AFFECTING YOUNG WITNESSES

437

that were part of their own scripts. The tendency to do this generally declines with age (Collins, 1970; Collins & Wellman, 1982; Collins, Wellman, Keniston, & Westby, 1978), and script-based errors can be reduced by pre-interview counseling or instruction (Saywitz & Snyder, 1993). Children also tend to remember unusual events better than specific events that are congruent with their general or script memories (Davidson, 1991). In investigative contexts, children may nonetheless provide scripted accounts because they are unaware of the level of detail needed by forensic investigators. Interviewers must thus communicate their needs for narrative accounts of specific incidents and motivate children to be maximally informative witnesses. When children report being abused more than once, interviewers can focus the child on specific events by using questions like Tell me about the last time or Tell me about the time in Boston. When considering the ways in which information can be elicited from children (or other informants), the distinction between recall and recognition testing is crucial. When adults and children are asked to describe events with free recall prompts (Tell me everything you remember ... ), their accounts may be brief and sketchy, but are more likely to be accurate. When provided with open-ended prompts like Tell me more about that or And then what happened?, children often report additional details. When interviewers prompt with leading questions such as Did he have a beard?, Did he touch you with his private?, or Did this happen in the day or in the night?, however, they shift from recall to recognition testing, and the probability of error rises dramatically (e.g, Dent, 1982, 1986; Dent &Stephenson, 1979; Hutcheson, Baxter, Telfer, & Warden, 1995; Lamb &Fauchier, 2001; Oates & Shrimpton, 1991, Orbach & Lamb, 2001). When memory is probed with the use of open-ended prompts, respondents attempt to provide as much relevant information as they remember, whereas children may have to confirm or reject information provided by the interviewer when focused questions tapping recognition memory are asked. Recognition probes refocus the child on domains of interest to the investigator and exert greater pressure to respond, whether or not the respondent is sure of the response. Recognition probes are more likely to elicit erroneous responses in eyewitness contexts because of response biases (i.e., tendencies to say "yes" or "no" without reflection) and false recognition of details that were only mentioned in previous interviews or are inferred from the gist of the experienced events (Brainerd & Reyna, 1996). Effective interviewers should thus maximize the reliance on free recall by offering open-ended prompts so as to minimize the risk of eliciting erroneous information. Free recall reports are not always accurate, of course, especially when the events occurred long before the interview or there have been opportunities for either pre-event (Leichtman & Ceci, 1995) or post-event (Leichtman & Ceci, 1995; Poole & Lindsay, 1995, 1997; Poole & White, 1993; Warren & Lane, 1995) contamination, but they are likely to be much more accurate than reports elicited with the use of recognition cues or prompts. Analyses of forensic interviews have shown quite consistently that open-ended or invitational questions yield responses that are three to four times longer and three times richer in relevant details than responses to focused interviewer utterances (Lamb, Hershkowitz, Sternberg, Boat, et al., 1996; Lamb, Hershkowitz, Sternberg, Esplin, et al., 1996; Lamb et al., 2003; Orbach et al., 2000; Sternberg, Lamb, Hershkowitz, Esplin, et al., 1996; Sternberg, Lamb, Orbach, et al., 2001). The superiority of open-ended utterances

438

LAMBET AL.

is apparent regardless of the age of the children being interviewed. Because focused questions test recognition rather than recall, furthermore, the information elicited is more likely to be inaccurate. Unfortunately, focused utterances are much more common in the field than openended questions are. In the field sites we have studied, for example, more than 80o/o of the interviewer utterances were focused, whereas only 6o/o were invitations. Research in the United States, Britain, Sweden, and Israel shows that overreliance on focused questions is evident regardless of the children's age, the nature of the offenses, the professional background of the interviewers, or the use of props and tools like anatomical dolls (e.g., Cederborg et al., 2000; Craig, Sheibe, Kircher, Raskin, & Dodd, 1996; Lamb, Hershkowitz, Sternberg, Boat, et al.,1996; Lamb, Hershkowitz, Sternberg, Esplin, et al., 1996; Orbach et al., 2000; Sternberg et al., 2001; Sternberg et al., 1996; Sternberg, Lamb, Davies, & Westcott, 2001). Overall, narrative responses are more desirable because they represent children's relatively spontaneous free recall, they are more detailed, and they are more accurate. Consistent evidence that interviewers tend to continue using focused rather than open-ended questions, despite extensive training, prompted the development of a fully structured investigative interview protocol designed to translate empirically based research guidelines into a practical tool to be used by investigators conducting forensic interviews (Orbach et al., 2000). The National Institute of Child Health and Human Development's (NICHD) investigative interview protocol covers all phases of the investigative interview and is designed to translate research-based recommendations into operational guidelines in order to enhance the retrieval of informative, complete, and accurate accounts of alleged incidents of abuse by young victimwitnesses. This is accomplished by the creation of a supportive interview environment (before substantive rapport-building), adapting interview practices to children's developmental levels and capabilities (e.g., minimizing linguistic complexity and avoiding interruptions), preparing children for their tasks as information providers (by clarifying the rules of communication and training children to report event-specific episodic memories), and maximizing the interviewers' reliance on utterance types (e.g., invitations) that tap children's free recall. When following the protocol, interviewers maximize the use of open-ended questions and probes, introduce focused questions only after exhausting open-ended questioning modes, use option-posing questions (including yes/no questions) only to obtain essential information later in the interview, and eliminate suggestive practices. Interviewers are also encouraged to use information provided by the children themselves as cues to promote further free-recall retrieval. In essence, the protocol is thus designed to maximize the amount of information elicited with the use of recall prompts because information elicited in this way is more likely to be accurate. In addition, the structured interview protocol minimizes opportunities for contamination of the children's accounts. Analyses revealed dramatic improvements in the organization of interviews, the quality of questions asked by interviewers, and the quality of information provided by children when Israeli youth investigators followed the protocol when interviewing 50 4- to 13-year-old alleged victims of sexual abuse (Orbach et al., 2000). Almost all of the children interviewed using the structured protocol provided a narrative account of the

15.

FACTORS AFFECTING YOUNG WITNESSES

439

alleged abuse in response to the first invitation, and the interviewers offered more than five times as many open~ended invitations as they did in comparable interviews con~ ducted before the structured protocol was introduced. The number of option~posing questions dropped by almost 50% as well, and much more of the information was ob~ tained with the use of free recall rather than investigator~directed recognition probes in the protocol~guided interviews. Children in the protocol condition provided propor~ tionally more of the total number of details in their first narrative response than did chil~ dren in the nonprotocol condition, and they provided significantly more information be~ fore being asked the first option~posing question. More of the details they provided were ended prompts, whereas fewer were elicited by directive, option~posing, elicited by open~ and suggestive utterances. Although this was not studied systematically, the interviewers using the NICHD interview protocol also became better critics of their own and of their colleagues' interviews. Similar results were obtained when investigative interviews conducted by police officers in the western United States were studied (Sternberg, Lamb, Esplin, Orbach, & Hershkowitz, 2002; Sternberg, Lamb, Orbach, et al., 2001). In addition to being better prompts organized, interviewers using the structured protocol used more open~ended and suggestive questions than were used in the comparison and fewer option~posing (baseline) interviews. In the baseline condition, only 10% of the interviewers' questions were invitations, whereas a third of the interviewers' questions were invitations when the protocol was followed. The total amount of information elicited from free recall memory also increased dramatically; whereas only 16% of the information was elicited interviews, about half of the information with the use of free recall in the pre~protocol was obtained with the use of free recall in the protocol interviews. The protocol also and suggestive prompts. In the baseline in~ reduced the use of directive, option~posing, terviews, 41% of the information was obtained with the use of option~posing and sug~ gestive questions compared with 24% in the protocol interviews. Furthermore, this pat~ tern of results was similar regardless of the children's age. Although younger children provided shorter and less detailed responses than did older children, analyses of inter~ views with 4~ to 6~year~old children revealed that the interviewers relied heavily on in~ vitations (34% of their questions) and succeeded in eliciting a substantial amount of information (49% of the total) using free~recall prompts. To further clarify the ability of preschoolers to address open~ ended questions, Lamb et al. (2003) studied forensic interviews of 130 4~ to S~year~olds. Like Sternberg, Lamb, Orbach, et al. (2001), they showed clearly that children as young as 4 years of age can indeed provide substantial amounts of information about alleged abuse in response to open~ended questions included in well~structured and planned forensic interviews. On average, one~half of the information provided by the children came in response to open~ended utterances. Furthermore, even though the older children reported more details in total and in response to the average invitation than the younger children did, the proportion of invitations eliciting new details did not change, suggesting that preschoolers are not uniquely incapable of responding to open~ ended prompts. Because both laboratory analog and field studies consistently show that the information pro~ vided in response to invitations is more likely to be accurate, these results suggested

440

LAMBET AL.

that open-ended invitations are superior investigative tools regardless of the interviewees' ages. Lamb et al. (2003) found that many interviewers made effective use of cued invitations (e.g., You said that he touched your vagina; tell me more about that.) rather than risky focused questions (So did he put his finger in your vagina?). Interestingly, action-based cues (Tell me more about him touching you.) were consistently more effective than were timesegmenting cues (e.g., Tell me more about what happened after he came into the room.) at all ages. Developmental improvements were especially dramatic with respect to timesegmenting cues, however, and these were extremely effective when addressed to 8-yearolds. With younger children, by contrast, references to actions already mentioned by the child proved more effective, presumably because such cues are less cognitively demanding than cues that required awareness of temporal sequences. At all ages, furthermore, more information would likely have been elicited if the interviewers had made greater use of cued invitations. Even when events are remembered, the process of retrieving those memories may be difficult, especially with increasing delays. Flin and her colleagues (1992) reported that 6-year-old children reported less information than 9-year-old children and adults and that, like adults, 6- and 9-year-olds reported less information 5 months after the event. Interestingly, the amount of incorrect information retrieved did not increase over time. Remembering is a constructive process, however. Like adults, children actively work to retrieve and organize memories. Thus when children are repeatedly interviewed, as is often the case when sexual abuse has been alleged, memories may not only become consolidated (facilitating subsequent recall), but also shaped and changed (Ornstein, Larus, & Clubb, 1992). Lamb, Sternberg, and Esplin (2000) found that both delay and age affected the amount of information recalled in nonprotocol forensic interviews, although it was of course impossible to assess the accuracy of the children's accounts. Because inaccurate information can be provided in response to open-ended prompts, especially when delays are long or there have been opportunities for post-event contamination (e.g., Leichtman & Ceci, 1995; Poole & Lindsay, 1995; Warren & Lane, 1995), it is crucially important for investigators to inquire about all relevant interviews and experiences since the alleged event, documenting the evidentiary chain of custody as they would when the evidence was a gun or knife (Rosenthal, 1995). Many researchers have shown that contextual cues can prompt reports of information (e.g., Malpass, 1996; Pipe & Wilson, 1994; Price &Goodman, 1990; Tulving, 1983; Wilkinson, 1988), although forensic investigators have been wary of possible trauma associated with these cues. Hershkowitz et al. (1998) reported that no children were anxious about revisiting the scenes at which they had allegedly been abused. In addition, although the children had been exhaustively interviewed at the investigators' offices, they were able to provide many new details (about 25% of the total) when reinterviewed at the scene of the alleged events. In a later study, however, the number of details retrieved was similar when children were interviewed only once, either at the scene or in an office, suggesting limits to the value of physical context reinstatement (Orbach, Hershkowitz, Lamb, Sternberg, & Horowitz, 2000). In addition, the effects of mental context reinstatement (MCR; i.e., guiding children to mentally reconstruct the setting in which the

15.

FACTORS AFFECTING YOUNG WITNESSES

441

event occurred) have been promising. Bekerian, Dennet, Hill, and Hitchcock (1990) re~ ported that MCR led to significant increases in the recall of both accurate and inaccu~ rate details, whereas McCauley and Fisher (1995, 1996) found that second graders pro~ vided up to twice as many accurate details when MCR techniques were employed than when they were not. In forensic settings, Hershkowitz, Orbach, Lamb, Sternberg, and Horowitz (2001) found that MCR did not affect the total amount of information pro~ vided, but enhanced the amount of information retrieved with the use of open~ended prompts, especially for the youngest (4-6~year~old) children studied. Similar amounts of information were retrieved with the use of physical and mental context reinstatement procedures (Hershkowitz, Lamb, Sternberg, Orbach, & Horowitz, 2002).

Suggestibility Whatever the vagaries and strengths of children's memories, the competency of child witnesses is often doubted on the grounds that children are too susceptible to influence by misleading questions or other sources of misinformation (Ceci & Bruck, 1993, 1995). Suggestibility is a multifaceted concept that involves social, communica~ tive, and memory processes. Children may respond inaccurately because they (a) infer that the interviewer would prefer a particular response (Ceci & Bruck, 1993, 1995); (b) do not understand the questions, but are eager to be cooperative (e.g., Hughes & Grieve, 1980); (c) retrieve the most recently acquired information about the event in question, although they might be able to retrieve information about the actual event if prompted to do so (Newcombe & Siegal, 1996, 1997); or (d) suffer from genuine source~ monitoring confusion that prevents them from discriminating between the original event and misinformation about it (Poole & Lindsay, 1997, 2001; Roberts, 2000). Given the number of processes that underlie suggestibility, it is not surprising that, at first glance, research on children's suggestibility appears to reveal a mixed and confusing picture (Ceci & Bruck, 1993; Ceci, Ross, & Toglia, 1987b). In a series of studies, Good~ man and her colleagues (Goodman & Aman, 1990; Goodman, Aman, & Hirschman, 1987; Goodman, Bottoms, Schwartz~ Kenney, & Rudy, 1991; Goodman, Rudy, Bottoms, & Aman, 1990; Goodman, Wilson, Hazan, & Reed, 1989) showed that children as young as 3 to 4 years old were seldom misled by questions such as "Did he keep his clothes on?" "Did he kiss you?" and "He took your clothes off, didn't he?" which suggested actions quite different from those that were witnessed or experienced. Other results might have been obtained if the actions had been more ambiguous and the suggestions more plausible, however (Steller, 1991), or if misleading questions had referred to details ob~ served or experienced in other contexts instead of being totally unfamiliar (Roberts & Blades, 1998). In other laboratory settings, in fact, preschoolers appear especially sus~ ceptible to suggestion (e.g., Ceci, Ross, & Toglia, 1987a, 1987b; King & Yuille, 1987; Toglia, Ceci, & Ross, 1989; see McAuliff, Kovera, & Viswesvaran, 1998, for a review). These apparently contradictory findings are not difficult to reconcile. In the studies doc~ umenting the resistance to suggestion, researchers have not repeated misleading ques~ tions over a short time period, exposed children to misleading stereotypes about target individuals, provided incentives to respond falsely, or included conditions that are often

442

LAMBET AL.

associated with recognition errors, such as a combination of specific questions and dolls, or instructions to think about nonevents when asked to pretend or guess. All of these experimental conditions increase the susceptibility to suggestion (e.g., Bruck, Ceci, Francouer, & Barr, 1995; Bruck, Ceci, Francouer, & Renick, 1995; Cassel, Roebers, & Bjorklund, 1996; Ceci, Huffman, Smith, & Loftus, 1994; Garven, Wood, Malpass, & Shaw, 1998; Leichtman & Ceci, 1995; Thompson, Clarke-Stewart, & Lepore, 1997). Age trends in susceptibility among school-aged children are less clear. Some researchers report that suggestibility is largely confined to the preschool years, with schoolaged children as capable as adults at resisting misleading questioning (e.g., Cassel et al., 1996; Eisen, Qin, Goodman, & Davis, 2002). Others find that suggestibility continues to decline over the elementary school years (Ackil & Zaragoza, 1995; Coxon & Valentine, 1997; Lindberg, Keiffer, & Thomas, 2000; Robinson & Briggs, 1997). Yet others suggest that older children and adults are sometimes more suggestible than younger children (Brainerd & Reyna, 1991; Lindberg, 1991). Suggestions are less likely to affect children's accounts when they pertain to central or salient details (e.g., Cassel et al., 1996; Dent & Stephenson, 1979; Dodd & Bradshaw, 1980; King & Yuille, 1987) and when interviewers counsel children to report personally experienced events only (Poole & Lindsay, 1997). Finally, although much less research has been conducted on suggestibility regarding memories of stressful or traumatic incidents, Eisen et al. (2002) reported that stress did not generally increase (or decrease) susceptibility to misleading questioning. In theory at least, susceptibility to misleading suggestions should vary, depending on the child's motivation to be completely accurate and/or comply with the interviewer's implicit or explicit agenda (King & Yuille, 1987). Children may feel obliged to answer adults' questions no matter how bizarre (Hughes & Grieve, 1980) and may assume that the repetition of a question implies that the initial answer was unsatisfactory (Ceci & Bruck, 1993). Subtle differences in the interviewers' style may also affect children's suggestibility. Goodman et al. (1991) reported that 3- to 7-year-olds were equally resistant to suggestions by "nice" and more neutral interviewers, whereas Davies and Bottoms (1998) and Carter et al. (1996) found that 6- and 7-year-old children interviewed by supportive interviewers made fewer errors in response to misleading questions than did children interviewed by neutral or nonsupportive interviews. Goodman et al. (1989) reported that 7- and 10-year-old children were surprisingly likely to accept suggestions made "in an atmosphere of accusation" 4 years after the event being recalled (Goodman & Clarke-Stewart, 1991), and Ceci et al. (1987a, 1987b) reported that preschoolers were more likely to accept suggestions made by an adult than by a 7-year-old confederate. Overall, therefore, the effects of interviewer characteristics are less consistent and impressive than one might expect. Regardless of the resolution of the various controversies concerning children's suggestibility, most researchers agree that the manner in which children are questioned can have profound implications not only for what is reported but also for what is remembered, and this increases the importance of careful interviewing (Brainerd & Ornstein, 1991; Lamb et al., 1998; Poole & Lamb, 1998). An interviewer who engages in suggestive questioning can manipulate both young and old witnesses. Such questions

15.

FACTORS AFFECTING YOUNG WITNESSES

443

are most likely to be influential when the memory is not rich or recent, when the content was imagined rather than experienced, when the questions themselves are so complicated that the witness is confused, and when the interviewer appears to have such authority or status that the witness feels compelled to accept his or her implied construction of the events. Clearly, children do remember important details of incidents that they have observed or experienced, and although their accounts can be manipulated, interviewers who are aware of children's capacities and deficiencies can avoid many of the problems posed by questions that force children to operate at or beyond the limits of their capacities.

CONCLUSION AND IMPLICATIONS FOR PRACTICE

Years of systematic research in the laboratory and in authentic forensic settings have helped create a broad and growing understanding of the factors that influence the quality of the information that can be obtained from young children. As we have shown above, children's memorial and communicative capacities are clearly less than those of adults, and they (especially preschoolers) are more susceptible than adults to witting or unwitting suggestions by adults. It may often be impossible to differentiate between accurate and inaccurate components of the accounts provided by most 3- and many 4-year-olds, furthermore, and thus many are likely to be legally incompetent. On the other hand, children can be informative witnesses when they can remember and recount their experiences accurately; whether their accounts are comprehensible, detailed, and accurate depends on a variety of factors. Perhaps the most important factor is the quality of the investigative interview conducted to obtain the information. In practice, interviewers who elicit information with the use of open-ended prompts and avoid placing undue cognitive or social burdens on children are most likely to obtain relevant information. This approach is particularly successful when the incident occurred more recently and was salient, the child is 5 or older, and the child has not been subjected to suggestive interviewing by concerned relatives and inept investigators. Unfortunately, research suggests that many forensic interviews are poorly conducted and that widespread consensus about the desirable qualities of good interviews does not guarantee compliance, even by investigators who are confident that their interviews closely follow best-practice guidelines. Because we have found that intensive training that combines use of the structured interview protocol with regular reviews of recent interviews is necessary to enhance the quality of investigative practice, we have recommended that regular supervision and quality assessment become a regular part of agency practice. Like other researchers (e.g., Stevenson, Leung, & Cheung, 1992; Warren et al., 1999), we have found that brief intensive training sessions 1 to 5 days in length enhance the trainees' knowledge of what they should do, but have little effect on what they actually do when interviewing alleged victims. By contrast, interviewing practices improve when the intensive training is accompanied by practice with the structured protocol and regular opportunities to review actual interviews continue indefinitely (Lamb, Sternberg, Orbach, Esplin, & Mitchell, 2002; Lamb, Sternberg, Orbach, Hershkowitz, et al., 2002).

444

LAMBET AL.

REFERENCES Ackil, J. K., & Zaragoza, M. S. (1995). Developmental differences in eyewitness suggestibility and memory for source. Journal of Experimental Child Psychology, 60, 57-83. Aldridge, J., Lamb, M. E., Sternberg, K. J., Orbach, Y., Esplin, P. W., & Bowler, L. (2004). Using a human figure drawing to elicit information from alleged victims of child abuse. Journal of Consulting and Clinical Psychology, 72, 304-316. American Professional Society on the Abuse of Children. ( 1990). Guidelines for psychosocial evaluation of suspected sexual abuse in young children. Chicago, IL: Author. American Professional Society on the Abuse of Children. (1997). Guidelines for psychosocial evaluation of suspected sexual abuse in young children (Revised). Chicago, IL: Author. Bekerian, D. A., Dennet, J. L., Hill, K., & Hitchcock, R. (1990). Effects of detailed imagery on simulated witness recall. Paper presented at the Second European Conference on Law and Psychology, Nuremberg, Germany. Brainerd, C. J., & Ornstein, P. A (1991). Children's memory for witnessed events: The developmental backdrop. In J. Doris (Ed.), The suggestibility of children's recollections (pp. 10-20). Washington, DC: American Psychological Association. Brainerd, C. J., & Reyna, V. F. (1991). Fuzzy-trace theory and children's false memories. Journal of Experimental Child Psychology, 27, 351-369. Brainerd, C. J., & Reyna, V. F. (1996). Mere testing creates false memories in children. Developmental Psychology, 32, 467-476. Brennan, M., & Brennan, R. E. (1988). Strange language: Child victims under cross examination (3rd ed.). Wagga Wagga, NSW, Australia: Riverina Literacy Centre. Brown, R. (1973). A first language. Cambridge, MA: Harvard University Press. Bruck, M., Ceci, S. J., Francouer, E., & Barr, R. (1995). "I hardly cried when I got my shot!" Influencing children's reports about a visit to their pediatrician. Child Development, 66, 193-208. Bruck, M., Ceci, S. J., Francouer, E., & Renick, A (1995). Anatomically detailed dolls do not facilitate preschoolers' reports of a pediatric examination involving genital touching. Journal of Experimental Psychology: Applied, 1, 95-109. Bull, R. (1992). Obtaining evidence expertly: The reliability of interviews with child witnesses. Expert Evidence, 1, 5-12. Bull, R. (1995). Innovative techniques for the questioning of child witnesses, especially those who are young and those with a learning disability. In M. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 79-194). Thousand Oaks, CA: Sage. Bull, R. (1996). Good practice for video recorded interviews with child witnesses for use in criminal proceedings. In G. Davies, S. Lloyd-Bostock, M. McMarran, & C. Wilson (Eds.), Psychology, law, and criminal justice: International developments in research and practice (pp. 100117). Berlin/New York: Walter de Gruyter. Camparo, L. B., Wagner, J. T., & Saywitz, K. J. (2001). Interviewing children about real and fictitious events: Revisiting the narrative elaboration procedure. Law and Human Behavior 25, 63-80. Carter, C. A., Bottoms, B. L., & Levine, M. (1996). Linguistic and socioemotional influences in the accuracy of children's reports. Law and Human Behavior, 20, 335-358. Cassel, W. S., Roebers, C. E. M., & Bjorklund, D. F. (1996). Developmental patterns of eyewitness responses to repeated and increasingly suggested questions. Journal of Experimental Child Psychology, 61, 116-133. Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113, 403-439. Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific analysis of children's testimony. Washington, DC: American Psychological Association.

15.

FACTORS AFFECTING YOUNG WITNESSES

445

Ceci, S. J., Huffman, M. L. C., Smith, E., & Loftus, E. F. (1994). Repeatedly thinking about a non-event: Source misattributions among preschoolers. Consciousness and Cognition, 3, 388-407. Ceci, S. J., Leichtman, M.D., & Putnick, M. E. (Eds.). (1992). Cognitive and social factors in early deception. Hillsdale, NJ: Lawrence Erlbaum Associates. Ceci, S. J., Ross, D. F., & Toglia, M.P. (1987a). Suggestibility of children's memory: Psychological issues. Journal of Experimental Psychology: General, 116, 38-49. Ceci, S. J., Ross, D. F., & Toglia, M.P. (1987b). Age differences in suggestibility: Narrowing the uncertainties. InS. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.), Children's eyewitness memory (pp. 79-91). New York: Springer-Verlag. Cederborg, A., Orbach, Y., Sternberg, K. L., & Lamb, M. E. (2000). Investigative interviews of child witnesses in Sweden. Child Abuse & Neglect, 24, 1355-1361. Collins, W. A. (1970). Learning of media content: A developmental study. Child Development, 41, 1133-1142. Collins, W. A., & Wellman, H. (1982). Social scripts and developmental patterns in comprehension of televised narratives. Communication Research, 9, 380-398. Collins, W. A., Wellman, H., Keniston, A., & Westby, S. (1978). Age-related aspects of comprehensive and inference from a televised dramatic narrative. Child Development, 49, 389-399. Coxon, P., & Valentine, P. (1997). The effects of the age of eyewitness on the accuracy and suggestibility of their testimony. Applied Cognitive Psychology, 11, 415-430. Craig, R. A., Sheibe, R., Kircher, J., Raskin, D. C., Kircher, J. C., & Dodd, D. H. (1999). Interviewer questions and content analysis of children's statements of sexual abuse. Applied Developmental Science, 3, 77-85. Dale, P. S. (1976). Language development: Structure and function. New York: Holt, Rinehart & Winston. Davidson, D. (1991). Children's recognition and recall memory for typical and atypical actions in scriptbased stories. Paper presented to the meeting of the Society for Research in Child Development, Seattle, W A. Davies, S. L., & Bottoms, B. L. (1998, March). Effects of social support on children's eyewitness reports. Paper presented to the American Psychology-Law Society conference, Redondo Beach, CA. Davies, G., Tarrant, A., & Flin, R. (1989). Close encounters of the witness kind: Children's memory for a simulated health inspection. British Journal of Psychology, 80, 415-429. Davies, G., Westcott, H., & Horan, N. (2000). The impact of questioning style on the content of investigative interviews with suspected child sexual abuse victims. Psychology, Crime & Law, 6, 681-97. Dent, H. R. (1982). The effects of interviewing strategies on the results of interviews with child witnesses. In A. T rankell (Ed.), Reconstructing the past: The role of psychologists in criminal trials (pp. 279-297). Stockholm: Norstedt. Dent, H. R., & Flin, R. (Eds.). (1992). Children as witnesses. Chichester, England: John Wiley &Sons. Dent, H. R., & Stephenson, G. M. (1979). An experimental study of the effectiveness of different techniques of questioning child witnesses. British Journal of Social and Clinical Psychology, 18,41-51. de Villiers, J., & de Villiers, P. (1999). Language development. In M.H. Bornstein & M.E. Lamb (Eds.), Developmental psychology: An advanced textbook (4th ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Dodd, D. H., & Bradshaw, J. M. (1980). Leading questions and memory: Pragmatic constraints. Journal of Verbal Learning and Verbal Behavior, 19,695-704. Doris, J. (Ed.). (1991). The suggestibility of children's recollections: Implications for eyewitness testimony. Washington, DC: American Psychological Association.

446

LAMBET AL.

Eisen, M. Qin, J., Goodman, G. S., & Davis, S. L. (2002). Memory and suggestibility in maltreated children: Age, stress arousal, dissociation, and psychopathology. Journal of Experimental Child Psychology, 83, 167-212. Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques for investigating interviewing: The cognitive interview. Springfield, IL: Charles C. Thomas. Flin, R., Boon, J., Knox, A., & Bull, R. (1992). The effect of a five month delay on children's and adults' eyewitness memory. British Journal of Psychology, 83, 323-336. Garven, S., Wood, J. M., Malpass, R. S., & Shaw, J. S. (1998). More than suggestion: The effect of interviewing techniques from the McMartin Preschool case. Journal of Applied Psychology, 83,347-359 Goodman, G. S., & Aman, C. (1990). Children's use of anatomically detailed dolls to recount an event. Child Development, 61, 1859-1871. Goodman, G. S., Aman, C., & Hirschman, J. (1987). Child sexual and physical abuse: Children's testimony. In S. J. Ceci, M. P. Toglia, & D. P. Ross (Eds.), Children's eyewitness memory (pp. 1-23). New York: Springer-Verlag. Goodman, G. S., Bottoms, B. L., Schwartz-Kenney, B. M., & Rudy, L. (1991). Children's testimony about a stressful event: Improving children's reports. Journal of Narrative and Life History, 1, 69-99. Goodman, G. S., & Clarke-Stewart, K. A. (1991). Suggestibility in children's testimony: Implications for sexual abuse investigations. In J. Doris (Ed.), The suggestibility of children's recollections (pp. 92-105). Washington, DC: American Psychological Association. Goodman, G. S., & Reed, D. S. (1986). Age differences in eyewitness testimony. Law and Human Behavior, 10,317-332. Goodman, G. S., Rudy, L., Bottoms, B., & Aman, C. (1990). Children's concerns and memory: Issues of ecological validity on the study of children's eyewitness testimony. In R. Fivush & J. Hudson (Eds.), Knowing and remembering in young children (pp. 249-284). New York: Cambridge University Press. Goodman, G. S., Wilson, M. E., Hazan, C., & Reed, R. S. (April, 1989). Children's testimony nearly four years after an event. Paper presented to the Eastern Psychological Association, Boston, MA. Harner, L. (1975). Yesterday and tomorrow: Development of early understanding of the terms. Developmental Psychology, 11, 864-865. Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., & Horowitz, D. (2001). The effects of mental context reinstatement on children's accounts of sexual abuse. Applied Cognitive Psychology, 15, 235-248. Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., & Horowitz, D. (2002). A comparison of mental and physical context reinstatement in forensic interviews with alleges victims of sexual abuse. Applied Cognitive Psychology, 16, 429-441. Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., Horowitz, D., & Hovav, M. (1998). Visiting the scene of the crime: Effects on children's recall of alleged abuse. Legal & Criminological Psychology, 3, 195-207. Home Office & Department of Health. (1992). Memorandum of good practice on video recorded interviews with child witnesses for criminal proceedings. London: Her Majesty's Stationery Office. Hughes, M., & Grieve, R. (1980). On asking children bizarre questions. In M. Donaldson, R. Grieve, &C. Pratt (Eds.), Early childhood development and education (pp. 104-114). Oxford, England: Blackwell. Hunt, J. S., & Borgida, E. (2001). Is that what I said? Witnesses' responses to interviewer modifications. Law and Human Behavior, 25, 583-603. Hutcheson, G. D., Baxter, J. S., Telfer, K., & Warden, D. (1995). Child witness statement quality: Question type and errors of omission. Law and Human Behavior, 19,631-648.

15.

FACTORS AFFECTING YOUNG WITNESSES

447

Imhoff, M. C. & Baker-Ward, L. (1999). Preschoolers' suggestibility: Effects of developmentally appropriate language and interviewer supportiveness. Applied Developmental Psychology, 20, 407-429. Johnson, M. K., & Foley, M.A. (1984). Differentiating fact from fantasy: The reliability of children's memory. Journal of Social Issues, 40, 33-50. Jones, D.P. H. (1992). Interviewing the sexually abused child. Oxford: Gaskell. King, M. A., & Yuille,]. C. (1987). Suggestibility and the child witness. In S. ]. Ceci, D. F. Ross, & M.P. Toglia (Eds.), Children's eyewitness memory (pp. 24-35). New York: SpringerVerlag. Lamb, M. E., & Fauchier, A. (2001). The effects of question type on self-contradictions by children in the course of forensic interviews. Applied Cognitive Psychology, 15, 483-491. Lamb, M. E., Hershkowitz, I., Sternberg, K. ]., Boat, B., & Everson, M.D. (1996). Investigative interviews of alleged sexual abuse victims with and without anatomical dolls. Child Abuse and Neglect, 20, 1239-1247. Lamb, M. E., Hershkowitz, I., Sternberg, K. ]., Esplin, P. W., Hovav, M., Manor, T., et al. (1996). Effects of investigative utterance types on Israeli children's responses. International Journal of Behavioral Development, 19, 627-637. Lamb, M. E., Sternberg, K. ]., & Esplin, P. W. (1994). Factors influencing the reliability and validity of the statements made by young victims of sexual maltreatment. Journal of Applied Developmental Psychology, 15, 255-280. Lamb, M. E., Sternberg, K. ]., & Esplin, P. W. (1995). Making children into competent witnesses: Reactions to the amicus brief In re Michaels. Psychology, Public Policy, and the Law, 1, 438-449. Lamb, M. E., Sternberg, K. ]., & Esplin, P. W. (1998). Conducting investigative interviews of alleged sexual abuse victims. Child Abuse and Neglect, 22, 813-823. Lamb, M. E., Sternberg, K. ]., & Esplin, P. W. (2000). Effect of age and length of delay on the amount of information provided by alleged abuse victims in investigative interviews. Child Development, 71, 1586-1596. Lamb, M. E., Sternberg, K. ]., Orbach, Y., Esplin, P. W., & Mitchell, S. (2002). Is ongoing feedback necessary to maintain the quality of investigative interviews with allegedly abused children? Applied Developmental Science, 6, 35-41. Lamb, M. E., Sternberg, K. ]., Orbach, Y., Esplin, P. W., Stewart, H., &Mitchell, S. (2003). Age differences in young children's responses to open-ended invitations in the course of forensic interviews. Journal of Consulting and Clinical Psychology, 71, 926-934. Lamb, M. E., Sternberg, K. ]., Orbach, Y., Hershkowitz, I., Horowitz, D., & Esplin, P. W. (2002). The effects of intensive training and ongoing supervision on the quality of investigative interviews with alleged sex abuse victims. Applied Developmental Science, 6, 114-125. Lamb, M. E., & Thierry, K. L. (2005). Understanding children's testimony regarding their alleged abuse: Contributions of field and laboratory analog research. In D. M. Teti (Ed.), Handbook of research methods in developmental psychology. Malden, MA: Blackwell Publishers. Leichtman, M. D., & Ceci, S. ]. (1995). The effects of stereotypes and suggestions on preschoolers' reports. Developmental Psychology, 31, 568-578. Lindberg, M. (1991). An interactive approach to assessing the suggestibility and testimony of eyewitnesses. In]. Doris, (Ed.), The suggestibility of children's recollections (pp. 47-55). Washington, DC: American Psychological Association. Lindberg, M.A., Keiffer,]., & Thomas, S. (2000). Eyewitness testimony for physical abuse as a function of personal experience, development, and focus of study. Journal of Applied Developmental Psychology, 21, 555-591. Lindsay, D. S., & Johnson, M. K. (1987). Reality monitoring and suggestibility. In S. ]. Ceci, M.P. Toglia, & D. F. Ross (Eds.), Children's eyewitness testimony (pp. 79-91). New York: Springer-Verlag.

448

LAMBET AL.

Malpass, R. S. (1996). Enhancing eyewitness memory. In S. L. Sporer, R. S. Malpass, & G. Koehnken (Eds.), Psychological issues in eyewitness identification (pp. 177-204). Mahwah, NJ: Lawrence Erlbaum Associates. Marin, B. V., Holmes, D. L., Guth, M., & Kovac, P. (1979). The potential of children as eyewitnesses. Law and Human Behavior, 3, 295-305. Martin, C. L., & Halverson, C. F. (1983). The effects of sex-typing schemas on young children's memory. Child Development, 54, 563-574. McAuliff, B. D., Kovera, M. B., & Viswesvaran, C. (1998, March). Methodological issues in child suggestibility research: A meta-analysis. Paper presented to the American Psychology-Law Society Convention, Redondo Beach, CA. McCartney, K., & Nelson, K. (1981). Children's use of scripts in story recall. Discourse Processes, 4, 59-70. McCauley, M. R., & Fisher, R. P. (1995). Facilitating children's eyewitness recall with the revised Cognitive Interview. Journal of Applied Psychology, 80,510-516. McCauley, M. R., & Fisher, R. P. (1996). Enhancing children's eyewitness testimony with the Cognitive Interview. In G. Davies, S. Lloyd-Bostock, M. McMurran, & C. Wilson (Eds.), Psychology, law, and criminal justice: International developments in research and practice (pp. 127134). Berlin: Walter de Gruyter. McGough, L. (1994). Child witnesses. New Haven, CT: Yale University Press. Memon, A., & Bull, R. (Eds.). (1999). Handbook of the psychology of interviewing. New York: John Wiley & Sons. Myles-Worsley, M., Cromer, C., & Dodd, D. (1986). Children's preschool script reconstruction: Reliance on general knowledge as memory fades. Developmental Psychology, 22, 2-30. Nelson, K. (1986). Event knowledge: Afunctional approach to cognitive development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K., & Gruendel, J. (1981). Generalized event representations: Basic building blocks of cognitive development. In M. E. Lamb & A. L. Brown (Eds.), Advances in development psychology (Vol. 1, pp. 131-158). Hillsdale, NJ: Lawrence Erlbaum Associates. Newcombe, P. A., & Siegal, M. (1996). Where to look first for suggestibility in young children. Cognition, 59, 337-356. Newcombe, P. A., & Siegal, M. (1997). Explicitly questioning the nature of suggestibility in preschoolers' memory and retention. ]oumal of Experimental Child Psychology, 67, 185-203. Oates, K., & Shrimpton, S. (1991). Children's memories for stressful and non-stressful events. Medical Science and Law, 31, 4-10. Orbach, Y., Hershkowitz, 1., Lamb, M. E., Sternberg, K. J., Esplin, P. W., & Horowitz, D. (2000). Assessing the value of structured protocols for forensic interviews of alleged abuse victims. Child Abuse and Neglect, 24, 733-752. Orbach, Y., Hershkowitz, 1., Lamb, M. E., Sternberg, K. J., & Horowitz, D. (2000). Interviewing at the scene of the crime: effects on children's recall of alleged abuse. Legal and Criminological Psychology, 5, 135-147. Orbach, Y., & Lamb, M. E. (2001). The relationship between within-interview contradictions and eliciting interviewer utterances. Child Abuse & Neglect, 25, 323-333. Ornstein, P. A., Larus, D. M., & Clubb, P. A. (1992). Understanding children's testimony: Implications of the research on children's memory. In R. Vasta (Ed.), Annals of child development (Vol. 8, pp. 147-176). London: Jessica Kingsley. Paris, S. G. (1988). Motivated remembering. In F. E. Weinert & M. Perlmutter (Eds.), Memory development: Universal changes and individual differences (pp. 221-242). Hillsdale, NJ: Lawrence Erlbaum Associates. Pea, R. D. (1980). The development of negation in early child language. In D.R. Olson (Ed.), The social foundations of language and thought (pp. 156-186). New York: Norton. Perry, N. W., McAuliff, B. D., Tam, P., Claycomb, L., Dostal, C., & Flanagan, C. (1995). When lawyers question children: Is justice served? Law and Human Behavior, 19, 609-629.

15.

FACTORS AFFECTING YOUNG WITNESSES

449

Perry, N. W., & Wrightsman, L. S. (1991). The child witness: Legal issues and dilemmas. Newbery Park, CA: Sage. Pipe, M. E., & Wilson, C. (1994). Cues and secrets: Influences on children's event reports. Developmental Psychology, 30, 515-525. Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. Poole, D. A., & Lindsay, D. S. (1995). Interviewing preschoolers: Effects of nonsuggestive techniques, parental coaching, and leading questions on reports of nonexperienced events. Journal of Experimental Child Psychology, 60, 129-154. Poole, D. A., & Lindsay, D. S. (1997, April). Misinformation from parents and children's source monitoring: Implications for testimony. Paper presented to the Society for Research in Child Development, Washington, DC. Poole, D. A., & Lindsay, D. S. (2001). Children's eyewitness reports after exposure to misinformation by parents. Journal of Experimental Psychology: Applied, 7, 27-50. Poole, D. A., & White, L. T. (1993). Two years later: Effects of question repetition and retention intervals on the eyewitness testimony of children and adults. Developmental Psychology, 29, 844-853. Price, D. W. W., & Goodman, G. S. (1990). Visiting the wizard: Children's memory of a recurring event. Child Development, 61, 664-680. Raskin, D. C., & Esplin, P. W. (1991). Statement validity assessments: Interview procedures and content analyses of children's statements of sexual abuse. Behavioral Assessment, 13, 265-291. Raskin, D., & Yuille, J. (1989). Problems of evaluating interviews of children in sexual abuse cases. InS. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.), Perspectives on children's testimony (pp. 184-207). New York: Springer-Verlag. Reich, P. A. (1986). Language development. Englewood Cliffs, NJ: Prentice-Hall. Roberts, K. P. (2000). Children's source monitoring. InK. P. Roberts, & M. Blades (Eds.), An overview of theory and research on children's source monitoring (pp. 11-57). Bethesda, MD: National Institute of Child Health & Human Development. Roberts, K. P., & Blades, M. (1995). Children's discriminations of memories for actual and pretend actions in a hiding task. BritishJoumal of Developmental Psychology, 13, 321-333. Roberts, K. P., & Blades, M. (1998). The effects of interacting in repeated events on children's eyewitness memory and source monitoring. Applied Cognitive Psychology, 12, 489-503. Roberts, K. P., & Lamb, M.E. (1999). Children's responses when interviewers distort details during investigative interviews. Legal & Criminological Psychology, 4, 23-31. Roberts, K. P., Lamb, M. E., & Sternberg, K. J. (2004). The effects of rapport-building style on children's reports of a staged event. Applied Cognitive Psychology, 18, 189-202. Robinson, J., & Briggs, P. (1997). Age trends and eye-witness suggestibility and compliance. Psychology Crime & Law, 3, 187-202. Rosenthal, R. (1995). State of New Jersey vs. Margaret Kelly Michaels: An overview. Psychology, Public Policy, and Law, 1, 246-271. Sattler, J. (1998). Clinical and forensic interviewing of children and families. San Diego: Author. Saywitz, K. J. (1988). The credibility of the child witness. Family Advocate, 10, 38. Saywitz, K., Nathanson, R., Snyder, L., & Lamphear, V. (1993). Preparing children for the investigative and judicial process: Improving communication, memory and emotional resiliency. Los Angeles: National Center on Child Abuse and Neglect. Saywitz, K. J., & Snyder, L. (1993). Improving children's testimony with preparation. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving testimony (pp. 117-146). New York: Guilford. Saywitz, K. J., & Snyder, L. (1996). Narrative elaboration: Test of a new procedure for interviewing children. Journal of Consulting & Clinical Psychology, 64, 134 7-13 57. Shank, R., & Abelson, R. (1977). Scripts, plans, goals, and understanding. Hillsdale, NJ: Lawrence Erlbaum Associates.

450

LAMBET AL.

Spencer,]. R., & Flin, R. (1990). The evidence of children: The law and the psychology (2nd ed.). London: Blackstone Press. Steller, M. (1991). Commentary: Rehabilitation of the child witness. In J. Doris (Ed.), The suggestibility of children's recollections (pp. 106-109). Washington, DC: American Psychological Association. Sternberg, K. ]., Lamb, M. E., Davies, G. M., & Westcott, H. L. (2001). The memorandum of good practice: Theory versus application. Child Abuse & Neglect, 25, 669-681. Sternberg, K. ]., Lamb, M. E., Esplin, P. W., & Baradaran, L. P. (1999). Using a scripted protocol in investigative interviews: A pilot study. Applied Developmental Science, 3, 70-76. Sternberg, K. ]., Lamb, M. E., Esplin, P. W., Orbach, Y., &Hershkowitz, I. (2002). Using a structured protocol to improve the quality of investigative interviews. In M. Eisen, G. Goodman, & ]. Quas (Eds.), Memory and suggestibility in the forensic interview (pp. 409-436). Mahwah, NJ: Lawrence Erlbaum Associates. Sternberg, K. ]., Lamb, M. E., & Hershkowitz, I. (1996). Child sexual abuse investigations in Israel. Criminal]ustice and Behavior, 23, 322-337. Sternberg, K. J., Lamb, M. E., Hershkowitz, 1., Esplin, P. W., Redlich, A., & Sunshine, N. (1996). The relationship between investigative utterance types and the informativeness of child witnesses. Journal of Applied Developmental Psychology, 17, 439-451. Sternberg, K. J., Lamb, M. E., Hershkowitz, 1., Yudilevitch, L., Orbach, Y., Esplin, P. W., et al. (1997). Effects of introductory style on children's abilities to describe experiences of sexual abuse. Child Abuse and Neglect, 21, 1133-1146. Sternberg, K. ]., Lamb, M. E., Orbach, Y., Esplin, P. W., & Mitchell, S. (2001). Use of a structured investigative protocol enhances young children's responses to free-recall prompts in the course of forensic interviews. Journal of Applied Psychology, 86, 997-1005. Stevenson, K. M., Leung, P., & Cheung, K. M. (1992). Competency-based evaluation of interviewing skills in child sexual abuse cases. Social Work Research & Abstracts, 28, 11-16. Steward, M. S. (1993). Understanding children's memories of medical procedures: "He didn't touch me and it didn't hurt!" In C.A. Nelson (Ed.), Memory and affect in development (pp. 171225). Hillsdale, NJ: Lawrence Erlbaum Associates. Thompson, W. C., Clarke-Stewart, K. A., & Lepore, S. J. (1997). What did the janitor do? Suggestive interviewing and the accuracy of children's accounts. Law and Human Behavior, 21, 405-426. Thierry, K. L., Lamb, M. E., Orbach, Y., & Pipe, M. E. (2005). The utility of anatomical dolls during interviews with alleged sexual abuse victims. Poster session presented at the annual meeting of the American Psychology-Law Society, Scottsdale, AZ. Toglia, M. P., Ceci, S. J., & Ross, D. F. (April, 1989). Prestige vs. source monitoring in children's suggestibility. Paper presented to the Society for Research in Child Development, Kansas City, MO. Tulving, E. (1983). Elements of episodic memory. Oxford: Clarendon Press. Walker, A. G. (1993). Questioning young children in court: A linguistic case study. Law and Human Behavior, 17, 59-81. Walker, A. G. (1994). Handbook on questioning children: A linguistic perspective. Washington, DC: American Bar Association Center on Children and the Law. Walker, N., & Hunt, J. S. (1998). Interviewing child victim-witnesses: How you ask is what you get. In C. R. Thompson, D. Herrman, J.D. Read, D. Bruce, D. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 55-87). Mahwah, NJ: Lawrence Erlbaum Associates. Walker, A. G., & Warren, A. R. (1995). The language of the child abuse interview: Asking the questions, understanding the answers. InT. Ney (Ed.), True and false allegations of child sexual abuse: Assessment and case management (pp. 153-162). New York: Brunner/Mazel.

15.

FACTORS AFFECTING YOUNG WITNESSES

451

Warren, A. R., & Lane, P. (1995). Effects of timing and type of questioning on eyewitness accuracy and suggestibility. In M.S. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness Applied psychology: Individual, social, and community issues (pp. 44-60). Thousand Oaks, CA: Sage Publications. Warren, A. R., & McCloskey, L.A. (1997). Language in social contexts. InS. B. Gleason (Ed.), The development of language (4th ed., pp. 210-258). New York: Allyn & Bacon. Warren, A. R., Woodall, C. C., Hunt, J. S., & Perry, N. W. (1996). "It sounds good in theory, but... ": Do investigative interviewers follow guidelines based on memory research? Child Maltreatment, I, 231-245. Warren, A. R., Woodall, C. E., Thomas, M., Nunno, M., Keeney, J. M., Larson, S.M., et al. (1999). Assessing the effectiveness of a training program for interviewing child witnesses. Applied Developmental Science, 3, 128-135. Wilkinson, J. (1988). Context effects in children's event memory. In M. M. Gruneberg, P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of memory: Current research issues (Vol. 1, pp. 107-111). New York: Wiley. Wood, J. M., McClure, K. A., & Birch, R. A. (1996). Suggestions for improving interviews in child protection agencies. Child Maltreatment, I, 223-230.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

16 The Development of Event Memory: Implications for Child Witness Testimony Margaret-Ellen Pipe National Institute of Child Health & Human Development Karen L. Thierry Rutgers, The State University of New Jersey

Michael E. Lamb University of Cambridge

What can we reasonably expect child witnesses to remember about their experiences, personal or witnessed, over time, and how does this change with age? Although it is not the only determinant, the ability to remember events is a prerequisite if children are to provide useful testimony. In this chapter, we focus on developmental changes in children's ability to provide accurate verbal accounts of their experiences and on forensically relevant factors that affect event memory. We begin by outlining developmental changes in children's event memory, the effects of the nature of the event and children's understanding of it, and the special problems posed when children must recall repeated experiences in forensic contexts. We then discuss how memories change over time, the forensic implications of such changes, and the role of the supportiveness of the interview situation, the type of questions, and other interview techniques used to elicit details from children. In the final section, we briefly consider individual differences in cognitive and emotional development that influence children's recall.

453

454

PIPE, THIERRY, LAMB

STRENGTHS AND LIMITATIONS OF CHILDREN'S EVENT MEMORY

Earliest Memories From the time they are 2 or 3 years of age, it is clear that young children can remember and verbally recount a great deal of information about many of their experiences after short delays (for example, of 1 month or less; Baker-Ward, Gordon, Ornstein, Larus, & Clubb, 1993; Cassel & Bjorklund, 1995; Fivush & Hamond, 1990; Pillemer, 1993) and sometimes also after much longer delays (Fivush & Schwarzmueller, 1998; Hamond & Fivush, 1991). Indeed, even before the acquisition oflanguage, very young infants show evidence of remembering, sometimes over long time periods, if appropriate nonverbal measures of memory are used (Hildreth, Sweeny, & Rovee-Collier, 2003; Rovee-Collier, Hartshorn, & OiRubbo, 1999; see Rovee-Collier & Hayne, 2000, for a review). Interestingly, however, although early experiences undoubtedly influence cognitive, social, and/or emotional development, they generally do not become part of children's "autobiographical memory" systems (see Howe & Courage, 2004, for a discussion). As adults, our very earliest memories remain elusive to conscious recall, a phenomenon known as infantile amnesia (Cowan & Davidson, 1984; Pillemer & White, 1989). This difficulty in recalling events from the first years of life is not restricted to adults; young children, having acquired the language necessary for verbal recall, also have difficulty recalling their earliest memories. Furthermore, although some events are more likely than others to be recalled from early ages (Neisser, 2004), even highly traumatic events do not appear to be available for later explicit recall when the events occurred very early in life. Terr (1988), for example, found that children who experienced traumatic events (e.g., dog bites) when they were younger than 18 months of age were unable to verbally recall the events when they were tested at older ages. The most detailed and coherent accounts of events were provided by children who were at least 2.5 to 3 years of age at the time of the experience. Researchers examining children's memories of stressful experiences, such as medical procedures and injuries, have reported similar findings (Howe, Courage, & Peterson, 1994; Peterson & Rideout, 1997). Explanations for why these early experiences should be so difficult to recall, despite functional memory systems capable of encoding and retaining information over long time periods, are numerous and controversial (see, for example, Howe & Courage, 1993; Howe, Courage, & Edison, 2003; Neisser, 2004; Nelson & Fivush, 2004). Language clearly plays a role; memories acquired during infancy are very fragile, in part because these memories are only encoded in nonverbal modalities, involving perceptually based attributes (Hayne & Rovee-Collier, 1995). In order to recall nonverbal representations of events verbally, these memories must be recoded into language form. Although some anecdotal evidence suggests that this recoding is possible (e.g., Myers, Perris, & Speaker, 1994), the recoding of nonverbal memories into verbal form appears to depend to some extent on children's linguistic abilities when they initially experienced the event (Bauer & Wewerka, 1997; Simcock & Hayne, 2002). Simcock and Hayne (2002) provide compelling evidence that verbal, although not preverbal, memories are accessible to very

16.

DEVELOPMENT OF EVENT MEMORY

455

young children (see also Bauer & Wewerka, 1997). In their study, 2 7-month-olds observed a novel event and were asked to verbally recall the event 6 months or 1 year later. Although the children had the vocabulary to verbally recall the event at the time of testing, they failed to do so and could only recall the event with the use of photographs and behavioral reenactments (Simcock & Hayne, 2002). Bauer and colleagues (2004) have shown, however, that events from an even younger age (20 months) were recalled verbally when prop items (but not photographs) were re-presented at the time of recall, 3-6 months later. Social interactionist perspectives on the emergence of autobiographical memory also highlight the importance of language development. For instance, Nelson and her colleagues argued that children start to form long-term memories only when they begin talking about their experiences with others, thereby creating meaningful and enduring autobiographical records of their experiences (Nelson, 1989, 1993; Reese, 2002). The social construction of personal narratives indeed influences memory reports, including the quantity and quality of children's narratives (Hudson, 1990b; Ratner, 1984; Reese, Haden, & Fivush, 1993). Others have argued that the development of children's selfconcepts (Howe & Courage, 1993; Howe et al., 2003) and their awareness of how memories were acquired (i.e., "source" knowledge) affect the emergence of autobiographical memory (see Howe et al., 2003; Nelson & Fivush, 2004; and Perner, 2000, for a review and discussion).

Developmental Changes in Children's verbal Accounts Once children begin to recall and talk about their experiences, their abilities are often impressive, although significant developmental changes continue through early childhood. Young children typically recall significantly less information than older children, particularly in response to very general prompts such as Tell me what happened and may omit much information that adults consider important (see Ornstein, Baker-Ward, Gordon, & Merritt, 1997; Saywitz, Goodman, Nicholas, & Moan, 1991; and Schneider & Pressley, 1997, for reviews). Preschoolers thus typically require more specific prompts from interviewers (Hamond & Fivush, 1991), to which they respond less accurately than older children do (Bjorklund, Bjorklund, Brown, & Cassel, 1998; Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994). Recent field research shows that children as young as 4 years of age provide proportionally as much information in response to open-ended questions as older children, although the brevity of their responses makes it necessary for interviewers to ask many questions, for example, using the child's prior responses as cues to trigger further recall (Lamb et al., 2003; Chapter 15, this volume). Developmental differences in children's recall are not restricted to mundane experiences or those with a positive emotional valence, but are also evident when children are interviewed about painful, distressing, or traumatic experiences, such as accidental injuries and intrusive medical procedures (e.g., Goodman et al., 1994; Howe, Courage, & Peterson, 1994, 1995; Merritt, Ornstein, & Spieker, 1994; Ornstein, 1995; Peterson, 1999; Peterson & Bell, 1996; Peterson & Whalen, 2001; Salmon, Price, & Pereira, 2002;

456

PIPE, THIERRY, LAMB

see also Peterson, Moores, & White, 2001; see Cordon, Pipe, Sayfan, Melinder, & Goodman, 2004, for a review). Goodman et al. (1994), for example, found differences between 3- to 4, 5- to 6-, and 7- to 10-year-old children in the amount and accuracy of information they recalled about the VCUG, a painful diagnostic procedure involving genital contact (see also Baker-Ward, Ornstein, & Principe, 1997). Similar age differences are evident in children's accounts of abusive experiences (Lamb, Sternberg, & Esplin, 2000; Sternberg, Lamb, Orbach, Esplin, & Mitchell, 2001). Lamb et al. (2003) reported near-linear increases with age in the total amount of information and the amount (though not proportion) of information elicited in response to open-ended prompts as well as in the amount of information provided in the average response provided by alleged victims describing their experiences. Although most studies do not report developmental changes in the accuracy of open-ended free narrative reports, the accuracy of responses to prompts and questions is likely to decrease markedly for younger children, particularly when they must disagree with the interviewer to answer correctly (see Chapter 14 on suggestibility, this volume, for a review; Cassel, Roebers, & Bjorklund, 1996; Greenstock & Pipe, 1996). In sum, although even young children can accurately describe previous experiences, developmental changes in remembering nonetheless take place. Age emerges as an important determinant of event memory in part because it is correlated with other variables that influence memory, including children's prior knowledge and understanding of events, and the effectiveness of the retrieval strategies used. As we shall see, it is sometimes possible to attenuate or even reverse developmental changes in memory when the confounding effects of variables usually associated with age, such as knowledge, are controlled. We next discuss knowledge and other determinants of the way in which events are encoded, which, in turn, affect children's recall.

Encoding Personal Experiences Knowledge and understanding affect how much not only children, but also adults, remember. To the extent that there are age-related changes in the understanding, knowledge, and perceived significance of experiences, age differences in memory are likely to occur. In the classic comparisons of expert and novice chess players, for example, adult experts recalled more chess positions than adult novices (Chase & Simon, 1973). Similar findings have also been obtained with children (Bjorklund, 1987; Bjorklund & Thompson, 1983; Bjorklund & Zeman, 1982; Chi, 1978; Chi & Ceci, 1987; Chi & Koeske, 1983; Landis, 1982), such that the usual age differences in memory can be eliminated or reversed when knowledge and age are pitted against each other (Chi, 1978; Lindberg, 1980). To quote Neisser (2004), "Because young children are less skilled and less knowledgeable than adults, they generally do not structure their experience in memorable ways" (p. 2). Children who have more knowledge about experienced events later recall more details about those events than children with less knowledge (Greenhoot, 2000; McGuigan & Salmon, 2004; Sutherland, Pipe, Schick, Murray, & Gobbo, 2003). For very young children, especially, direct experience is an important source of event knowledge (e.g., Nelson, 1986, 1996). In general, directly participating in an event is

16.

DEVELOPMENT OF EVENT MEMORY

457

likely to result in stronger and/or more accessible memory traces, for both adults and children, than is being a bystander or observer, or audience for a story, about the same event (Murachver, Pipe, Gordon, Fivush, & Owens, 1996; Roediger, McDermott, Pisani, & Gallo, 2004; Rudy & Goodman, 1991; Tobey & Goodman, 1992). Tobey and Good~ man (1992), for example, found that 4~year~old children who participated in a real~life event (a Simon Says game) freely recalled central actions more accurately than children who merely observed the same event on video, and Rudy and Goodman (1991) found that 4~year~olds who were direct participants in a real~life event were less susceptible to misleading questions than children who observed the real~ life event. Similarly, Murachver et al. (1996) found that children who participated in a contrived interaction with an adult "pirate" recalled more information than those who read a story about visiting the pirate. Their free recall was also more accurate than that of children who only watched the event or heard about the event. Even when recall was supported by behavioral reen~ actment, children who were read the story were significantly less accurate than those who had participated or observed. Whether direct participation leads to stronger mem~ aries than other sources for younger children has not been examined directly, however. Participation may strengthen the resulting memory trace as well. Theorists agree that memory trace strength can vary (Brainerd & Reyna, 1990; Ceci, Toglia, & Ross, 1988), such that stronger or weaker memory traces can be created. Comparing 3~ and 5~year~old children's memories of events about which they heard a story (narrative con~ clition) or in which they directly participated, Gobbo, Mega, and Pipe (2002) found that children who participated recalled more details accurately than children in the narrative condition. To determine whether this difference in recall was due to differences in the strength of the memory traces created by participating rather than only hearing a story about the event, Gobbo et al. (2002) equated children's level of learning (or encoding) by having children in each condition reach a criterion level of learning. This criterion was achieved by repeated exposure of the children to the event. Children who heard about the event to a criterion level of learning recalled as many details as children who participated in the event (see Murachver et al., 1996, for compatible findings). Thus, although participating in an event creates a stronger memory trace than merely being told about it, repeated experiences can reduce or eliminate some of the differences be~ tween sources. Adults often talk to children about anticipated as well as past events and discuss the activities in which they are taking part (Fivush, 1998; Fivush, Haden, & Adam, 1995; Haden & Fivush, 1996). Talking about events while (Ornstein, Principe, Hudson, Gar~ don, & Merritt, 1997; Tessler & Nelson, 1994) or after (Goodman et al., 1994; Hudson, 1990b) they are taking place may enhance children's event recall. In the study by Tessler and Nelson (1994), children's recall of an event in which they had participated reflected those aspects of the event talked about by the parent and/or the child during the event. Ornstein, Principe, et al. (1997) reported compatible findings in a study of children's memory of a painful medical procedure (see also Haden, Ornstein, Eckerman, & Oidow, 2001). Goodman et al. (1994) similarly found that parents' retrospective reports of their post~event interactions with their children concerning a painful medical procedure that they had experienced were correlated with the children's subsequent ability to resist

458

PIPE, THIERRY, LAMB

misleading questions about the procedure. In other words, although parental discussion was not related to correct recall, it was negatively correlated with errors in response to misleading questions. Furthermore, Sutherland et al. (2003) found that information presented prior to an experience is also useful, at least when the information is specific to the experience rather than being globally related to the topic (cf. DeMarie-Dreblow, 1991), although talk during and after an event is more effective than talk before (McGuigan & Salmon, 2004). Event-related discussion may strengthen memories for several reasons. Discussing events in advance, for example, increases knowledge about the event, thereby rendering it more memorable, whereas post-event discussion may ensure rehearsal, which consolidates the memory or, following long delays, ensures retrieval, which reactivates the memory. Moreover, prior, contemporaneous, or retrospective discussions may all highlight important factors on which children should focus and provide appropriate verbal labels for actions and objects, thereby facilitating memory. Many of the experiences explored in forensic investigations may have been poorly understood by alleged victims, especially when the alleged victims were young children. In addition, disclosures of sexual abuse are often made months or even years after the abusive incident(s) (London, Bruck, Ceci, & Shuman, 2005; in press), with little or no discussion with others in the intervening period. Victims of childhood sexual abuse are frequently embarrassed or afraid, or have been threatened not to tell (Cederborg, Lamb, & Laurell, in press). In the absence of discussion and opportunities for verbal recall, such experiences may not be remembered in detailed or coherent narrative form.

Repeated Experiences Children's knowledge and the resulting memory representations can also be affected by the number of times they have experienced an event. Victims of sexual abuse are frequently abused repeatedly, sometimes over long periods of time. When children experience similar events, they tend to form general event representations (or scripts) of "typical" events rather than particular incidents (Farrar & Goodman, 1992; Hudson, Fivush, & Kuebli, 1992; Nelson & Gruendel, 1981). These general event representations can help children to predict what is going to occur, understand what is happening during an event, and guide the recall and retrieval of familiar events (Brainerd & Ornstein, 1991; Brewer &Nakamura, 1984; Hudson, 1986; Nelson, 1986, 1993; Nelson & Gruendel, 1981). Repeated experience may strengthen event memories, with children recalling more details than if they experienced the event only a single time (Bauer & Fivush, 1992; Fivush, Kuebli, & Clubb, 1992; Hudson, 1990a; Hudson & Nelson, 1986; Murachver et al., 1996; see Powell & Thomson, 2003 for review). Furthermore, children's memories of details that recur across experiences are also more accurate and more resistant to suggestion and misinformation effects (Connolly & Lindsay, 2001; Gobbo et al., 2002; McNichol, Shute, & Tucker, 1999; Powell, Roberts, Ceci, & Hembrooke, 1999). In contrast, aspects of the experience that change across reoccurrences tend to be omitted from children's event reports, so children's accounts may reflect the common features and basic structure but ignore the details that vary from time to time. The changing components may also be more vulnerable to suggestion, at least under

16.

DEVELOPMENT OF EVENT MEMORY

459

some conditions (Connolly & Lindsay, 2001; Fasig, 1999; McNichol et al., 1999; but cf. Powell et al., 1999; Powell, Roberts, & Thomson, 2000). Prior experiences can also have adverse effects on children's event recall. Ornstein et al. (1998) showed that, after a 12-week delay, children's recall of a pediatric exam became more reliant on their general knowledge, as the children reconstructed the examination or filled in missing details with information about what usually happens (see also Myles-Worsley, Cromer, & Dodd, 1986). That is, with the passage of time, script-related intrusions began to appear in children's accounts. Other sources of knowledge, not only children's own experiences, can also contribute to script-related errors. Fivush, Hudson, and Nelson (1984; Hudson & Fivush, 1991) similarly found that children relied more on general knowledge about an event (visiting a museum) to provide basic information about a specific instance of that event (a visit to an archaeological museum) over time, although they were still able to access some information about specific events after very long delays (of up to 6 years). It appears that the confusion of details across episodes is more likely among younger than older children, particularly over time (Farrar & Goodman, 1992; Powell et al., 1999). Younger children are more likely to forget the source of their memories than are older children (Foley & Johnson, 1985; Lindsay, Johnson, & Kwon, 1991). Moreover, the passage of time between the experience and recall, which is likely to be months or even years in forensic contexts, increases the tendencies both to rely on general event knowledge (Myles-Worsley et al., 1986; Slackman & Nelson, 1984) and to confuse details from the different episodes (Farrar & Goodman, 1992; Hudson, 1990a; Roberts & Powell, 2001; Slackman & Nelson, 1984). The effects of knowledge and repeated experience on memory are quite significant from a practical perspective. In forensic contexts, especially when child sexual abuse is alleged, it is not uncommon for children to be asked about incidents that have occurred repeatedly, sometimes over long periods of time. In the courtroom, what happened on one specific occasion is often critically important. Yet the requirement that children recall a single episode distinct from other similar experiences may be very challenging for young children. Despite remarkable memory for details of what happened (Fivush & Hamond, 1990; Powell et al., 1999), young children are generally not as accurate when identifying details associated with one particular incident of a repeated experience (Farrar & Goodman, 1990; Roberts & Powell, 2001). For example, when children between the ages of 4 and 8 years were asked to recall the final instance of an event that was experienced six times, with minor variations in some details but the same basic event structure, children frequently recalled details from the earlier instances, rather than the final instance (Farrar & Goodman, 1992; Powell & Thomson, 1997). Powell et al. (1999) similarly found that children provided less accurate accounts of repeatedly experienced events because they reported details from one incident as if they had occurred in another incident.

Monitoring the Source of Memories Because of the forensic importance, researchers have begun to examine how to enhance the accuracy with which children recount specific experiences, as distinct from other, similar experiences (see Roberts, 2002; Roberts & Powell, 2001, for review). The ability

460

PIPE, THIERRY, LAMB

to recall specific incidents of a repeated event or distinguish between personal and vicarious experiences requires engaging in a process known as source monitoring (Johnson, Hashtroudi, & Lindsay, 1993). In order to recall details about specific incidents, children must discriminate the source of the detail, including whether it happened the first time or the last time, in one place or another, or in the course of discussions with another person. Field work examining relations between children's source awareness and their recall of multiple experiences of alleged abuse supports the importance of source monitoring for episodic recall. For example, Thierry, Lamb, and Orbach (2003) found that 3- to 11-year-old alleged sexual abuse victims who were more aware of the source of their knowledge recalled more episodic details (but not more generic details) about multiple experiences of abuse than children who were less aware of source. Attempts to improve children's source monitoring performance and ability to recall specific events without intrusion of information about other similar experiences have met with mixed success. In general, children are less likely to make source errors when asked for open-ended, free recall accounts than when asked specific questions (Roberts & Powell, 2001). Explicitly asking children about the source of information they have reported-for example, whether it was something that they saw or something that someone told them about-can be useful with older children (Lindsay & Johnson, 1989; Zaragoza & Lane, 1994), but not with very young (3- or 4-year-old) children (Leichtman, Morse, Dixon, & Spiegel, 2000; Quas, Schaaf, Alexander, & Goodman, 2000; see Roberts & Powell, 2001; Roberts, 2002, for reviews). Thierry, Spence, and Memon (2001) found that 5- to 6-year-old, but not 3- to 4-year-old children who were required to actively monitor the source of information about something they either witnessed or saw on television were better at monitoring source when subsequently asked misleading questions. Poole and Lindsay (2002) also found that having children monitor the source of seen and heard events in a training phase helped 7- to 8- year-olds, but not 3- to 4-year-olds to distinguish among activities. When children were given more extensive source-monitoring training, however, the memory accuracy of 3- to 4-year-olds was improved in the study of Thierry and Spence (2002). Conversely, asking children to recall experiences regardless of source and then asking for source attributions appears not to reduce confusion (Thierry, Goh, Murray, & Pipe, 2005; Powell & Thomson, 1997). Moreover, adults who repeatedly recalled seen and imagined objects without regard to their source confused more details and made more source monitoring errors in subsequent memory tests (Henkel, 2004). The fact that children may have difficulty recalling specific incidents of multiply experienced events has led many court systems in the United States to relax the requirement that child witnesses identify the time of the individual incidents of multiply experienced events (Roberts, 2002), because confusion between episodes does not necessarily cast doubt on the accuracy or credibility of young witnesses (see also Roberts & Powell, 2001). Children's general event representations or scripts are affected, however, not only by repeated experiences, but also by conversations, television, books, and so on (Roberts & Powell, 2001; Sutherland et al., 2003). The extent to which these other sources of information influence children's generalized event accounts remains largely unexplored, although recognition procedures show that children confuse not only episodes they ex-

16.

DEVELOPMENT OF EVENT MEMORY

461

perienced, but also those they experienced and those heard about (Roberts & Powell, 2001). In forensic contexts, recognition procedures are generally not recommended, so it is important to determine whether information obtained from vicarious sources is incorporated into children's free narrative accounts of their own experiences. Preliminary evidence suggests that this is less likely (Sutherland et al., 2003), although this pattern changes when children have been exposed to repeated suggestions (Bruck, Melnyk, & Ceci, 2000; Leichtman & Ceci, 1995).

Remembering Events over Long Delays Once in memory, how durable are children's memories of their experiences? When children (and adults) recall neutral or positive events over long time periods, forgetting is typically extensive (e.g., Jones & Pipe, 2002; Ornstein, Baker-Ward, et al., 1997; Salmon & Pipe, 2000), and children may require many cues and props to facilitate recall (Hudson & Fivush, 1991; but cf. Fivush & Schwarzmueller, 1998), if, indeed, they can do so at all (Goodman, Batterman-Faunce, Schaaf, & Kenney, 2002; Pillemer, Picariello, & Pruett, 1994). Pillemer and colleagues showed, for example, that although both 3- and 4-year-old children remembered what happened at school when a fire alarm went off and they were interviewed soon after, none of the younger and only some of the older children remembered it when interviewed 7 years later (Pillemer, 1993; Pillemer, Picariello, & Pruett, 1994; see also Drummey & Newcombe, 1995). Children can remember other experiences after very long, forensically relevant delays. Ornstein and colleagues showed that even quite young children recalled a pediatric examination extremely well after delays of up to 6 weeks (e.g., Baker-Ward, Gordon, Ornstein, Larus, & Clubb, 1993; Clubb, Nida, Merritt, & Ornstein, 1993; Merritt et al., 1994; Ornstein, Shapiro, Clubb, Follmer, & Baker-Ward, 1997). Medical examinations are likely to be familiar events for many young children, and the good recall in this study might reflect children's knowledge about and understanding of the examination. Other studies suggest that some traumatic experiences about which children presumably had little prior knowledge may be remembered better over long delays than more mundane or neutral experiences. In a study of children's recall of the VCUG (the painful diagnostic procedure described earlier), although children did recall somewhat less when interviewed 6 weeks as opposed to shortly after the test, the change was not statistically significant (Merritt et al.,1994). Burgwyn-Bailes, Baker-Ward, Gordon, and Ornstein (2001) similarly reported good recall of hospital treatment that included suturing of facial lacerations. Children who were between the ages of 3 and 7 years at the time of the injury recalled a similar number of features when interviewed 1 year later and a few days and 6 to 8 weeks after the suturing. However, more false alarms occurred in response to suggestive questions after the long delay, suggesting that the memory representations did change over time. Children also remembered injuries and the hospital treatment that resulted, in studies by Peterson and her colleagues (e.g., Peterson & Bell, 1996; Peterson & Whalen, 2001). When children were interviewed 6 months after the injury, they reported significantly less information about both the injury and the hospital events than when they were

462

PIPE, THIERRY, LAMB

interviewed soon after the events (Peterson & Bell, 1996), but when re~interviewed 2 years and 5 years after the injury, the children recalled less about the hospital event than they had soon after, but as much as before about the injuries (Peterson, 1999; Pe~ terson & Whalen, 2001). Nonetheless, the children's accounts of both the injury and hospital events were less accurate after the very long delays than in the initial interview, suggesting changes in the content of the memories, if not in the amount recalled. Reinterviewing children about Hurricane Andrew, however, Fivush, Sales, Gold~ berg, Bahrick, & Parker (2004) showed that children's reports can become more rather than less detailed after long delays. The children, who had been 3 to 4 years old at the 6 years later. The effects were quite dra~ time of the hurricane, were re~interviewed matic, with the now 9~ to lO~year~old children reporting almost twice as much informa~ tion as when interviewed shortly after the event (Parker, Bahrick, Lundy, Fivush, & Levitt, 1998). Why are some experiences remembered better over time than others? Stress, arousal, and even trauma seem unlikely, on their own, to provide useful explanations. Research on the effects of stress on children's event reports has yielded inconsistent findings. Some researchers have reported that high levels of stress are associated with improved memory in the laboratory (Goodman, Hirschman, Hepps, & Rudy, 1991; Warren & Swartwood, 1992), some have reported that high levels of stress are associated with poorer memory (Bugental, Blue, Cortez, Fleck, & Rodriguez, 1992; Merritt et al., 1994; Peters, 1987, 1991), and others have reported no relationship at all between stress and et al., 1993; Howe et al., 1994). For example, Howe et al. (1994) recall (Baker~Ward found no relationship between the level of stress (reported by the parents) and the amount of information recalled by their children either 3 to 5 days or 6 months after an emergency room procedure. By contrast, Goodman et al. (1991) found that children who showed higher levels of anxiety during a medical procedure reported the incident more accurately than children who simply had a washable tattoo applied. Furthermore, children when Bahrick, Parker, Fivush, and Levitt (1998) classified 3~ and 4~year~old medium~, and low~stress groups based on the extent of their exposure to into high~, and low~stress groups recalled Hurricane Andrew, they found that children in the high~ the least amount of information about the hurricane, whereas children in the medium~ stress group recalled the most information, suggesting that some stress improves recall, whereas too much stress impedes it. However, these differences were not maintained 6 years later. when the children (now 9 and 10 years old) were re~interviewed Fivush (1998) concluded that, although children may recall more details about stressful than about nonstressful events, developmental changes in the representation and recall of these events appear similar. Other scholars have also concluded that the same variables that influence memories of more mundane or positive experiences affect memories of stressful and traumatic experiences (e.g., Cordon et al., 2004; Howe, 1997), but recent work suggests that children may recall negative and positive events somewhat differently (Fivush, Hazzard, Sales, Safati, & Brown, 2003). For instance, when narrat~ ing emotionally positive events (e.g., a trip to a amusement park), 5~ to 12~year~olds re~ called more information about objects and people and included more descriptive de~ tails than when recalling emotionally negative events (e.g., illness or death of a family

16.

DEVELOPMENT OF EVENT MEMORY

463

member). When recalling emotionally negative events, in contrast, the children included more information about their thoughts and emotions and recounted these experiences more coherently than when recalling positive events (Fivush et al., 2003). Because a diverse range of experiences has been studied, it is not surprising that no simple relation among stress, trauma, and memory has been found. Whereas some stressful experiences are highly distinctive and salient, others are much less so. In addition to distinctiveness, some traumatic experiences may be associated with numerous reminders, whereas other traumatic experiences may have few, if any, reminders. We know from studies of infants (Rovee-Collier & Hayne, 2000), toddlers (Sheffield, 2004), and children (Priestley, Roberts, & Pipe, 1999) that even relatively brief reminders can powerfully affect the accessibility of memories over long time periods. Not surprisingly, in studies showing that traumatic experiences are well recalled over long delays, children were likely to have been exposed to repeated reminders, providing opportunities for reactivation of the memories (e.g., Fivush et al., 2004; Peterson & Whalen, 2001). In addition, Peterson and Whalen (2001) suggest that increases in recollection may be attributable to improvements in narrative skill, rehearsal (as a result of repeated interviews), and discussions within the family, leading to more detailed, albeit second-hand, information. Consistent with evidence that repeated recall attempts may lead to improved recall (hypermnesia) in children (LaRooy, Pipe, & Murray, 2005) as well as in adults (e.g., Bluck, Levine, & Laulehre, 1999; Bornstein, Liebel, & Scarberry, 1998; Dunning & Stern, 1992; Scrivner & Safer, 1988), this enhancement in recall over time might also reflect greater memory accessibility. Few researchers have studied the association between the severity of abuse (presumably a correlate of stressfulness) and children's recall, probably because there is no consensus regarding the ways in which the severity of abuse should be measured, and a variety of events (ranging from exposure to rape at gunpoint) are defined as sexual abuse. In addition, whether the abuse is even stressful to children may vary, depending on such factors as the age of the child and the identity of the perpetrator (see Cederborg et al., in press). With better definitions of child abuse and access to accounts of a larger number of cases, researchers may be able in the future to determine whether different types of abuse are recalled differently as a function of the many interrelated factors that characterize real-world experiences of abuse.

Developmental Differences in the Retrieval and Reporting of Information Although many researchers have examined the development of memory and factors affecting encoding, others have focused on the retrieval of information and strategies for facilitating the retrieval and reporting of memories. In forensic interviews, the interviewer has a critical influence on how much and how accurately information is retrieved from memory. In general, younger children are less likely to retrieve event details from memory spontaneously and need more focused prompting or cueing-for example, in the form of questioning-than older children (Fivush & Hamond, 1989; Hamond & Fivush, 1991; Lamb et al., 2003). Laboratory research suggests that developmental

464

PIPE, THIERRY, LAMB

differences in memory performance are greater in recall than in recognition memory tasks (e.g., Mandler & Stein, 1974), but recognition tasks that take the form of yes-no or multiple-choice questions pose a problem when used in real-world contexts, because they are more likely to elicit erroneous or inaccurate responses (see Lamb et al. and suggestibility chapters in this volume). A number of alternative strategies for enhancing young children's event reports, including the use of dolls and prop items, drawing during the interview, and reinstatement of the event context, have been explored in both analogue and field studies, and it is to these that we now tum. Several researchers have examined whether dolls and props aid children's event recall. Comparing the amount of information that children report with and without dolls, some researchers have shown that children report more information when dolls are used than without dolls, but at the expense of accuracy (Bruck, Ceci, Francouer, & Renick, 1995; Goodman, Quas, Betterman-Faunce, Riddlesberger, & Kuhn, 1997). For example, 5- to 7-year-old girls reported more free-recall details when the dolls were used as demonstration aids (Saywitz et al., 1991), but these free-recall details (collapsed across verbal reports and behavioral demonstrations) included more inaccurate details than when the dolls were not used. Similarly, Goodman et al. (1997) found that 3- and 4-yearolds reported more information about genital touching during free recall with the dolls than without the dolls, with more errors included in the doll reports. Unlike Saywitz et al. (1991), Goodman et al. (1997) found that the dolls enhanced the amount of information recalled by the 5- to 6- and 7- to 10-year-olds without any increase in errors. In field studies, where accuracy cannot be assessed, however, Lamb, Hershkowitz, Stemberg, Boat, and Everson (1996) reported no increase in the amount offorensically relevant detail reported when dolls were employed. In addition, Thierry, Lamb, Orbach, and Pipe (2005) found that dolls were associated with play, ambiguous enactments, and contradictions. Concern about the inaccuracy of doll-associated reports has prompted decreases in the forensic use of dolls in the United States, over the last decade, although some forensic experts still advocate their use during interviews with allegedly abused children (Everson & Boat, 2002). Others have advocated greater caution, especially when young children are involved (Poole & Lamb, 1998). Allowing children to draw while they tell about experienced events may also aid the retrieval of information from memory. When allowed to draw and tell what happened, children reported more details about salient events than when only asked to tell (Gross & Hayne, 1998; Wesson & Salmon, 2001). For example, Wesson and Salmon (2001) found that 5- and 8-year-olds who drew and enacted emotionally laden events reported twice as much verbal information as when they simply told what happened. It appears that drawing may help cue children's recall of details (Butler, Gross, & Hayne, 1995; Gross & Hayne, 1999) and may also help children with less developed expressive language abilities by providing external cues that help to organize the children's narratives (Salmon, Roncolato, & Gleitzman, 2003). As with dolls (Thierry et al., 2005), however, asking children to draw may also result in less accurate reports under suggestive interviewing conditions (Bruck et al., 2000) and may not always enhance recall (Salmon & Pipe, 2000).

16.

DEVELOPMENT OF EVENT MEMORY

465

In other efforts to enhance children's memory retrieval, Saywitz and her colleagues (Saywitz & Snyder, 1996) developed a memory training technique, known as the narra~ tive elaboration technique (NET). This technique helps to structure children's recall by providing external cues that might aid retrieval of event details. Children are presented with cards that cue their memories of various event components, such as the partici~ pants, settings, actions, and affective states associated with the experienced events. Seven~ to 8~ and 10~ to 11~year~olds trained with the use of the NET produced more ac~ curate details about an event than children in control groups (Saywitz & Snyder, 1996). Although the narrative elaboration procedure has not been used in any forensic studies, the NICHD interview protocol (see Lamb et al., this volume) is designed to enhance narrative elaboration with the use of somewhat similar techniques. In the NET, the in~ terviewer provides category cues, however, whereas the child's own narrative provides cues to prompt further recall in the NICHD interview protocol. Reinstating the context in which the alleged experiences occurred, either physically (i.e., returning to the scene of the crime) or mentally (i.e., guided reflection on aspects of the context of the crime), can also aid retrieval, although the findings have been in~ consistent. The prediction that contextual cues should enhance memory performance is based on Tulving's encoding specificity principle, whereby cues that were present during memory encoding help to bring into consciousness details of an event that might not have been retrieved without such cues (Tulving, 1983). Many researchers have shown in the laboratory that children recall more information, without adverse effects on ac~ curacy, if they are interviewed where an event occurred (Gee & Pipe, 1995; Pipe & Wil~ son, 1994; Price & Goodman, 1990). A field study by Hershkowitz et al. (1998) sug~ gested that taking children back into the context in which the alleged incidents had occurred led the children to report additional information. In this study, however, all children were interviewed both at the investigators' office and back at the scene, and hence the effects of context reinstatement (returning to the scene) were confounded with the effect of repeated interviewing. In a subsequent study, Orbach, Hershkowitz, Lamb, Sternberg, and Horowitz (2000) reported no difference between the numbers of details reported by victims who were interviewed in the investigators' offices and those who were interviewed at the scene of the alleged abuse. Hershkowitz, Orbach, Lamb, Sternberg, and Horowitz (2001) also showed that men~ tal context reinstatement facilitated children's recall. Four~ to 13~year~olds who were required to mentally reinstate the context of an alleged sexual abuse incident (children were instructed to "close your eyes and think about that time ... think about sounds or voices you could hear ... smells you could smell") produced proportionally more sub~ stantive details in response to open~ended invitations than children who were not in~ structed to reinstate the context mentally (Hershkowitz et al., 2001). Because these were field studies, the accuracy of the details could not be assessed, but, taken together with laboratory~based studies, the results of these studies suggest that mental context reinstatement may be a promising technique. Other useful techniques that could be incorporated into interview protocols have not been studied systematically in the field or laboratory. For instance, the finding that children's source awareness is related to recall of specific instances of abuse (Thierry,

466

PIPE, THIERRY, LAMB

Lamb, & Orbach, 2003) supports Ferner's assertion about the importance of source memory for episodic recall. Providing children with opportunities to recall events from different sources during the preliminary phases of interviews might enhance children's source awareness and facilitate their subsequent recall of specific events. Source awareness training might have the added advantage of enhancing the accuracy of children's recall as well. Researchers have shown that preschoolers are less likely to incorporate other sources of information (e.g., details heard or seen on television) into their reports of real-life experiences if they are given source-monitoring training (Ghetti & Alexander, 2004; Thierry & Spence, 2002; Thierry et al., 2001).

Individual Differences in Recall There are often marked individual differences in the amount and/or accuracy of the information children recall. Individual differences may be forensically relevant to the extent that they predict, for example, whether a child is likely to be suggestible or easily misled, to lie, or to benefit from a particular interview strategy (Gordon et al., 1993; Pipe & Salmon, 2002). More generally, understanding sources of variation in children's recall may help place their testimony in context and help interviewers understand why some children may say less about their experiences than others do. Most studies examining the role of individual differences in memory have focused on identifying those children who are likely to be vulnerable to suggestive questioning (see Bruck & Melnyk, 2004; Quas, Qin, Schaaf, & Goodman, 1997). A small but growing number of studies have also been concerned with both cognitive and personality variables that moderate children's recall. In addition to knowledge (see above), intelligence is positively correlated with event recall (Elischberger & Roebers, 2001; Geddie, Fradin, & Beer, 2000), although the relation may be stronger for older (e.g., 8-10-yearold) children (Roebers & Schneider, 2001) and may also depend on how interviews are conducted (Brown & Pipe, 2003). Gordon et al. (1993) found that language and narrative skills were positively related to recall of a pediatric exam by 5-year-old but not 3-year-old children (see also Salmon et al., 2003), but Greenhoot, Ornstein, Gordon, and Baker-Ward (1999) did not find a similar correlation, and Quas (1998; Quas et al., 1997) failed to find any relation between language ability (as measured with the Peabody Picture Vocabulary Test) and children's free recall. Brown and Pipe (2003) found that WISC-III vocabulary scores were related to event recall, whereas a measure of narrative ability was not (see also Kleinknecht, 2001). The ability to form long-term memories may be affected by children's understanding of the source of their knowledge (Gopnik & Graf, 1988; Perner, 2000; Welch-Ross, 1995)-for example, whether the events occurred in real life or were merely imagined or heard about Oohnson et al., 1993)-particularly because the ability to attribute source emerges around the same age (i.e., 3 to 5 years) as the first autobiographical memories that adults later remember (Pratt & Bryant, 1990). In fact, Perner (2000) suggested that children are incapable of recalling personally experienced events (i.e., episodic recall) if they lack "experiential awareness" and are unaware of the origins of knowledge. Perner and Ruffman (1995) found that 3- to 4-year-old children who correctly identified the

16.

DEVELOPMENT OF EVENT MEMORY

467

source of their knowledge (i.e., whether they saw where an item was located or were told by someone where it was located) reported more information in free recall than children who did not correctly attribute the source of their knowledge. Children who could identify the origin of their knowledge were thus less dependent on cues provided by the interviewer than those who could not identify its origin (Perner & Ruffman, 1995). Social and emotional factors, particularly those relating to attachment and temperament, may affect children's event reports, by mediating the way in which experiences are appraised, encoded, and organized in memory and how they are subsequently retrieved. Goodman et al. (1997) argued that attachment may influence children's accounts of emotionally laden experiences. For example, using parental attachment style as an indicator of the child's attachment security, they argued that children whose parents had an "anxious-ambivalent" attachment style (i.e., not securely attached) may be preoccupied with locating a source of security during stressful experiences, with the result that their encoding of the experience will differ from that of children whose parents showed a secure attachment style. In turn, securely attached children may be better able to regulate their emotions, aiding both encoding and subsequent recall. Furthermore, Goodman and colleagues found that parental attachment style accounted for the relation between stress and memory in two studies (Goodman et al., 1994, 1997). Children whose parents reported insecure attachment were both more stressed during a painful and distressing medical procedure and subsequently made more errors when recounting it than children whose parents reported secure attachment styles. Laboratory research with young (3-year-old) children also suggests a relation between memory and child attachment status, with securely attached children recognizing more positive events than negative events from a puppet show, whereas the reverse was true for children who were insecurely attached (Belsky, Spritz, & Crnic, 1996). With respect to children's temperament, findings have been mixed. Gordon et al. (1993) examined temperament in relation to event memory with the use of the Temperament Assessment Battery (TABC; Martin, 1988) and found that three dimensions from the battery (approach-withdrawal, emotionality, and adaptability) were related to recall of a medical examination by both 3- and 5-year-old children. Merritt et al. (1994) likewise found that adaptability and approach-withdrawal were related to children's recall of the VCUG procedure. However, Baker-Ward, Burgwyn, Ornstein, and Gordon (1995) did not find any association between temperament and recall of minor surgery for facial lacerations, although coping style during the surgery was related to recall (see Pipe & Salmon, 2002; Quas, 1998; Quas et al., 1997, for reviews). Roebers and Schneider (2001) found that shy children answered specific questions less accurately than did children who were less shy, and manageability was positively related to the number of intrusions reported by 3-year-olds when they re-enacted an event using dolls and props (Greenhoot et al., 1999). Greenhoot et al. (1999) suggested that the more manageable or "easy" children may be more compliant and eager to please, which, in turn, may have led them to produce more intrusion errors. In addition, Greenhoot et al. (1999) found that less persistent 3-year-olds were more likely to produce errors than children who were more persistent and thus better able to attend to the tasks. When showing what happened, behaviorally, Salmon et al. (2003) found that children with higher levels of

468

PIPE, THIERRY, LAMB

"effortful control" (i.e., the ability to shift and refocus attention in order to regulate behaviors and emotions) produced more details than children with lower levels of effortful control. More research is needed before we can make the kinds of predictions most useful to the legal system about the reliability of particular children, or the best interviewing strategies to adopt with particular children. Nonetheless, an understanding that individual differences are likely to exist and that the quality of children's testimony is affected by personality and temperament may ultimately enhance the quality of interviewing.

CONCLUSIONS

The quality and quantity of information that children report about events depend on a number of factors surrounding the alleged incident(s), the children themselves, and the way in which the report was obtained. Children who merely observe events may not be able to report as many details as those who are direct participants (Murachver et al., 1996; Rudy & Goodman, 1991; Tobey & Goodman, 1992). Likewise, children who have experienced an event for which they lack knowledge or understanding may not report as many details, particularly when there has been a long delay since the event being recalled. Multiple experiences of an event can have both positive and negative effects on memory. On the positive side, multiple experiences of similar events may strengthen the resulting memory for details that consistently occurred in each incident (e.g., Gobbo et al., 2002). On the negative side, younger children are especially likely to confuse details that varied across individual incidents of repeated events (Powell et al., 1999). In addition, children sometimes intrude script-consistent information into their event reports when they have general knowledge about said events (Ornstein et al., 1998). Researchers have had some success developing procedures that enhance the amount of information children report without adverse effects on accuracy. In the laboratory setting, for example, the NET (Saywitz & Snyder, 1996) appears to help children as young as 5 years of age report more information. Although its utility has not yet been examined in laboratory analogue studies, the NICHD investigative interview protocol (Lamb et al., 2003; Orbach, Hershkowitz, Lamb, Sternberg, Esplin, et al., 2000; Sternberg et al., 2001) enhances the amount of information that children report in response to more openended prompts. Because details elicited from open-ended prompts are more likely to be accurate than details elicited from focused questions (e.g., leading or suggestive prompts), the protocol encourages the use of open-ended questions and discourages the use of leading and suggestive questions. If children must be repeatedly interviewed about alleged events, the use of such protocols in these instances might be less problematic than unstructured interviews, because more of the information would be obtained from recall memory. During the past two decades, there has been a dynamic interchange between researchers working in the laboratory and those working in the field (see Pipe, Lamb, Orbach, & Esplin, 2004, for a review), and this must continue; there are areas in which, at present, we have insufficient evidence on the basis of which to make strong recommendations, in particular, relating to the most appropriate facilitative techniques for helping

16.

DEVELOPMENT OF EVENT MEMORY

469

children of different ages recount their experiences. Moreover, although memory is the cornerstone of accurate eyewitness testimony, in forensic contexts remembering often takes place in very specific motivational and emotional contexts not easily captured in the laboratory setting. Laboratory and analog research on memory development is most informative with respect to investigative processes when there is convergence of findings with those based on more applied research studies, such as those conducted in the field. Continued collaboration between laboratory and field researchers will help to further our understanding of both basic processes in memory development and the strengths and limitations of child witnesses.

REFERENCES Bahrick, L., Parker,]., Fivush, R., & Levitt, M. (1998). The effects of stress on young children's memory for a natural disaster. Journal of Experimental Psychology: Applied, 4, 308-331. Baker-Ward, L., Burgwyn, E. 0., Ornstein, P. A., & Gordon, B. N. (1995, April). Children's reports of a minor medical emergency procedure. Paper presented at the biennial meeting of the Society for Research in Child Development, Indianapolis, IN. Baker-Ward, L., Gordon, B. N., Ornstein, P. A., Larus, D. M., &Clubb, P. A. (1993). Young children's long-term retention of a paediatric examination. Child Development, 64, 1519-1533. Baker-Ward, L., Ornstein, P. A., & Principe, G. F. A. (1997). Revealing the representation: Evidence from children's reports of events. In P. W. van den Broek, P. ]. Bauer, & T. Bourg (Eds.),

Developmental spans in event comprehension and representation: Bridging fictional and actual events (pp. 79-107). Mahwah, NJ: Lawrence Erlbaum Associates. Bauer, P., & Fivush, R., (1992). Constructing event representations: Building on a foundation of variation and enabling relations. Cognitive Development, 7, 381-401. Bauer, P., Van Abbema, D., Wiebe, S., Cary, M., Phill, C., Burch, M., & Bauer, P. (2004). Props,

not pictures, are worth a thousand words: Verbal accessibility of early memories under different conditions of contextual support. Applied Cognitive Psychology, 18, 373-392. Bauer, P. ]., & Wewerka, S. S. (1997). Saying is revealing: Verbal expression of event memory in the transition from infancy to early childhood. In P. W. van den Broek, P. ]. Bauer, & T. Bourg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and actual events (pp. 139-168). Mahwah, NJ: Lawrence Erlbaum Associates. Belsky,]., Spritz, B., & Crnic, K., (1996). Infant attachment security and affective-cognitive information processing at age 3. Psychological Science, 7, 111-114. Bjorklund, D. F. (1987). How age changes in knowledge base contribute to the development of children's memory: An interpretive review. Developmental Review, 7, 93-130. Bjorklund, D., Bjorklund, B., Brown, R., & Cassel, W. (1998). Children's susceptibility to repeated questions: How misinformation changes children's answers and their minds. Applied Developmental Science, 2, 99-111. Bjorklund, D. F., & Thompson, B. E. (1983). Category typicality effects in children's memory performance: Qualitative and quantitative differences in the processing of category information. Journal of Experimental Child Psychology, 35, 329-344. Bjorklund, D. F., & Zeman, B. R. (1982). Children's organization and metamemory awareness in their recall of familiar information. Child Development, 53, 799-810. Bluck, S., Levine, L. ]., & Laulhere, T. M. (1999). Autobiographical remembering and hypermnesia: A comparison of older and younger adults. Psychology and Aging, 14, 671-682. Bornstein, B. H., Liebel, L. M., & Scarberry, N.C. (1998). Repeated testing in eyewitness memory: A means to improve recall of a negative emotional event. Applied Cognitive Psychology, 12, 119-131.

470

PIPE, THIERRY, LAMB

Brainerd, C., & Ornstein, P. (1991). Children's memory for witnessed events: The developmental backdrop. In]. Doris (Ed), The suggestibility of children's recollections (pp. 10-20). Washington, DC: American Psychological Association. Brainerd, C.]., & Reyna, V. F. (1990). Gist is the grist: Fuzzy-trace theory and the new intuitionism. Developmental Review, 10,3-47. Brewer, W. F., & Nakamura, G. V. (1984). The nature and functions of schemas. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social cognition (Vol. 1, pp. 119-160). Hillsdale, NJ: Lawrence Erlbaum Associates. Brown, D., & Pipe, M.-E. (2003). Individual differences in children's event memory reports and the narrative elaboration technique. Journal of Applied Psychology, 88, 195-206. Bruck, M., Ceci, S. ]., Francouer, E., & Renick, A (1995). Anatomically detailed dolls do not facilitate preschoolers' reports of a pediatric examination involving genital touching. Journal of Experimental Psychology: Applied, 1, 95-109. Bruck, M., & Melnyk, L. (2004). Individual differences in children's suggestibility: A review and synthesis. Applied Cognitive Psychology. Special Issue (A Crossman, M. H. Scullin, & L. Melnyk, Eds.). Individual and Developmental Differences in Suggestibility, 18, 947-996. Bruck, M., Melnyk, L., & Ceci, S. ]. (2000). Draw it again Sam: The effect of drawing on children's suggestibility and source monitoring ability. Journal of Experimental Child Psychology, 77, 169-196. Bugental, D. B., Blue, J., Cortez, V., Fleck, K., & Rodriguez, A (1992). Influences of witnessed affect on information processing in children. Child Development, 63, 774-786. Burgwyn-Bailes, E., Baker-Ward, L., Gordon, B. N., & Ornstein, P. A (2001). Children's memory for emergency medical treatment after one year: The impact of individual difference variables on recall and suggestibility. Applied Cognitive Psychology, 15, S25-S48. Butler, S., Gross, J., &Hayne, H. (1995). The effect of drawing on memory performance in young children. Developmental Psychology, 31,597-608. Cassel, W., & Bjorklund, D. (1995). Developmental patterns of eyewitness memory and suggestibility: An ecologically based short-term longitudinal study. Law & Human Behavior, 19, 507-532. Cassel, W. S., Roebers, C. E. M., & Bjorklund, D. F. (1996). Developmental patterns of eyewitness responses to repeated and increasingly suggestive questions. Journal of Experimental Child Psychology, 61, 116-133. Ceci, S., Toglia, M., & Ross, D. (1988). On remembering ... more or less: A trace strength interpretation of developmental differences in suggestibility. Journal of Experimental Psychology: General, 117, 201-203. Cederborg, C-A., Lamb, M. E., & Laurell, 0. (in press). Delay of disclosure, minimization, and denial when the evidence is unambiguous: A multivictim case. In M. Pipe, M. E. Lamb, Y. Orbach, & A-C. Cederborg (Eds.), Child sexual abuse: Disclosure, delay and denial. Mahwah, NJ: Lawrence Erlbaum Associates. Chase, W. G., & Simon, H. A (1973). Perception in chess. Cognitive Psychology, 4, 55-81. Chi, M. T. H. (1978). Knowledge structures and memory development. In R. S. Siegler (Ed.), Children's thinking: What develops? (pp. 73-96). New York: Wiley. Chi, M. T. H., & Ceci, S. J. (1987). Content knowledge: Its role, representation, and restructuring in memory development. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 20, pp. 91-142). New York: Academic Press. Chi, M., & Koeske, R. (1983). Network representation of a child's dinosaur knowledge. Developmental Psychology, 19, 29-39. Clubb, P. A., Nida, R. E., Merritt, K., & Ornstein, P. A (1993). Visiting the doctor: Children's knowledge and memory. Cognitive Development, 8, 361-372. Connolly, D. A., & Lindsay, D. S. (2001). The influence of suggestions on children's reports of a unique experience versus an instance of a repeated experience. Applied Cognitive Psychology, 15, 205-223.

16.

DEVELOPMENT OF EVENT MEMORY

471

Cordon, I. M., Pipe, M.-E., Sayfan, L., Melinder, A., & Goodman, G. S. (2004). Memory for traumatic experiences in early childhood. Developmental Review, 24, 101-132. Cowan, N., & Davidson, G. (1984). Salient childhood memories. Journal of Genetic Psychology, 145, 101-107. DeMarie-Dreblow, D. (1991). Relation between knowledge and memory: A reminder that correlation does not imply causality. Child Development, 62, 484-498. Drummey, A. B., & Newcombe, N. S. (1995). Remembering versus knowing the past: Children's explicit and implicit memories for pictures. Journal of Experimental Child Psychology. Special Issue: Early memory, 59, 549-565. Dunning, D., & Stern, L. B. (1992). Examining the generality of eyewitness hypermnesia: A close look at time delay and question type. Applied Cognitive Psychology, 6, 643-657. Elischberger, H. B., & Roebers, C. M. (2001). Improving young children's free narratives about an observed event: The effects of nonspecific verbal prompts. International Journal of Behavioral Development, 25, 160-166. Everson, M.D., & Boat, B. W. (2002). The utility of anatomical dolls and drawing in child forensic interviews. In M. L. Eisen, J. A. Quas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 383-408). Mahwah, NJ: Lawrence Erlbaum Associates. Farrar, M. J., & Goodman, G. S. (1990). Developmental differences in the relation between scripts and episodic memory: Do they exist? In R. Fivush &J. A. Hudson (Eds.), Emory symposia in cognition: Vol. 3. Knowing and remembering in young children (pp. 30-64). New York: Cambridge University Press. Farrar, M. J., & Goodman, G. S. (1992). Developmental changes in event memory. Child Development, 63, 173-187. Fasig, L. (1999, April). The influence of script memory on children's suggestibility. In M. K. WelchRoss & L. Fasig (Chairs), Moderators of suggestibility and eyewitness memory in young children. Symposium conducted at the biennial meeting of the Society of Research in Child Development, Albuquerque, NM. Fivush, R. (1998). Gendered narratives: Elaboration, structure, and emotion in parent-child reminiscing across the preschool years. In C. P. Thompson, D. J. Herrmann, D. Bruce, J. D. Read, D. G. Payne, & M.P. Toglia (Eds.), Autobiographical memory: Theoretical and applied perspectives (pp. 79-103). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., Haden, C., & Adam, S. (1995). Structure and coherence of preschoolers' personal narratives over time: Implications for childhood amnesia. Journal of Experimental Child Psychology. Special Issue: Early memory, 60, 32-56. Fivush, R., & Hamond, N. R. (1989). Time and again: Effects of repetition and retention interval on 2 year olds' event recall. Journal of Experimental Child Psychology, 47, 259-273. Fivush, R., & Hamond, N. R. (1990). Autobiographical memory across the preschool years: Toward reconceptualizing childhood amnesia. In R. Fivush & J. A. Hudson (Eds.), Emory symposia in cognition: Vol. 3. Knowing and remembering in young children (pp. 223-248). New York: Cambridge University Press. Fivush, R., Hazzard, A., Sales, J. M., Sarfati, D., & Brown, T. (2003). Creating coherence out of chaos? Children's narratives of emotionally positive and negative events. Applied Cognitive Psychology, 17, 1-19. Fivush, R., Hudson, J., &Nelson, K. (1984). Children's long-term memory for a novel event: An exploratory study. Merrill-Palmer Quarterly, 30,303-316. Fivush, R., Kuebli, J., & Clubb, P. A. (1992). The structure of events and event representations: A developmental analysis. Child Development, 63, 188-201. Fivush, R., Sales, J. M., Goldberg, A., Bahrick, L., & Parker, J. (2004). Weathering the storm: Children's long-term recall of Hurricane Andrew. Memory, 12, 104-118. Fivush, R., & Schwarzmueller, A. (1998). Children remember childhood: Implications for childhood amnesia. Applied Cognitive Psychology, 12, 455-473.

472

PIPE, THIERRY, LAMB

Foley, M.A., &Johnson, M. K. (1985). Confusions between memories for performed and imagined actions: A developmental comparison. Child Development, 56, 1145-1155. Geddie, L., Fradin, S., & Beer, J. (2000). Child characteristics which impact accuracy of recall and suggestibility in preschoolers: Is age the best predictor? Child Abuse & Neglect, 24, 223-235. Gee, S., & Pipe, M.-E. (1995). Helping children to remember: The influence of object cues on children's accounts of a real event. Developmental Psychology, 31, 746-7 58. Ghetti, S., & Alexander, K. W. (2004). "If it happened, I would remember it": Strategic use of event memorability in the rejection of false autobiographical events. Child Development, 75, 542-561. Gobbo, C., Mega, C., & Pipe, M.-E. (2002). Does the nature of the experience influence children's suggestibility? A study of children's event memory. Journal of Experimental Child Psychology, 81, 502-530. Goodman, G. S., Batterman-Faunce, J. M., Schaaf, J. M., & Kenney, R. (2002). Nearly 4 years after an event: Children's eyewitness memory and adults' perceptions of children's accuracy. Child Abuse & Neglect, 26, 849-884. Goodman, G. S., Hirschman, J. E., Hepps, D., & Rudy, L. (1991). Children's memory for stressful events. Merrill-Palmer Quarterly, 37, 109-157. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesburger, M. M., & Kuhn, J. (1994). Predictors of accurate and inaccurate memories of traumatic events experienced in childhood. Consciousness & Cognition: An Intemational]oumal. Special Issue: The recovered memory/ false memory debate, 3, 269-294. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1997). Children's reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge, and parental attachment. Applied Developmental Science, 1, 54-7 5. Gopnik, A., & Graf, P. (1988). Knowing how you know: Young children's ability to identify and remember the sources of their beliefs. Child Development, 59, 1366-1371. Gordon, B. N., Ornstein, P. A., Nida, R. E., Follmer, A., Crenshaw, M. C., & Albert, G. (1993). Does the use of dolls facilitate children's memory of visits to the doctor? Applied Cognitive Psychology, 7, 459-474. Greenhoot, A. F. (2000). Remembering and understanding: The effects of changes in underlying knowledge on children's recollections. Child Development, 71, 1309-1328. Greenhoot, A. F., Ornstein, P. A., Gordon, B. N., & Baker-Ward, L. (1999). Acting out the details of a pediatric check-up: The impact of interview condition and behavioral style on children's memory reports. Child Development, 70, 363-380. Greenstock, J., & Pipe, M.-E., (1996). Interviewing children about past events: The influence of peer support and misleading questions. Child Abuse & Neglect, 20, 69-80. Gross, J., & Hayne, H. (1998). Drawing facilitates children's verbal reports of emotionally laden events. ]oumal of Experimental Psychology: Applied, 4, 163-179. Gross, J., & Hayne, H. (1999). Drawing facilitates children's verbal reports after long delays. ]oumal of Experimental Psychology: Applied, 5, 265-283. Haden, C. A., & Fivush, R. (1996). Contextual variation in maternal conversational styles. MerrillPalmer Quarterly, 42, 200-227. Haden, C., Ornstein, P., Eckerman, C., & Didow, S. (2001). Mother-child conversational interactions as events unfold: Linkages to subsequent remembering. Child Development, 72, 1016-1031. Hamond, N. R., & Fivush, R. (1991). Memories of Mickey Mouse: Young children recount their trip to Disney World. Cognitive Development, 6, 433-448. Hayne, H., & Rovee-Collier, C. (1995). The organization of reactivated memory in infancy. Child Development, 66,893-906. Henkel, L.A. (2004). Erroneous memories arising from repeated attempts to remember. ]oumal of Memory and Language, 50, 26-46.

16.

DEVELOPMENT OF EVENT MEMORY

473

Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., & Horowitz, D. (2001). The effects of mental context reinstatement on children's accounts of sexual abuse. Applied Cognitive Psychology, 15, 235-248. Hershkowitz, I., Orbach, Y., Lamb, M. E., Sternberg, K. J., Horowitz, D., & Hovav, M. (1998). Visiting the scene of the crime: Effects on children's recall of alleged abuse. Legal & Criminological Psychology, 3, 195-207. Hildreth, K., Sweeny, B., & Rovee-Collier, C. (2003). Differential memory-preserving effects of reminders at 6 months. Journal of Experimental Child Psychology, 84, 41-62. Howe, M. L. (1997). Children's memory for traumatic experiences. Learning & Individual Differences, 9, 153-174. Howe, M. L., & Courage, M. L. (1993). On resolving the enigma of infantile amnesia. Psychological Bulletin, 113, 305-326. Howe, M. L., & Courage, M. L. (2004). Demystifying the beginnings of memory. Developmental Review, 24, 1-5. Howe, M. L., Courage, M. L., & Edison, S.C. (2003). When autobiographical memory begins. Developmental Review, 23, 471-494. Howe, M. L., Courage, M. L., & Peterson, C. (1994). How can I remember when "I" wasn't there? Long-term retention of traumatic memories and emergence of the cognitive self. Consciousness and Cognition, 3, 32-355. Howe, M. L., Courage, M. L., & Peterson, C. (1995). Intrusions in preschoolers' recall of traumatic childhood events. Psychonomic Bulletin & Review, 2, 130-134. Hudson, J. A. (1990a). Constructive processing in children's event memory. Developmental Psychology, 26, 180-187. Hudson, J. A. (1990b). The emergence of autobiographical memory in mother-child conversation. In R. Fivush &J. A. Hudson (Eds.), Knowingandrememberinginyoungchildren (pp. 166196). New York: Cambridge University Press. Hudson, J. A., & Fivush, R. (1991). As time goes by: Sixth graders remember a kindergarten experience. Applied Cognitive Psychology, 5, 347-360. Hudson, J. A., Fivush, R., & Kuebli, J. (1992). Scripts and episodes: The development of event memory Special Issue: Memory in everyday settings. Applied Cognitive Psychology, 6, 483-505. Hudson, J., & Nelson, K. (1986). Repeated encounters of a similar kind: Effects of familiarity on children's autobiographic memory. Cognitive Development, I, 253-2 71. Johnson, M., Hashtroudi, S., & Lindsay, D. (1993). Source monitoring. Psychological Bulletin, 114,3-28. Jones, C., & Pipe, M.-E. (2002). How quickly do children forget events? A systematic study of children's event reports as a function of delay. Applied Cognitive Psychology, 16, 755-768. Kleinknecht, E. (2001). Emerging autobiographies: The role of social-cognition in the development of event memory skills. Dissertation Abstracts International: Section B: the Sciences & Engineering, 61, 5019. Lamb, M. E., Hershkowitz, I., Sternberg, K. J., Boat, B., & Everson, M. D. (1996). Investigative interviews of alleged sexual abuse victims with and without anatomical dolls. Child Abuse and Neglect, 20, 1239-1247. Lamb, M. E., Sternberg, K. J., & Esplin, P. W. (2000). Effects of age and delay on the amount of information provided by alleged sex abuse victims in investigative interviews. Child Development, 71, 1586-1596. Lamb, M. E., Sternberg, K. J., Orbach, Y., Esplin, P. W., Stewart, H., & Mitchell, S. (2003). Age differences in young children's responses to open-ended invitations in the course of forensic interviews. Journal of Consulting and Clinical Psychology, 71, 926-934. Landis, T. Y. (1982). Interaction between text and prior knowledge in children's memory for prose. Child Development, 53, 811-814. La Rooy, D. J., Pipe, M.-E., & Murray, J. E. (2005). Reminiscence and hypermnesia in children's eyewitness memory. Journal of Experimental Child Psychology, 90, 235-254.

474

PIPE, THIERRY, LAMB

Leichtman, M.D., & Ceci, S. J. (1995). The effects of stereotypes and suggestions on preschoolers' reports. Developmental Psychology, 31 , 568-5 78. Leichtman, M. D., Morse, M. B., Dixon, A., & Spiegel, R. (2000). Source monitoring and suggestibility: An individual differences approach. InK. P. Roberts & M. Blades (Eds.), Children's source monitoring (pp. 257-288). Mahwah, NJ: Lawrence Erlbaum Associates. Lindberg, M.A. (1980). Is knowledge base development a necessary and sufficient condition for memory development? Journal of Experimental Child Psychology, 30, 401-410. Lindsay, D. S., & Johnson, M. K. (1989). The eyewitness suggestibility effect and memory for source. Memory & Cognition, 17, 349-358. Lindsay, D. S., Johnson, M., &Kwon, P. (1991). Developmental changes in memory source monitoring. Journal of Experimental Child Psychology, 52, 297-318. London, K., Bruck, M., Ceci, S. J., & Shuman, D. W. (in press). Disclosure of child sexual abuse: A review of the contemporary empiticalliterature. In M. Pipe, M. E. Lamb, Y. Orbach, & A.-C. Cederborg (Eds.), Child sexual abuse: Disclosure, delay and denial. Mahwah, NJ: Lawrence Erlbaum Associates. London, K., Bruck, M., Ceci, S. J., & Shuman, D. W. (2005). Disclosure of child sexual abuse: What does the research tell us about the ways that children tell? Psychology, Public Policy, and the Law, 11, 194-226. Mandler, J. M., & Stein, N. L. (1974). Recall and recognition of pictures by children as a function of organization and dis tractor similarity. Journal of Experimental Psychology, 102, 65 7-669. Martin, M. (1988). Individual differences in everyday memory. In M. M. Gruneberg, P. E. Morris, & R.N. Sykes (Eds.), Practical aspects of memory: Current research and issues: Vol. 1. Memory in everyday life (pp. 466-471). Oxford, England: John Wiley & Sons. McGuigan, F., & Salmon, K. (2004). Time to talk: The influence of the timing of adult-child talk on children's event memory. Child Development, 75, 669-686. McNichol, S., Shute, R., & Tucker, A. (1999). Children's eyewitness memory for a repeated event. Child Abuse & Neglect, 23, 1127-1139. Merritt, K. A., Ornstein, P. A., & Spieker, B. (1994). Children's memory for a salient medical procedure: Implications for testimony. Pediatrics, 94, 17-23. Murachver, T., Pipe, M.-E., Gordon, R., Fivush, R., & Owens, J. L. (1996). Do, show, and tell: Children's event memories acquired through direct experience, observation, and stories. Child Development, 67, 3029-3044. Myers, N. A., Perris, E. E., & Speaker, C. J. (1994). Fifty months of memory: A longitudinal study in early childhood. In R. Fivush (Ed.), Long-term retention of infant memories. Memory (pp. 383415). Hove, England: Lawrence Erlbaum Associates. Myles-Worsley, M., Cromer, C., & Dodd, D. (1986). Children's preschool script reconstruction: Reliance on general knowledge as memory fades. Developmental Psychology, 22, 2-30. Neisser, U. (2004). Memory development: New questions and old. Developmental Review, 24, 154-158. Nelson, K. (1986). Event knowledge: A functional approach to cognitive development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1989). Remembering: A functional developmental perspective. In P. R. Solomon, G. R. Goethals, C. M. Kelley, & B. R. Stephens (Eds.), Memory: Interdisciplinary approaches (pp. 127150). New York: Springer-Verlag. Nelson, K. (1993). The psychological and social origins of autobiographical memory. Psychological Science, 4, 85-92. Nelson, K. (1996). Memory development from 4 to 7 years. The five to seven year shift: The age of reason and responsibility. In A. J. Sameroff & M. M. Haith (Eds.), The John D and Catherine T MacArthur Foundation series on mental heath and development (pp. 141-160). Chicago, IL: University of Chicago Press. Nelson, K., & Fivush, R. (2004). The emergence of autobiographical memory: A social cultural developmental model. Psychological Review, 111, 486-511.

16.

DEVELOPMENT OF EVENT MEMORY

475

Nelson, K., & Gruendel, J. (1981). Generalized event representations: Basic building blocks of cognitive development. In M. E. Lamb & A. L. Brown (Eds.), Advances in development psychology (Vol. 1, pp. 131-158). Hillsdale, NJ: Lawrence Erlbaum Associates. Orbach, Y., Hershkowitz, I., Lamb, M. E., Sternberg, K.J., Esplin, P. W., &Horowitz, D. (2000). Assessing the value of structured protocols for forensic interviews of alleged abuse victims. Child Abuse and Neglect, 24, 733-752. Orbach, Y., Hershkowitz, I., Lamb, M. E., Sternberg, K. ]., & Horowitz, D. (2000). Interviewing at the scene of the crime: Effects on children's recall of alleged abuse. Legal and Criminological Psychology, 5, 135-147. Ornstein, P. A. (1995). Children's long-term retention of salient personal experiences. Journal of Traumatic Stress, 8, 581-605. Ornstein, P. A., Baker-Ward, L., Gordon, B. N., &Merritt, K. A. (1997). Children's memory for medical experiences: Implications for testimony. Applied Cognitive Psychology, 11, S87-S 104. Ornstein, P. A., Merritt, K. A., Baker-Ward, L., Furtado, E., Gordon, B. N., & Principe, G. (1998). Children's knowledge, expectation, and long-term retention. Applied Cognitive Psychology, 12, 387--405. Ornstein, P. A., Principe, G., Hudson, A. E., Gordon, B. N., & Merritt, K. (1997, April). Procedural information and stress as mediators of children's long-term recall. Paper presented at the meeting of the Society for Research in Child Development, Albequerque, NM. Ornstein, P. A., Shapiro, L. R., Clubb, P. A., Follmer, A., & Baker-Ward, L. (1997). The influence of prior knowledge on children's memory for salient medical experiences. InN. Stein, P. A. Ornstein, B. Tversky, & C.]. Brainerd (Eds.), Memory for everyday and emotional events (pp. 83-112). Hillsdale, NJ: Lawrence Erlbaum Associates. Parker, J. F., Bahrick, L., Lundy, B., Fivush, R., & Levitt, M. (1998). Effects of stress on children's memory for a natural disaster. In C. P. Thompson, D.]. Herrmann, J.D. Read, D. Bruce, D. G. Payne, & M.P. Toglia (Eds.), Eyewitness memory: Theoretical and applied perspectives (pp. 3154). Mahwah, NJ: Lawrence Erlbaum Associates. Perner, J. (2000). Memory and theory of mind. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 297-312). London, Oxford: Oxford University Press. Perner, J., & Ruffman, T. (1995). Episodic memory an autonoetic consciousness: Developmental evidence and a theory of childhood amnesia. Journal of Experimental Child Psychology. Special Issue: Early memory, 59,516-548. Peters, D.P. (1987). The impact of naturally occurring stress on children's memory. InS.]. Ceci, M.P. Toglia, & D. F. Ross (Eds.), Children's eyewitness testimony (pp. 122-141). New York: Springer-Verlag. Peters, D.P. (1991). The influence of stress and arousal on the child witness. In J. Doris (Ed.), The suggestibility of children's recollections (pp. 60-76). Washington, DC: American Psychological Association. Peterson, C. (1999). Children's memory for medical emergencies: 2 years later. Developmental Psychology, 35, 1493-1506. Peterson, C., & Bell, M. (1996). Children's memory for traumatic injury. Child Development, 67, 3045-3070. Peterson, C., Moores, L., & White, G. (2001). Recounting the same events again and again: Children's consistency across multiple interviews. Applied Cognitive Psychology, 15, 353-371. Peterson, C., & Rideout, R. (1997). Memory for medical emergencies experienced by 1- and 2-year-olds. Developmental Psychology, 34, 1059-1072. Peterson, C., & Whalen, N., (2001). Five years later: Children's memory for medical emergencies. Applied Cognitive Psychology, 15, S7-S24. Pillemer, D. B. (1993). Preschool children's memories of personal circumstances: The fire alarm study. In E. Winograd & U. Neisser (Eds.), Emory symposia in cognition: Vol. 4. Affect and accuracy in recall: Studies of"flashbulb" memories. (pp. 121-137). New York: Cambridge University Press.

476

PIPE, THIERRY, LAMB

Pillemer, D. B., Picariello, M. L., & Pruett, J. C. (1994). Very long-term memories of a salient preschool event. Applied Cognitive Psychology, 8, 95-106. Pillemer, D. B., & White, S. H. (1989). Childhood events recalled by children and adults. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 21, pp. 297-340). New York: Academic Press. Pipe, M.-E., Lamb, M. E., Orbach, Y., & Esplin, P. W. (2004). Recent research on children's testimony about experienced and witnessed events. Developmental Review, 24, 440-468. Pipe, M.-E., & Salmon, K. (2002). What children bring to the interview context: Individual differences in children's event reports. In M. L. Eisen (Ed.), Memory and suggestibility in the forensic interview. Personality and clinical psychology series (pp. 235-261). Mahwah, NJ: Lawrence Erlbaum Associates. Pipe, M.-E., & Wilson, C. (1994). Cues and secrets: Influences on children's event reports. Developmental Psychology, 30, 515-525. Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington, DC: American Psychological Association. Poole, D. A., & Lindsay, D. S. (2002). Reducing child witnesses' false reports of misinformation from parents. Journal of Experimental Child Psychology, 8I, 117-140. Powell, M. B., Roberts, K. P., Ceci, S. ]., & Hembrooke, H. H. (1999). The effects of repeated experience on children's suggestibility. Developmental Psychology, 35, 1462-1477. Powell, M. B., Roberts, K. P., & Thomson, D. M. (2000). The effect of a suggestive interview on children's memory of a repeated event: Does it matter whether suggestions are linked to a particular incident? Psychiatry, Psychology, & Law, 7, 182-191. Powell, M. B., & Thomson, D. M. (1997). Contrasting memory for temporal-source and memory for content in children's discrimination of repeated events. Applied Cognitive Psychology, II, 339-360. Powell, M. B., & Thomson, D. M. (2003). Improving children's recall of an occurrence of arepeated event: It is a matter of helping them to generate options? Law Human Behavior, 27, 365-384. Pratt, C., & Bryant, P. (1990). Young children understand that looking leads to knowing (so long as they are looking into a single barrel). Child Development, 6I, 97 3-982. Price, D. W. W., & Goodman, G. S. (1990). Visiting the wizard: Children's memory of a recurring event. Child Development, 6I, 664-680. Priestley, G., Roberts, S., & Pipe, M.-E. (1999). Returning to the scene: Reminders and context reinstatement enhance children's recall. Developmental Psychology, 35, 1006-1019. Quas, J. A. (1998). Children's memory of experienced and nonexperienced events across repeated interviews. Unpublished doctoral dissertation, University of California, Davis. Quas, J. A., Qin, J. ]., Schaaf, J., & Goodman, G. S. (1997). Individual differences in children's and adult's suggestibility and false event memory. Learning and Individual Differences, 9, 359-390. Quas, J. A., Schaaf, J. M., Alexander, K. W., & Goodman, G. S. (2000). Do you really remember it happening or do you only remember being asked about it happening? Children's source monitoring in forensic contexts. InK. P. Roberts & M. Blades (Eds), Children's source monitoring. (pp. 197-226). Mahwah, NJ: Lawrence Erlbaum Associates. Ratner, H. H. (1984). Memory demands and the development of young children's memory. Child Development, 55, 2173-2191. Reese, E. (2002). Social factors in the development of autobiographical memory: The state of the art. Social Development, II, 124-142. Reese, E., Haden, C. A., & Fivush, R. (1993). Mother-child conversations about the past: Relationships of style and memory over time. Cognitive Development, 8, 403-430. Roberts, K. P. (2002). Children's ability to distinguish between memories from multiple sources: Implications for the quality and accuracy of eyewitness statements. Developmental Review, 22, 403-435.

16.

DEVELOPMENT OF EVENT MEMORY

477

Roberts, K. P., & Powell, M. B. (2001). Describing individual incidents of sexual abuse: A review of research on the effects of multiple sources of information on children's reports. Child Abuse & Neglect, 25, 1643-1659. Roebers, C. M., & Schneider, W. (2001). Individual differences in children's eyewitness recall: The influence of intelligence and shyness. Applied Developmental Science, 5, 9-30. Roediger, H. L., McDermott, K. B., Pisani, D. B., & Gallo, D. A. (2004). Illusory recollection of voices. Memory, 12, 586-602. Rovee-Collier, C., Hartshorn, K., & DiRubbo, M. (1999). Long-term maintenance of infant memory. Developmental Psychology, 35, 91-102. Rovee-Collier, C., & Hayne, H. (2000). Memory in infancy and early childhood. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 267-282). London: Oxford University Press. Rudy, L., & Goodman, G. S. (1991). Effects of participation on children's reports: Implications for children's testimony. Developmental Psychology, 27, 527-538. Salmon, K., & Pipe, M.-E. (2000). Recalling an event one year later: The impact of props, drawing and a prior interview. Applied Cognitive Psychology, 14, 99-120. Salmon, K., Price, M., & Pereira, J. K. (2002). Factors associated with young children's long-term recall of an invasive medical procedure: A preliminary investigation. Journal of Developmental & Behavioral Pediatrics, 23,347-352. Salmon, K., Roncolato, W., & Gleitzman, M. (2003). Children's reports of emotionally laden events: adapting the interview to the child. Applied Cognitive Psychology, 17, 65-79. Saywitz, K. J., Goodman, G. S., Nicholas, E., & Moan, S. F. (1991). Children's memories of a physical examination involving genital touch: Implications for reports of child sexual abuse. Journal of Consulting & Clinical Psychology, 59, 682-691. Saywitz, K. J., & Snyder, L. (1996). Narrative elaboration: Test of a new procedure for interviewing children. Journal of Consulting & Clinical Psychology, 64, 1347-1357. Schneider, W., & Pressley, M. (1997). Memory development between two and twenty (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Scrivner, E., & Safer, M.A. (1988). Eyewitnesses show hypermenesia for details about a violent event. Journal of Applied Psychology, 73,371-377. Sheffield, E. G. (2004). But I thought it was Mickey Mouse: The effects of new postevent information on 18-month-olds' memory. Journal of Experimental Child Psychology, 87, 221-238. Simcock, G., & Hayne, H. (2002). Breaking the barrier? Children fail to translate their preverbal memories into language. Psychological Science, 13, 225-231. Slackman, E., & Nelson, K. (1984). Acquisition of an unfamiliar script in story form by young children. Child Development, 55,329-340. Sternberg, K. J., Lamb, M. E., Orbach, Y., Esplin, P. W., & Mitchell, S. (2001). Use of a structured investigative protocol enhances young children's responses to free-recall prompts in the course of forensic interviews. Journal of Applied Psychology, 86, 997-1005. Sutherland, R., Pipe, M.-E., Schick, K., Murray, J., & Gobbo, C. (2003). Knowing in advance: The impact of prior event information on memory and event knowledge. Journal of Experimental Child Psychology, 84, 244-263. Terr, L. C. (1988). What happens to early memories of trauma? A study of twenty children under age five at the time of documented traumatic events. Journal of the American Academy of Child and Adolescent Psychiatry, 27, 96-104. Tessler, M., & Nelson, K. (1994). Making memories: The influence of joint encoding on later recall by young children. Consciousness & Cognition: an International Journal, 3, 307-326. Thierry, K. L., Goh, C. L., Murray, J., & Pipe, M.-E. (2005). Source recall enhances children's discrimination of seen and heard events. Journal of Experimental Psychology: Applied, 11, 33-44. Thierry, K. L., Lamb, M. E., & Orbach, Y. (2003). Awareness of the origin of knowledge predicts child witnesses' recall of alleged sexual and physical abuse. Applied Cognitive Psychology, 17, 953-967.

478

PIPE, THIERRY, LAMB

Thierry, K. L., Lamb, M. E., Orbach, Y., & Pipe, M.-E. (2005). Developmental differences in the function and use of anatomical dolls during interviews with alleged sexual abuse victims. Journal of Consulting and Clinical Psychology, 73, 1125-1134. Thierry, K. L., & Spence, M. ]. (2002). Source-monitoring training facilitates preschoolers' eyewitness memory performance. Developmental Psychology, 38, 428-437. Thierry, K. L., Spence, M. ]., & Memon, A. (2001). Before misinformation is encountered: Source monitoring decreases child witness suggestibility. Journal of Cognition & Development, 2, 1-26. Tobey, A. E., & Goodman, G. S. (1992). Children's eyewitness memory: Effects of participation and forensic context. Child Abuse & Neglect, 16, 779-796. Tulving, E. (1983). Elements of episodic memory. Oxford: Clarendon Press. Warren, A. R., & Swartwood,]. N. (1992). Developmental issues in flashbulb memory research: Children recall the Challenger event. In E. Winograd & U. Neisser (Eds.), Emory symposia in cognition: Vol. 4. Affect and accuracy in recall: Studies of "flashbulb" memories. (pp. 95-120). New York: Cambridge University Press. Welch-Ross, M. K. (1995). An integrative model of the development of autobiographical memory. Developmental Review, 15, 338-365. Wesson, M., & Salmon, K. (2001). Drawing and showing: Helping children to report emotionally laden events. Applied Cognitive Psychology, 15, 301-320. Zaragoza, M. S., & Lane, S. M. (1994). Source misattributions and the suggestibility of eyewitness memory.Journal of Experimental Psychology: Learning, Memory, & Cognition, 20, 934-945.

17 False Memory in Children: Data, Theory, and Legal Implications Valerie F. Reyna Britain Mills Steven Estrada Charles J. Brainerd Cornell University

False memory is an area of intense research interest for both theoretical and practical reasons. From a theoretical perspective, false memories have been the subject of hot debates about the nature of human memory and a focal point for old and new memory theories. From a practical perspective, false memories are a threat to the validity of eyewitness testimony, a misleading source of autobiographical information in psychotherapy, and a biased representation of lessons taught in educational settings. True memories represent events as they really happened, whereas false memories shade, distort, or entirely misrepresent what really happened. False memories may sometimes seem to be only subtly different from strictly accurate reports of experienced events, as when witnesses misremember true inferences about events as though the inferences were directly experienced. Such false memories, which are often confidently reported as directly witnessed, are ubiquitous in everyday life. However, the distinction between inference and experience is crucial in many real-world contexts, especially in legal settings (Reyna & Brainerd, 1998). Eyewitnesses are not called upon to report their theories of events, but, rather, what they actually remember experiencing. Although some inferences turn out to be consistent with the facts, others are based on prejudice, ignorance, and unfounded suppositions (Nisbett & Ross, 1980). Thus, justice requires that we minimize the influence of false memories in legal proceedings at every stage in which children are interviewed, from the crime scene to the courtroom In this chapter we review evidence concerning false memories of multiple types in children, whose perceived competence as witnesses has shifted dramatically in the last 20 years. The general question that concerns us is whether children's memory is essentially 479

480

REYNAET AL.

accurate (and under what conditions), and whether their memory is more or less accurate compared to that of adults. To these ends, we have divided the published literature on children's false memory into three sections: spontaneous false memory for meaningful materials (e.g., sentences, pictures, numbers, or stories), false memory in the DeeseRoediger-McDermott word-list paradigm, and resistance to and rejection of false memories in a variety of paradigms. Within each section, we review the major findings, discuss explanations for those findings (including schema, activation, source-monitoring, and fuzzy-trace theories), and, in a final section, draw out their social and legal policy implications. As an advance organizer, it is useful to describe the content and approach of this chapter. We review the latest research concerning children's false memory, mainly focusing on laboratory tasks because those tasks cleanly test hypotheses about which factors produce and reduce false memories. For those readers who are most interested in practical applications, the last section provides an overview of social and legal implications of false-memory research. However, understanding the evidentiary basis for recommended practice, discussed in the initial sections of this chapter, is essential in court. Ironically, basic research using laboratory tasks is often more applicable to real-life legal cases than research that mimics the complexities of real life (see Brainerd & Reyna, 2005; Brainerd, Reyna, & Poole, 2000; Reyna, Holliday, & Marche, 2002, for details). The reason for this difference in trustworthiness is that well-designed laboratory tasks isolate causes for human behavior (rather than jumbling up multiple possible causes and effects, as happens in real life). Results of sound scientific studies, especially those that have been replicated, provide evidence for general causal principles that predict behavior in real life. For example, asking child witnesses to generate inferences about what might have happened in order to draw them out is likely to produce not only false testimony but also false memories-confident beliefs that whatever was produced actually happened. In this chapter we explain how experiments conducted under controlled conditions show that repeatedly encouraging inferences (with the use of a variety of methods) leads to later confident assertions that those inferences were experienced. Thus, readers who are interested primarily in practical applications of research should think about laboratory experiments as analogous to events in real life and try to extract the general principles or rules of thumb that experiments have demonstrated to be true about human behavior. Specifically, they should think of the words, sentences, pictures, and other presented material in experiments as though they were experienced events in the real world and think of memory tests as analogous to investigative interviews. Are there examples of failure of laboratory tasks to generalize to the real world? Indeed, there are such examples, but they are fewer than literal thinkers would have us believe, and, more generally, no single experiment or real-life event for that matter is completely analogous in every respect to another event. Professionals need to think carefully about the underlying causal mechanisms for behavior and about whether those mechanisms apply to the real-life situation, in order to judge whether the analogy holds. Assumptions about what can and cannot be generalized from laboratory tasks ultimately depend on one's theory of a phenomenon-in this case, of false memory. In that

17. FALSEMEMORY

481

connection, readers should evaluate the following theoretical explanations for false memories in light of the evidence presented below: Are false memories caused by con~ structing, or making up, memories based on prior knowledge and inference; by activat~ ing past associations between the experienced event and related events; by confusing the sources of memories (i.e., whether they were thought up or experienced); or by using mechanisms such as semantic activation and inference to store gist memories of experi~ enced events in parallel with and independent of verbatim memories for the same events? Although these theories of false memory all seem plausible and perhaps indistinguish~ able, as we discuss, they make strikingly different predictions that have been tested, and the theories have not been uniformly supported by the results of those tests.

SPONTANEOUS FALSE MEMORY FOR MEANINGFUL MATERIALS

In classic sentence memory experiments, Bransford and Franks (1971) showed that adults would falsely recognize sentences that were semantically consistent with presented sen~ tences (and that integrated ideas presented in separate sentences), often with high con~ fidence that those sentences had been presented. For example, given the sentences The ants ate the sweet jelly and The sweet jelly was on the table (mixed with other sentences of varying lengths), subjects would confidently assert that they had been presented with the sentence The ants ate the sweet jelly that was on the table. Confidence was highest for test sentences that combined four ideas that had been expressed separately during the earlier study phase and was significantly higher compared with sentences that, in fact, had been previously presented. In line with earlier findings of schematic memory in adults (Bartlett, 1932), this finding was thought to support the conclusion that human memory is interpretive rather than strictly accurate. Numerous investigators set out to determine whether this dramatic demonstration of false recognition of sentences could be replicated in children. In one of the earliest demonstrations, Paris and Carter (1973) presented children with sentences expressing spatial relationships much like those that had been given to adults. For example, a child might be told that The bird is in the cage and The cage is under the table, and given a filler sentence to complete a short narrative. Like adults, children also falsely recognized true inferences, such as The bird is under the table, as having been presented earlier. Other studies (e.g., Johnson & Scholnick, 1979; Liben & Posnansky, 1977; Paris & Mahoney, 1974; Prawatt & Cancelli, 1976; Small & Butterworth, 1981) produced similar findings: test sentences that had not been previously presented during the study phase, but which preserved the meaning of presented sentences, were mis~ recognized more often than sentences that did not preserve meaning. Meaningfully re~ lated pictures (Brainerd & Reyna, 1993; Paris & Mahoney, 1974) and numbers (Brainerd & Gordon, 1994) also elicited false recognition of nonpresented materials that were con~ sistent with the patterns of presented materials. For example, given stories that contained numerical information and relationships, such as Farmer Brown has 10 cows, 7 horses, 4 pigs, 3 sheep, and 2 chickens (some stories contained relational statements such as

482

REYNAET AL.

Farmer Smith has more ducks than turkeys), children falsely recognized the statement that Farmer Brown has more cows than horses as having been presented. Brown, Smiley, Day, Townsend, and Lawton (1977) obtained analogous effects for children's recall, namely, intrusions of thematic ideas in children's retention of stories. The traditional explanation for these kinds of false recognition and recall findings has been that memory is schematic or constructive, which implies that it is difficult to discriminate what has actually been experienced from what has been constructed in the mind based on knowledge, inference, and interpretation. A developmental corollary of the general notion that memory is constructive is the idea that false recognition and recall should increase with development. Predictions about development usually highlight increasing accuracy and improved performance with increasing age. However, if, as Piaget and lnhelder (1973, p. 382) and many of their cognitive developmental descendants claimed, "the schemata used by the memory are borrowed from the intelligence," then certain kinds of memory errors that are consistent with inferential reasoning should become more common with age. In other words, younger children would be less likely to misrecognize true inferences as having been presented, because they are less likely to make those inferences in the first place. This prediction of schema theory or constructivism was not borne out by the evidence, however. The evidence for the developmental prediction of greater false recognition of inferences with age or developmental level is mixed. For example, Paris and Carter (1973) and Liben and Posnansky (1977) did not find developmental differences according to age or developmental level, respectively, in types of false recognition. (In some experiments, younger children misrecognized true inferences more often than older children, but they also misrecognized other nonpresented sentences as well.) However, Johnson and Scholnick (1979) and Prawatt and Cancelli (1976) did find such differences (increases in false recognition of inferences) associated with cognitive level. In other studies, the opposite developmental trend was observed: developmental decreases in spontaneous false memories for inferences (Ackerman, 1992, 1994; Reyna & Kiernan, 1994, 1995) and for nonexperienced events that are broadly consistent with actually experienced events (Pipe, Gee, Wilson, & Egerton, 1999; Poole & White, 1991). Finally, some investigators reported different developmental trends in different conditions (Brown et al., 1977; Paris & Mahoney, 1974). Several investigators (Liben & Posnansky, 1977; Paris & Mahoney, 1974; Small and Butterworth, 1981) have pointed out that the conclusion that memory is schematic or constructive is subject to an important methodological confound. In most studies of verbal materials, true test sentences-presented sentences and true inferences-contained words that were presented, whereas false sentences contained new words that had not been presented. Thus, simple familiarity rather than consistency with meaning or inferences is sufficient to account for false recognition of true sentences. Reyna and Kiernan (1994) showed that when familiar wording and truth value of sentences are unconfounded (i.e., true sentences sometimes contain familiar words and other times contain new words, and similarly for false sentences), older children are less likely than younger ones to misrecognize true inferences, regardless of the familiarity of their wording. Older children demonstrated much better verbatim recognition of presented true sentences

17.

FALSE MEMORY

483

compared with true inferences or true paraphrases; 9-year olds were 50% more likely to say "old" to presented verbatim sentences versus their nearest true competitors (i.e., the most often misrecognized true nonpresented sentences), whereas 6-year-olds were only 24% more likely to say "old" to the presented sentences versus true nonpresented sentences. The differences between true inferences and false sentences were about the same across age. These results contrast with predictions of schema theory or constructivism, which suggest that memory errors that preserve meaning or inferential gist should be more prevalent among older children; instead, meaning-preserving false recognitions were less likely among older children (see also Reyna & Kiernan, 1995, for another study in which familiarity and meaning consistency were unconfounded, with similar results). Another prediction of schema theory and constructivism is that recognition judgments of presented sentences and of nonpresented inferences ought to be related. Memory judgments about presented sentences and inferences should both be affected either by common schemata (Piaget & lnhelder, 1973) or by their consistency with a semantic representation of presented information (Paris & Carter, 1973). For example, Paris and Carter argue explicitly that false recognition of inferences should be more likely when presented sentences (that support the inferences) are remembered. This prediction of positive dependency-that the tendency to say "old" to true inferences is positively related to the tendency to say "old" to presented sentences that support those specific inferences-was disconfirmed for both children (Reyna & Kiernan, 1994) and for adults (Lim, 1993) with the use of identical materials. Although recognition judgments of presented sentences were independent of judgments of meaning-preserving paraphrases and inferences, judgments of meaning-preserving paraphrases and inferences (i.e., nonpresented true sentences) were positively dependent on one another. Semantic integration occurred for nonpresented true sentences but was independent of memory for verbatim sentences. Similar false recognition and independence effects were found for metaphors and corresponding nonpresented sentences that expressed their interpretations (Reyna & Kiernan, 1995). Activation theories often apply to recognition of presented and semantically related words (Roediger, Watson, McDermott, & Gallo, 2001) because words are readily represented as fixed items in a mental lexicon (or dictionary), but activation theories of sentences (Anderson, Budiu, & Reder, 2001) have also been developed. Activation theories generally account for false recognition by assuming ancillary activation of semantic neighborhoods or semantically (or associatively) related links in a mental network. For example, if someone hears the word apple, it is as if the node for apple lights up in the brain, and some activation spreads to related words, such as banana. Because activation is shared along a network linking presented items with related nonpresented items, recognition judgments of presented and nonpresented items should clearly be dependent-but, as we have seen, they have been found to be independent. Activation theories would also be hard pressed to account for higher recognition rates for nonpresented related items than presented items (as has been found by Brainerd & Reyna, 1998), because activation dissipates from presented items to related nodes. Thus, it is difficult for activation to be higher for related items than for items that were actually presented, all other factors being equal (Reyna & Lloyd, 1997). If one assumes that activation from

484

REYNAET AL.

related nodes is summated in order to accommodate this effect, it then becomes difficult to account for the fact that people do not routinely mistake concepts related to experienced events as having been directly experienced. Why is remembering that you heard banana when you actually heard apple not the norm? The significance of this point is that any adequate theory has to simultaneously explain true and false memories. Source-monitoring theory makes a prediction about dependency that is similar to that made by activation theory Oohnson, Hashtroudi, & Lindsay, 1993). According to this theory, source judgments depend on qualities of the memory for the experienced event (e.g., the event of witnessing a crime or the event of hearing a list of sentences). Source-monitoring theory predicts positive dependency between presented and semantically related items (contrary to data) because both are related to memory for the experienced event. People decide that true inferences were externally presented rather than internally generated, according to this theory, because inferences share features with memory for the experienced event (i.e., shared semantic features), and people do not necessarily retrieve cues from their memory that would distinguish the source of the inference (e.g., evidence of internal cognitive operations, such as remembering thinking of the inference). Source-monitoring theorists have consistently predicted (based on developmental improvements in memory judgments and source discrimination) and often found source monitoring differences between children and adults (and between younger children and older children; for a review, see Roberts [2002]). As Reyna and Lloyd (1997) review in some detail, source-monitoring theory describes many false memory effects in children and adults, but has difficulty accounting for inconsistent developmental patterns (discussed earlier), findings of verbatim-gist independence, as well as other false-memory effects. Source-monitoring theory continues to be an excellent framework for characterization of situations in which memories for information from multiple sources remain intact, even though the sources of the memories are confused. The source-monitoring framework has also generated a special task (in which people judge the specific source of recognized items) that reduces false memories attributable to source confusions. A range of mechanisms from divergent theories, however, may be used to explain these source-confusion errors (see Brainerd & Reyna, 2005). The most recent developmental theory of false memory is fuzzy-trace theory, which was designed to build on findings from earlier theories as well as generate new predictions (e.g., Reyna & Brainerd, 1995, 1998; Reyna, Holliday, & Marche, 2002). According to fuzzy-trace theory, children extract separate verbatim and gist representations of experience (i.e., these representations are not semantically integrated) that become inaccessible at different rates over time. With development, older children become better able to acquire and retain verbatim information, compared with when they were younger. They also become better able to understand and spontaneously connect the semantic gist or meaning of related events. The experiments of Reyna and Kiernan (1994, 1995) were generated to test predictions of fuzzy-trace theory, and results were found to be supportive of the theory. Reyna and Brainerd (1995) outline a number of factors that predictably affect the accessibility of verbatim versus gist memories and, thus, the tendency to falsely recognize and recall gist-consistent materials, including words, sentences, pictures, and numbers. These factors include delay (verbatim memory becomes

17.

FALSEMEMORY

485

inaccessible more rapidly over time than memory for gist), age (verbatim and gist memory develop with age), materials (pictures and metaphors, for example, support more enduring verbatim memories compared with words and literal sentences, respectively), and instructions (instructions to use verbatim memory, as opposed to meaning instructions, reliably affect recognition and recall judgments). These assumptions account for the results we have reviewed thus far, as well as other findings, for instance, that judgments of presented sentences and true inferences become positively dependent after a retention interval (e.g., an interval of a week or more between presentation of sentences and memory testing). According to this theory, once verbatim memories for presented sentences are no longer accessible after a delay, judgments of both presented and true related sentences are made on the basis of semantic gist. Additional false-memory phenomena generated by fuzzy-trace theory (and demonstrated in children) include mere-memory testing (prior nonsuggestive memory tests of semantically related material increase false recognition for that material), false memory persistence (gist-based false memories are more persistent over time than initially verbatim-based true memories), repeated cuing of gist (gist repetition, especially unopposed by verbatim repetition, produces vivid false recognition), and false-recognition reversal (semantically related items can be misrecognized less often than unrelated items under conditions of high verbatim priming or accessibility; see Brainerd & Reyna, 2002, 2005; Reyna & Brainerd, 1995; and Reyna & Lloyd, 1997 for discussions). There is some debate over whether fuzzy-trace theory uniquely predicts these effects or whether sourcemonitoring theory could be extended to cover these findings (Lindsay &Johnson, 2000; Reyna, 2000). The bottom line of these variants on standard false-memory effects is that they demonstrate the ease with which natural properties of human memory conspire to create false memories in everyday life and that conditions that strengthen or cue gist memories are likely to increase spontaneous false memories in children. As Reyna and Lloyd (1997) discuss, internally generated inferences and semantic gist can be made more accessible through externally provided cues that work in synergy with a prepared mind

FALSE MEMORY IN THE DEESE-ROEDIGERMCDERMOTT WORD-LIST PARADIGM

In tasks involving spontaneous false memory in which memory is tested shortly after materials are presented, the typical pattern of findings for elementary school-aged children and for adults is that presented items are recognized more readily than semantically related false items are misreported as "recognized." (Younger preschool-aged children are more variable in their responses and have less accurate verbatim memories.) The extent to which children and adults discriminate between presented items and semantically related items hinges on the nature of instructions. Reyna and Kiernan (1994) determined that it was essential to explicitly present true nonpresented sentences and explain that they should be rejected; otherwise, children and many adults assumed that instructions to affirm only sentences that were the "same" as those that were presented

486

REYNAET AL.

meant that they should affirm sentences that were the same in meaning, not in surface form. Verbatim identity instructions are unnatural in everyday life and, thus, require special procedures (e.g., concrete examples), although they mimic the constraints of legal testimony. Investigations subsequent to those of Bransford and Franks (1971), in which familiarity of elements and other methodological confounds are better controlled, have shown that their dramatic false recognition effects are difficult to replicate when instructions to reject meaning-consistent items are clear (for a review, see Brainerd & Reyna, 2005). However, dramatic false recognition and recall effects have been reliably obtained in adults with the use of the Deese-Roediger-McDermott (DRM) paradigm. The DRM paradigm is a procedure originally devised by Deese (1959) and recently rediscovered by Roediger and McDermott (1995) and Read (1996). The task involves presenting a list of 12-15 words that are semantic associates of a word that is not presented (the critical lure), for example, nurse, sick, health, hospital, ill ... (critical lure: doctor); or, bed, rest, tired, awake ... (critical lure: sleep). For both recognition and recall, the typical finding is that intrusions of the critical lure occur with a probability that is only about 10% lower than recall of studied items, and false recognition of the critical lure occurs with a probability that is comparable to that of correct recognition of studied items. In addition, subjects report high degrees of confidence in these judgments and often claim to remember experiencing the lure at study when asked whether they "remember" or just "know" that the item was presented. Although it might seem at first blush that false memories for words from semantically related lists have little to do with forensically relevant memories for real-life events, at a deeper level, the DRM task illustrates many of the common features of forensically relevant memories, and the findings derived from this task appear to apply broadly to memory for complex events. For example, in real life, child victims of crimes often experience repeated, related events (e.g., of abuse). Experiencing subsequent events is likely to cue memories for earlier events that are substantively similar, especially if victims realize that the events are not exact repetitions but are related. If victims connect similar events over time, despite accurate memory for actual details of separate instances, they may nevertheless become confused about whether events that are consistent with the theme of repeated events occurred. In addition, investigative interviews and sworn testimony are likely to take place days, weeks, months, or even years after crimes occurred, making meaning-based memory distortions more likely. Furthermore, the form of memory questions in legal contexts can vary from highly specific, if not leading, questions to open-ended free recall, another factor investigated in DRM studies (see Reyna, 1998, and Reyna, Holliday, & Marche, 2002, for other forensically relevant examples). Before we proceed to a review of the published research on children's propensity to exhibit false memories in the DRM paradigm (and what this means for the accuracy of their memory reports), it is important to outline some of the major explanations for this dramatic effect. Understanding the nature of the effect in adults provides a context in which effects with children can be interpreted. Deese originally characterized the probability of lure intrusion as a function of the mean associative strength of the list items, defined as the average probability that list words (targets), all of which are associates of

17. FALSEMEMORY

487

the lure, also elicit the lure as an associate of themselves (e.g., the average probability that words such as bed and rest, which are associates of sleep, also elicit sleep as an associate of themselves). However, Robinson and Roediger (1997) manipulated the number of associates on the studied list by adding lower (i.e., more weakly related) associates or by holding the list length constant and replacing associates with unrelated items. Inasmuch as the mean associative strength of the lists decreases in either situation, Deese's account implies that the probability of falsely recalling the critical lure should decrease under both conditions. In the first condition, Robinson and Roediger found that the probability of falsely recalling the critical item increased with the number of associates on the list; in the second manipulation (holding the list length constant), adding unrelated words had no effect on the probability of lure intrusion. Taken together, these observations suggest that it is the summed backward associative strength-rather than the mean-that determines the probability of lure intrusion. Children's associative networks for words are known to be well established as of the preschool years, and therefore their patterns of activation from that point onward ought to be similar to those of adults (Bjorklund, 1987). Thus, if summed associative strength were the source ofDRM effects, then children ought to display those effects. As we discuss in greater detail below, children do not display such effects, which is evidence against the summed activation account. Variants of Underwood's (1965) notion of an implicit associative response (IAR) have been incorporated into many explanations of the DRM effect. In its original version, the high association between studied items and the lure results in the conscious activation of the lure during study. The lure thus behaves as if it were actually presented in the original list, and the high proportion of "remember" responses given to lures supports this idea. However, this explanation is refuted by the finding by Robinson and Roediger that increasing the number of studied words has opposite effects on the recall of presented words versus critical lures. "Highly associated non-presented items, implicitly activated during study, should reveal the same patterns of recall as do studied words. Clearly, veridical recall drops as a function of the number of studied associated items, whereas false recall increases" (p. 236). Robinson and Roediger entertain a modified activation hypothesis in which the critical lure is activated multiple times, which might explain the increase in the probability of its recall with more opportunities for repetition (although, as discussed below, this account does not explain why critical lures generated in tandem with the study list are not forgotten much the way studied words are). According to Robinson and Roediger (1997), extant findings are consistent with both a modified activation account (with associative strength equivalent to activation) and a fuzzy-trace theory account based on repeated cuing of gist (each word repeatedly cues the gist theme of the list-and the critical lure is an excellent cue for that gistalthough none of the targets are repeated verbatim; e.g., Reyna & Lloyd, 1997). Repeated cuing of the same gist produces vivid false memories that behave, paradoxically, like gist memories in terms of resistance to forgetting and dependence on meaning, but they acquire an illusory intensity that we have called phantom recollection (Brainerd, Wright, Reyna, & Mojardin, 2001). Crucially, for repeated cuing of the same gist to be

488

REYNAET AL.

effective, people must connect the gist from one instance to another. Otherwise, different gists are each cued once, instead of repeated cuing and strengthening of the same gist theme across items. As Holliday, Reyna, and Hayes (2002) and Reyna and Brainerd (1998) discuss, it is important to separate phenomenological experiences (vivid recollection versus vague familiarity) associated with memory representations from the nature of memory representations themselves (precise verbatim memories versus fuzzy gist memories). Although the norm is that vivid recollective phenomenology accompanies retrieval of verbatim memories (and vague familiarity is generally associated with gist memories), when gist is repeatedly cued as in the DRM paradigm whereas each verbatim word is presented only once, gist memories become associated with vivid (but illusory) recollective phenomenology (Brainerd et al., 2001; Brainerd, Payne, Wright, & Reyna, 2003). This fuzzy-trace theory account of the effect of repeated cuing of gist on conscious experience (or phenomenology) of memories was originally applied to memories for inferences (Reyna, 1995; Reyna & Brainerd, 1995; Titcomb & Reyna, 1995) and subsequently applied to the DRM paradigm, without modification. Source-monitoring theory is sometimes taken to imply, in combination with activation theories, that the critical lure is implicitly activated at study (or at test) but not successfully monitored or gated out at test (Roediger, Watson, McDermott, & Gallo, 2001). However, both the modified activation account and the latter source-monitoring account fail to explain why memory for critical lures endures over time (consistent with memory for gist) rather than following a forgetting process, similar to that of targets, that is initiated at study (e.g., Brainerd et al., 2001, 2003; McDermott, 1996; Payne et al., 1996; Toglia, Neuschatz, & Goodwin, 1999). Source monitoring also encompasses shifts in response criteria or thresholds (Johnson, Hashtroudi, & Lindsay, 1993), that is, that people sometimes adopt looser or stricter criteria for judging a memory as true or experienced (in addition to basing source judgments on perceptual and semantic similarity). However, the DRM effect is unlikely to be influenced by social demand factors or caused exclusively by strategic criterion shifts (compare Miller & Wolford, 1999, with Roediger & McDermott, 1999). The structure of the DRM task minimizes the contribution of demand influences, and this effect is routinely elicited in computerized versions of the task with no experimenter immediately present. According to the criterion shift account, meta-knowledge concerning the gist of studied items makes subjects more likely to respond "old" if the test item is consistent with that gist. However, Gallo, McDermott, Percer, and Roediger (2001) showed that providing detailed instructions about the nature of the DRM task to subjects after study had no effect on false recognition when subjects have studied multiple lists that may or may not have included the critical lure. These considerations illustrate that demand factors or conscious strategizing is not critical to this false memory phenomenon. The failure to find a large role for demand effects or conscious strategies would imply robust false-memory effects in children because conscious strategizing is a late-developing skill. Hence, the DRM false-memory effect does not require metacognitive skills that children lack. As this review of explanatory mechanisms in the DRM paradigm suggests, traditional developmental predictions vary from early demonstrations of false memory effects (i.e., in younger elementary school-aged children, if not younger) followed by little developmental change, which would seem to fall out of activation theories, to developmental

17. FALSE MEMORY

489

declines in false memory effects as clearly predicted by source-monitoring theory. Fuzzytrace theory's dual process assumptions lead to predictions of divergent developmental trends under specific conditions (Brainerd & Reyna, 2005). According to fuzzy-trace theory, children store separate representations of both the exact surface form of their experiences (called verbatim traces, although they apply equally to pictures, numbers, events, and other meaningful stimuli in addition to verbal stimuli) and of their interpretation of the meaning of those experiences (called gist traces, which capture relations, patterns, inferences, and other elaborations of experience filtered through children's understanding). Both verbatim and gist memory develop; in particular, the ability to use verbatim memories to reject gist-consistent false memories improves during childhood as does the ability to spontaneously notice conceptual (or gist) relations that hold across a series of instances. We discuss the former verbatim developmental trend in the section that follows. With respect to the latter (connecting-the-gist limitations), ample evidence can be found in studies of free recall of semantically related word lists. When studied word lists can be organized into categorically related subgroups, adults typically cluster those subgroups of words together when they recall the list. When young children recall the same types oflists, they do not cluster words into related subgroups (e.g., Bjorklund, 1987). A series of studies has shown that the lack of clustering is not related to retrieval difficulties, but instead reflects more fundamental differences in noticing and storing meaningful relations that hold across items. Children also fail to show proactive interference (Bjorklund & Hock, 1982). Adult recall of a series of word lists involving semantically related items becomes progressively worse on successive related lists; children do not show this effect until adolescence. Hence, based on these data, we would expect that children would be less apt to note that the words bed, rest, tired, awake and so on all have to do with sleep-despite understanding each individual word. If the DRM false memory effect is caused by repeated cuing of related gist, then children should be less likely to show the effect because they are less likely to notice the gist theme and thus less likely to accumulate the impact of multiple cues to the same gist. In the first article published on the DRM effect in children, Brainerd, Reyna, and Forrest (2002) showed that the DRM effect was not present to any significant extent among a sample of 60 five-year-olds in a recall task. For the sake of comparability to adult data, identical lists and procedures were used as in many experiments with adults; children recalled 10 lists seriatim with 12 words per list. Children recalled 23% of studied words per list on average. Intrusions of critical lures occurred for only 6% of lists, compared with about half of the lists in experiments with adults using the same materials (Payne et al., 1996; Roediger & McDermott, 1995). The qualitative adult patterns were also absent in the young children. Adults readily appreciate the gist theme of each list, and not only do they recall critical lures that are consistent with the gist theme of the lists, they also rarely intrude items from different (categorically unrelated) lists. For children, although the critical nonpresented word was recalled for only 6% of lists, 22% of lists produced recall of unrelated words from earlier lists. Two subsequent experiments replicated and extended these findings. In the second experiment, 5- and 7-year-olds were compared to determine whether onset of formal schooling or other developmental changes from 5 to 7 might increase false memories

490

REYNAET AL.

(Sameroff & Haith, 1996). Furthermore, lists known to produce the highest levels of false recall in adults were included (and compared with those that produce low levels) to maximize the opportunity to observe an effect. In contrast to patterns for adults, "low" lists produced better recall of presented words than "high" lists for 5- and 7-year-oldsbut there was no difference between high and low lists for false recall. There was an age increase in true recall, but not in false recall, which remained very low, as it had been in the first experiment. For example, 5-year-olds recalled 34% and 7-year-olds recalled 40% of the studied items on the low lists. However, only 7% of lists produced false recall of the critical lure. In a third experiment, both recall and recognition were tested, and the age span was widened to include 5-year-olds, 11-year-olds, and young adults. Across this wider age span, differences in false recall were now detected, as one might expect. For young children, false recall of critical words remained very low (7% pooled across the three experiments), high-low list differences were nonexistent but then emerged with age, and intrusions from earlier lists were more prevalent than intrusions of critical words. Adolescents showed qualitative and quantitative patterns that were intermediate between those of young children and those of adults. For example, by adolescence, false recall increased to 27% (compared with 53% in adults). Recognition elicited the same developmental trends in false memory, although false memory effects could be detected even in young children because recognition is a more sensitive measure of memory. These results are clearly consistent with fuzzy-trace theory's prediction that young children will have difficulty connecting the conceptual dots among list words (which results in repeated cuing of the same gist) and thus will be less likely to show DRM falsememory effects. Brainerd, Holliday, and Reyna (2004) also studied false recognition in children and adolescents with the use of both the DRM paradigm and standard categorized word lists. The novel feature of this research was that an experimental procedure, called conjoint recognition, was used that delivers separate measurements of processes that contribute to the DRM illusion, including distinguishing the usual gist-based memory reports that are accompanied by familiarity phenomenology versus gist-based memory reports that are accompanied by phantom recollection (i.e., vivid illusory recollection). In the DRM experiment, children ranged in age from 7 to 14 years and were presented with three blocks of DRM lists, and each block was followed by a recognition test. The probes on recognition tests consisted of studied targets, semantically related distractors (such as critical lures), and other distractors that were unrelated to list themes. Each block consisted of three different DRM lists, for a total of nine lists. Consistent with the earlier findings of Brainerd et al. (2002) for recognition, the false alarm rates for critical lures were substantial at 51% for 7-year-olds (the youngest age tested), but rates increased appreciably during this age range to near-adult levels of 71%. The conjoint recognition model was used to mathematically separate the relative contributions of different memory processes to these effects. Application of the conjoint recognition model revealed the reasons for this developmental trend: it was due wholly to phantom recollection. Specifically, the tendency of children to falsely recognize critical lures on the basis of familiarity did not change with age, whereas the parallel tendency to falsely recognize critical lures on the basis of phantom recollection increased

17.

FALSEMEMORY

491

dramatically with age. Thus, the results of this experiment not only replicate the earlier developmental trend in the DRM effect; they localize the trend within this intriguing illusory phenomenology (which is associated with repeated cuing of gist). In a second experiment, Brainerd et al. (2004) used the conjoint recognition procedure with categorized lists to test the idea that repeated cuing of gist induces phantom recollection. The subjects in this experiment were in the 5-11-year age range. The procedure was very similar to that of the first experiment. The children received three lists, with a recognition test following each list. Each list was composed of three subgroups of words that were taxonomically related (e.g., exemplars of furniture, colors, and animals). As in the first experiment, the recognition test that followed each list was composed of studied targets, semantically related distractors that were unstudied exemplars of each category, and unrelated distractors that were exemplars of nonpresented categories. Note that there were three gist themes per study list and each gist theme was cued by fewer exemplars, compared with the much larger number of cues to shared gist in the DRM paradigm. Therefore, we would expect less phantom recollection with these materials. In line with this prediction, measured levels of phantom recollection were much lower than in the first experiment, though phantom recollection increased with age as in the first experiment. In another study using the DRM paradigm, Howe, Cicchetti, Toth, and Cerrito (2004) focused on special populations of children and conducted both recognition and recall tests. Howe et al. were concerned with the effects of chronic stress associated with child maltreatment on basic memory processes. Differences in basic memory processes between maltreated and nonmaltreated children were examined in an experiment in which middle SES (socioeconomic status), low-SES maltreated, and low-SES nonmaltreated children (aged 5-7,8-9, and 10-12 years) studied 12 DRM lists. They found that both true and false memories increased with age regardless of whether recall or recognition measures were used, and, contrary to some speculation, these trends did not differ as a function of a history of maltreatment. Like results for older children in the two studies by Brainerd et al., false memories were more likely with high than with low lists, although true recall did not differ. This pattern held for all samples regardless of SES. However, there were differences in memory performance as a function of socioeconomic status (low-SES children had lower accuracy), consistent with other findings in the literature. Brainerd et al. (2002, 2004) and Howe et al. (2004) used word lists that had been composed for adults and normed on adults. In several unpublished studies, Reed, Mangan, Warren, Price, and Metzger (2003) and Metzger et al. (2004) used "child appropriate" lists (ones generated by the children themselves) and obtained the same age trends found in the studies of Brainerd et al. (2002, 2004) and Howe et al. Price, Metzger, Williams, Phelps, and Phelps (2001) also found that false recall and recognition of the critical nonpresented words increased between childhood and adolescence and again between adolescence and adulthood. Thus, the adult materials used by Howe et al. and other studies do not seem to produce spurious age trends in true and false memories due to correlations between age and word difficulty. The finding of Howe et al. that false memories increased with age is consistent with the other findings we have reviewed for the DRM procedure (e.g., Brainerd et al., 2002

492

REYNAET AL.

and 2004, as well as the unpublished studies discussed above and dissertations discussed below) and is consistent with fuzzy-trace theory's gist-connection explanation. Because it is critical that gist relations among list items be extracted for strong false memories to be generated, younger children whose gist processing is less well developed than that of older children are less likely to be susceptible to the DRM illusion. Indeed, across studies using the DRM paradigm, younger children's overall accuracy (especially for recall) was often found to be higher than older children's because false memory increased more with age than true memory did. Thus, contrary to the literature on children's suggestibility, in which younger children are generally portrayed as being more susceptible to misinformation than older children (see review by Bruck & Ceci, 1999), spontaneous, gist-based false memories in the DRM paradigm are more likely in older than in younger children. Three additional unpublished studies echo this theme; they all showed developmental increases in the DRM effect (Forrest, 2002; Holliday, Reyna, & Brainerd, 2006; Karibian, 2003). In an unpublished dissertation, Karibian (2003) administered the DRM task to 64 second and fifth graders, 32 at each grade level. Half of the children at each grade level were identified (and had been formally evaluated) as learning disabled and the other half, drawn from the same schools, were not. Karibian presented eight 12-word lists (four high lists, those that produce high levels of false memory in adults, and four low lists) followed by a terminal recognition test. Children either recalled the lists prior to recognition or did not, another factor known to increase levels of false memory on subsequent recognition tests in adults. Analyses comparing ages and disability groups were conducted to determine effects of general learning ability on false memory. Recognition tests included targets, critical lures, other semantically related distractors (the remaining three items from the original 15-item lists), and unrelated items taken from nonpresented DRM lists. Children were tested immediately and after a 1-week delay. Although null effects (findings of no difference) are subject to the criticism that the study lacked sufficient statistical power to detect differences, the general pattern of results is similar to prior studies with larger sample sizes. As in earlier studies, for recall, the critical distractor was not the most common intrusion. Predominant intrusions were not related to the list theme, in contrast to intrusions for adults, which are almost exclusively related to the list theme. Although recalling DRM word lists increases false recognition in adults, there was no such effect in children. For recognition, again as in earlier studies, false recognition effects were obtained at all age levels. Although younger children were not significantly less likely to show false recognition effects (the "younger children" were 8.6 years old), learning-disabled children at each age level showed less false recognition than nondisabled children. At the delayed test, both age groups showed false recognition effects, and a small age increase in false recognition emerged. In another unpublished dissertation, Forrest (2002) had substantially more power to detect differences, with 60 children at each of two age levels, and a wider age range: kindergartners and sixth graders. Forrest presented sixteen 12-word lists, eight high and eight low lists, and half of the lists contained words ordered from strongest to weakest

17. FALSE MEMORY

493

association to the critical word (and the other half were ordered weakest to strongest). The recognition test consisted of the same types of items as on Karibian's test, and recall versus no recall was again manipulated (but within subjects; subjects recalled half of their lists). Finally, half of the lists were tested immediately and all of the lists (previously tested and not) were tested after a delay. Although adult false recognition has been found to be greater for high (vs. low) lists, strong-to-weak ordered lists (vs. weak-to-strong ordered lists), and lists that have been previously recalled (vs. not recalled), none of these effects were found for the younger children. Only the high-low effect was evident for the older children. As demonstrated in other studies, false recognition and false recall of critical distractors increased substantially with age. True recognition and true recall also increased with age, but as in previous studies, not as substantially as false recognition and recall. Levels of false recall and recognition in younger children, for example, were almost identical to levels observed for the same age group, kindergartners, in the study by Brainerd et al. (2002), and, like that study, levels for older children were higher, but did not quite reach adult levels. There were qualitative similarities to prior work as well. For younger children, intrusions on the recall test were not predominantly semantic; intrusions were mainly from prior lists that were semantically unrelated to list themes. For older children, approximately half of the intrusions were semantically related to list themes, consisting of a mix of critical lures and other semantically related words. The effects obtained on immediate tests were also obtained on delayed tests. Naturally, the effect of prior (recall) testing could only be evaluated on the delayed (recognition) test. As in the mere-memory testing effect discussed earlier for standard word lists (e.g., Brainerd & Reyna, 1996; Reyna & Lloyd, 1997), prior testing increased false recognition of semantically related items. These findings further support a gist-based explanation for the DRM false memory effect and for its development with age. Thus far, the studies we have discussed on the DRM paradigm have uncovered similar developmental trends, and their results have resembled one another quantitatively and qualitatively. We have reviewed three published articles containing a total of five experiments (Brainerd et al., 2002, 2004; Howe et al., 2004): two unpublished dissertations (Forrest, 2002; Karibian, 2003) and three additional unpublished studies (Metzger et al., 2004; Price et al., 2001; Reed et al., 2003), which together suggest that false memories increase with development and that memory errors shift to reflect the gist themes of semantically related word lists. However, in one of the first published studies using the DRM paradigm, Ghetti, Qin, & Goodman (2002) found seemingly contradictory results. As shall become apparent, however, Ghetti et al. implemented methodological changes that place this study in the category of research on resistance to false memories rather than false-memory susceptibility (see the section below on recollection rejection). The key evidence for this claim is that adults failed to show the usual falsememory effects, and children did not differ significantly from this adult pattern. Ghetti et al. presented 10 lists of 7 words each (rather than the usual12-15 words) to children (5- and 7-year-olds) and adults. Half of the subjects watched a videotape of an experimenter reading the words, and the other half watched a videotape of the experimenter reading the words but also showing a line drawing of the object named by the

494

REYNAET AL.

word. Source was manipulated by having a male and a female experimenter each read 5 of the 10 lists, in counterbalanced order. Recall occurred after each list was presented, and after all 10 lists were presented, children received a recognition test. The recognition test consisted of 60 items, 30 studied words (3 items from each of the 10 lists) and 30 nonstudied words: 10 critical lures (one from each list), 15 items from nonstudied word lists, and 5 critical lures from nonstudied lists. Children also provided confidence judgments for their recognition responses, using a child-friendly 3-point scale, anchored by a picture of a confident-looking child and an uncertain-looking child. Ghetti et al. found an improvement in true recall for studied items at each age level, but no difference across age in false recall. However, the overall percentage of false recall was only 20%, and the picture manipulation decreased levels of false recall. Even in the no-picture condition, adults did not show the usual DRM pattern; they recalled 88% of the studied items and a meager 15% of the critical lures. (As we have discussed, the decrease in the number of studied words would be expected to lower false memory levels [Robinson & Roediger, 1997].) When differences in false recall were not detected, Ghetti et al. calculated an alternative measure of false recall and subjected it to planned comparisons involving age. Using this measure (which involved dividing the number of critical lures recalled by the total number of studied and nonstudied words recalled), they found that 5-year-olds had significantly higher false recall scores than older children and adults. However, this measure will yield higher false recall scores in younger children if true recall improves with age and false recall does not change because the same numerator (false recall) would be divided by a bigger and bigger denominator (i.e., true recall that is increasing with age). Thus, it is not accurate to conclude from the alternative scores analysis that false recall is decreasing with age when it is actually not changing. For recognition, the picture manipulation both increased true recognition and decreased false recognition. As in recall, there was a significant effect of age on true recognition (adults differed from both groups of children), but no age difference for false recognition. Confidence ratings also differed for studied items and critical lures; studied items were endorsed with higher confidence. The typical developmental trend in source discrimination accuracy was obtained (younger children were more accurate than older children and adults in correctly judging which experimenter had read the word), but it was significant only for the picture condition. Apparently, the younger children were inferior in source-monitoring ability (null effects are inconclusive, not evidence against a hypothesis, meaning that the age difference in the picture condition is the only relevant evidence pertaining to source monitoring). However, source monitoring could not explain age differences in false recognition because there were no differences. As in recall, the recognition data for adults were unusual; adults in the no-picture condition correctly recognized 88% of studied items but falsely recognized only 36% of the critical lures (and an even lower percentage of lures in the picture condition, 28%). These unusually low levels of false recall and recognition for adults would make it more difficult to detect age differences, especially age increases in false recall and recognition. Wellknown effects of using pictures (memory performance is better for pictures than words, and even better when dual presentation modes are used) and shortening the word lists

17. FALSE MEMORY

495

explain why false recall and recognition were so low. (Perhaps presenting word lists via videotape as opposed to audiotape was also engaging and enhanced encoding.) Therefore, the Ghetti et al. study provides a straightforward test of the main explana~ tory hypothesis concerning recollection rejection discussed in the next section, namely, that enhancing verbatim memory in traditional ways reduces false memory. Ghetti et al. compare the major theories of false memory, differentiating the theories' developmental predictions concerning age trends as well as their implications for phe~ nomenological experience. They point out, for example, that the distinctiveness heuris~ tic (e.g., Israel & Schacter, 1997; Schacter, Israel, & Racine, 1999) is based on metacog~ nitive awareness that true memories of studied items should include recollection of distinctive details. For example, if every word has been presented with a picture, people should be less likely to say that a word has been presented if they cannot retrieve the distinctive picture. Because this is a strategic decision process involving adopting a con~ servative response bias, it is unlikely that young children could implement the distinc~ tiveness heuristic. However, use of item~ specific information to better discriminate stud~ ied from nonstudied items (or improved encoding or accessibility of verbatim memory traces) would be within the abilities of young children. Thus, the fact that the picture manipulation significantly reduced false recall and recognition in children argues against the use of the distinctiveness heuristic in favor of more basic memory discrimination processes. Ghetti et al. also note that complex inferential processes used to attribute sources to memories (as described by Mather, Henkel, & Johnson, 1997) are also un~ likely to be within the abilities of young children. Ghetti et al. made true memories more distinctive and thereby reduced false mem~ aries. However, what underlies the more typical pattern that false memories increase theory explanation for age differences in DRM effects were with age? If the fuzzy~trace correct, it would pinpoint a highly delimited factor to explain the low levels of false~ memory effects in young children. The undoing of that factor ought to be correspond~ ingly simple. If children fail to notice that the words all fit a gist theme, then false mem~ aries ought to be induced by telling children the gist theme (e.g., names of fruit, words about sleep) of each list prior to study. (Children may also fail to semantically elaborate on each item, which would not necessarily be greatly affected by a single cue prefacing the entire word list.). Brainerd et al. (2004) implemented this simple manipulation of cuing the gist theme for categorized lists (which generally show less false recognition than DRM lists, even in young children), and Holliday et al. (2006) implemented it for DRM lists. As predicted, Brainerd et al. found that false recognition and corresponding levels of phantom recollection increased significantly for both younger and older chil~ dren (the latter group of fifth graders would be at the threshold of showing clustering effects in free recall and thus could still benefit from cues to connect the gist across items). Holliday et al. also found that false recall rose to significant levels in young chil~ dren, as predicted. In focusing on the main explanation for the main findings concerning the DRM effect in children, we do not mean to imply that the inability to connect gist is the only determinant of children's false memory performance. Children's false memory performance is also known to be influenced by less in~ depth understanding of the gist of infrequent

496

REYNAET AL.

words, the ability to engage in some rejection processes (i.e., memory for verbatim targets that allows rejection of related lures, reviewed below), and the general developmental trend of increasing dissociation between verbatim and gist processing with age (e.g., Reyna, 1995; Reyna & Kiernan, 1994). Hence, for some difficult or infrequent words on DRM lists, children may have less understanding of their gist; they may be less successful in extracting the gist of such words (although similar age trends have been obtained with easier words). Regarding dissociation, younger children are more likely to rely on gist memories in a verbatim task (such as the DRM or sentence memory tasks that we have discussed) or on verbatim memories in a gist task (e.g., responding on the basis of the literal contents of memory in a comprehension task; Brainerd & Reyna, 1993). Thus, they would be more apt to interpolate gist memories into the DRM task, assuming that those gist memories had been successfully extracted for each item and connected across items. It is unlikely that children engage in some of the other processes that may affect adult performance in the DRM task, such as strategic monitoring (metacognition, or strategic thinking about one's own cognition, is a late-developing skill). This discussion highlights that there is evidence for multiple, countervailing influences on children's false memory. However, it is important to point out that these multiple influences are predicted to influence false memories in specific ways under specific circumstances and that mathematical models have been used to disentangle the relative contributions of underlying opponent processes on manifest behavior. Therefore, despite countervailing influences, it is possible to derive clear expectations for children's false memory performance under a range of concrete conditions. These countervailing or opponent processes are illustrated in the study by Holliday et al. (2006). In addition to investigating gist cuing, Holliday et al. examined the effect of verbatim repetition on children's false recall and recognition. Children (7-, 9-, 11-, 13-, and 15-year-olds) studied six DRM lists, recalled each list after it was presented, and then responded to a recognition test. Once again, false-alarm rates for nonpresented words increased with age in free recall and recognition tests as predicted by fuzzytrace theory. As alluded to above, gist cuing increased false recognition of nonpresented words at all age levels. Also as predicted by fuzzy-trace theory, verbatim repetition of lists increased true recall and recognition of targets and decreased false recall and recognition of nonpresented words. As we now discuss, the latter effect pertains to processes that support correct rejections of false items in recognition and resistance to false intrusions in recall.

RESISTING FALSE MEMORIES

The phenomena we have discussed involve factors that generally induce or increase false memories. Although we have noted that false memories result from natural processes of encoding and storing meaningful events, are there equally natural processes for rejecting false memories? If we understood such processes, it should be possible to create conditions that would facilitate accurate memory reports. In many legal cases, memory reports are the main and sometimes the sole evidence concerning a crime-especially for child

17.

FALSE MEMORY

497

victims of sexual abuse (for illustrative cases, see Brainerd & Reyna, 2005). Improving the accuracy of testimony of child victims and witnesses is an attainable goal if we apply the scientific research that is currently available. The memory-independence finding (e.g., that memory for inferences and other true nonpresented sentences is independent of memory for the presented sentences on which the inferences are based) and the other evidence favoring separate gist and verbatim memory representations suggest a very specific mechanism for rejecting false memories. Remembering exact events becomes increasingly difficult as time passes. The independence effect occurs at an intermediate point in time at which experienced events can still be recognized (by retrieving verbatim memories). At a later point in time, only gist memories are accessible and are used to recognize both experienced events (e.g., presented sentences) and gist-consistent items (e.g., true inferences), producing positive dependency: remembering the gist of what happened supports saying "yes" in a memory test to both actually experienced events and gist-based reconstructions. What about the earliest point in time, immediately after events have occurred while they are fresh in one's mind? In that case, verbatim memories would be accessible, given any cue related to the event. An item that expresses the gist of what occurred would remind the witness of what actually happened, which could still be retrieved. Naturally, a verbatim replica of what was experienced would also be an excellent reminder of what actually occurred. Note that retrieving a verbatim memory for what actually occurred should lead to opposite responses: "yes" to the verbatim replica but "no" to the gistbased facsimile. The better the witness remembers the original experience, the easier it would be to reject related (but nonexperienced) lures. The latter relation is one of negative dependency. Reyna and Kiernan (1995) reasoned that the key to inducing negative dependency, the ability to reject nonexperienced lures based on accurate verbatim memories, would be maximizing verbatim accessibility. As we discussed earlier, there are a number of factors that are known to increase the strength or accessibility of verbatim memories, such as materials to be remembered, age of the witness, and delay between the experience and the memory report (e.g., Reyna & Brainerd, 1995). Reyna and Kiernan (1995) obtained the predicted negative dependency with memorable novel metaphors (for which verbatim wording is crucial in conveying meaning) as the to-be-remembered materials, young adult subjects as "witnesses" (who have better verbatim memories than children and older adults), and immediate rather than delayed testing after metaphor presentation. (Metaphors were presented along with other sentences as part of narratives in the study phase, and multiple types of true and false sentences were tested in a recognition task immediately and after a week's delay.) As these levels of factors were replaced, for example, testing children rather than adults, negative dependency decreased and eventually became insignificant. Brainerd and Reyna (2002) re-analyzed these and other data, using the conjoint recognition model discussed earlier. Parameters of the mathematical model estimated the contribution of verbatim-memory-based rejection of false memories and gist-memorybased acceptance of false memories. Brainerd and Reyna referred to the former process as "recollection rejection." Model parameters estimating these processes responded

498

REYNAET AL.

appropriately to experimental manipulations (e.g., the gist memory parameter increased after delays), providing some assurance that the parameters measure what they purport to measure. (In 12 experiments, Brainerd, Reyna, Wright, & Mojardin [2003] also demonstrated that the recollection rejection parameter responded predictably to verbatim manipulations, such as list repetition and presentation in unusual type fonts.) Consistent with expectations about developmental differences in verbatim memory, there were significant improvements in recollection rejection with development (defined either as age differences or as differences between developmentally delayed and control children of the same age). When the task was not the DRM or other paradigms that involved repeated cuing of the same gist, developmental improvements in recollection rejection overtook developmental improvements in gist-based memory processing, resulting in a net decrease in false memory reports. Also, when younger children did correctly reject false-but-gist-consistent items, their basis for doing so was sometimes different from that of older children. Younger children rejected gist-consistent items, in part, because they failed to access gist representations. For older children, rejection of gist-consistent items occurred despite access to gist representations. However, it should be stressed that recollection rejection significantly contributed to performance even at the youngest age groups (6 years old) tested. In the study by Brainerd et al. (2004) discussed earlier, with DRM materials and categorized lists, the conjoint recognition model was also used to estimate recollection rejection. Developmental trends for both types of word lists paralleled the results for sentences obtained by Brainerd and Reyna (2002), namely, that recollection rejection improved with age. The developmental relationship between true recollection (which concerned the targets that were actually presented) and recollection rejection was as one might expect if they were both based on verbatim memory for presented targets. For example, using the DRM procedure, the parameters for true recollection and recollection rejection both increased from age 7 to age 14 (and similarly for categorized lists from age 5 to age 11). Brainerd et al. (2004) also implemented a gist cuing manipulation (providing children with the gist theme in advance of each list) in connection with categorized lists. They found that this manipulation reduced recollection rejection (i.e., the recollection rejection parameter was lower) and increased phantom recollection, thereby attenuating age differences. Recollection rejection does not depend on metacognitive knowledge or complex decision strategies and inferences (though what is recollected may be one ingredient in a metacognitive strategy [Brainerd et al., 2003; Ghetti & Alexander, 2004]) and, thus, is within the capabilities of young children. Because recollection rejection depends on the retrieval of verbatim traces, memory test cuing manipulations that selectively enhance the accessibility of verbatim information (e.g., memory for surface information such as exact wording) should decrease false memory reports. As we have discussed, Ghetti et al. (2002) performed such manipulations, including presenting pictures along with words, as did Holliday et al. (2006), who varied verbatim repetition of targets. Both studies showed that these manipulations reduced false memory effects. Brainerd, Reyna, and Kneer (1995) also performed such manipulations. In five experiments, they presented lists of unrelated common nouns (including typical exemplars of familiar categories, such as cat from the category of animals, as well as category names, such as flowers) to subjects

17. FALSE MEMORY

499

ranging in age from 5 years old to adults, followed by recognition tests. Recognition tests in different experiments contained such items as presented words, typical exemplars of familiar categories, category names, semantic associates, and words unrelated to any that were presented. The accessibility of verbatim memory was enhanced by varying repeti~ tions of target words at study, among other manipulations. For example, in the first ex~ periment, 30 words were presented once, and the other 30 randomly intermixed words were presented three times each (at least 5 items intervened between presentations of repeated items). On the recognition test, some target~resembling distractors were pre~ ceded by presentation of the targets to which they were related. Placing the presented words either immediately prior to their related dis tractors or several words back (4 other test items intervened between the presented target and its related distractor) on the recognition test was also expected to enhance the accessibility of verbatim memory. Children were tested under two types of conditions in different experiments: easier verbatim recognition (when the same speaker presented words at study and test) and harder verbatim recognition (when different speakers presented words at study and test). Under more difficult conditions for verbatim recognition, younger children (kindergart~ ners) falsely recognized related distractors at a higher rate than unrelated distractors (gist memory rather than verbatim memory drove responses to distractors), but older children (third graders) showed the opposite result: they falsely recognized related dis~ tractors at a lower rate than unrelated distractors (verbatim memory rather than gist memory drove responses to distractors). The latter pattern was called "false recognition reversal" because the usual pattern of greater false recognition for related than for unre~ lated test items was inverted. As expected, repetition at study and immediacy of verbatim priming at test both sig~ nificantly enhanced false~recognition reversal. Under easier verbatim recognition condi~ tions, children as young as 5 years old displayed false recognition reversal, and reversals were again larger when studied items had been repeated and when verbatim priming occurred immediately prior to the related distractor than when it occurred several items reversal was attenuated or back on the test. As would be expected, false~recognition eliminated on delayed tests, presumably because verbatim memories had become less ac~ cessible. In another experiment, when related items preceded presented targets on the test (either immediately prior or four items prior to the presented target), the usual false recognition pattern was obtained. Thus, reversals occurred only when presented items preceded related words, arguing against general similarity or response bias strategies. Making what was actually presented more accessible in memory allowed children to re~ call that item and to reject distractors when presented with any reminder or related cue theory, the memory processes underlying to the original item. According to fuzzy~trace false recognition (gist~based comparisons) and false~recognition reversal (verbatim~ based comparisons) are within the purview of younger children as well as older ones (Brainerd & Reyna, 1993; Reyna & Kiernan, 1994, 1995). However, younger children reversal than older children because verbatim are less likely to display false~recognition memory ability continues to develop during childhood. As our brief review reveals, various phrases have been used to describe processes that involve resistance to false memories, beginning with verbatim~based rejection (e.g., Reyna & Kiernan (1994), negative dependency (e.g., Reyna & Kiernan, 1995), nonidentity

500

REYNAET AL.

judgments (e.g., Brainerd et al., 2001), recollection rejection (e.g., Brainerd & Reyna, 2002), false recognition reversal (Brainerd, Reyna, & Kneer, 1995), and, in studies with adults, recall-to-reject (e.g., Gallo, 2004; Rotella, Macmillan, & Van Tassel, 2000). The last phrase, recall-to-reject, resembles the others in that it involves retrieval of memories for experienced events. However, recall-to-reject assumes that the rejection process involves recall (which is more a description of a task rather than an underlying process) and that recall is a unitary process-that recall must involve retrieval of the verbatim trace rather than the gist trace (or, that distinguishing between these dual traces is not necessary). For example, recall-to-reject could logically refer to recalling gist and, thus, rejecting lures because they were not consistent with the gist of experienced events. The other phrases, in contrast, were all used in the context of fuzzy-trace theory and refer to the same process of retrieving a verbatim memory for experienced events in order to reject related items. Because this process involves verbatim memories, the properties of which are well known, we can predict the factors that will facilitate rejection of false memories. Among those factors is pictorial encoding (Brainerd & Reyna, 1993; Israel & Schacter, 1997). As we have discussed, pictures enhance verbatim memory performance relative to comparable words. Although pictures figure in research by Schacter et al. (1999), the use of the distinctiveness heuristic discussed by Schacter et al. should be distinguished from recollection rejection; the latter involves the retrieval of verbatim memories but not necessarily metacognitive strategies. The fact that young children exhibit recollection rejection is among the key pieces of evidence that recollection rejection is not identical with the distinctiveness heuristic. Manipulations, such as pictorial encoding, that improve recollection rejection of false memories do not merely enhance memories. Instead, they selectively enhance verbatim memories (e.g., for presented targets in the DRM paradigm). Other manipulations, such as engaging in deeper semantic processing, similarly improve memory for experienced events (e.g., presented targets), but they also increase acceptance of false memories (e.g., for critical lures) (Toglia et al., 1999). Therefore, dual representational assumptions that distinguish gist and verbatim memories are necessary to account for memory-enhancing manipulations that reduce false memories.

OVERVIEW: LEGAL AND SOCIAL IMPLICATIONS OF RESEARCH ON FALSE MEMORY

What does research suggest about the legal and social ramifications of false memory? Although it is now widely acknowledged that false memories can be implanted by actively suggesting misinformation to child witnesses, the threat posed by spontaneous false memories is less well known. Spontaneous false memories begin the moment people encounter and encode experiences. They occur when people experience meaningful stimuli, such as words, sentences, narratives, pictures, numbers, and events. Children and adults are more likely to falsely remember semantically consistent versions of events, compared with inconsistent, unrelated, or meaningless versions. Spontaneous false memories are not always entirely spontaneous (see Table 17.1 for a taxonomy of false

17.

FALSEMEMORY

SOl

TABLE 17.1 A Continuum of False Memories from Internal to External Suggestion: Illustrative Phenomena

1.

Autosuggestion (Bransford & Franks, 1971; Reyna & Kiernan, 1994) True Memory: The bird is in the cage. The cage is under the table. False Memory: The bird is under the table.

2.

Repeated Cuing (Ackerman, 1992; Reyna, 1996b; Roediger & McDermott, 1995) True Memory: nurse, sick, hospital, ill, physician, stethoscope, office, lawyer, patient, medicine ... False Memory: doctor

3.

Mere Memory Testing (Brainerd & Reyna, 1996; Payne et al., 1996) True Memory: dog, book, train, picture, color... False Memory: Question: Did the list contain "animal"? Answer: No On a later test or interview: Did the list contain "animal"? Answer: Yes

4.

Forced Confabulation (Reyna, Holliday, & Marche, 2002; Zaragoza, Payment, Ackil, Drivdahl, & Beck, 2001) True Memory: dog, book, train, picture, color... Forced Confabulation: Regardless of what you remember, go ahead and tell me the list, "animal, book, train, picture, color. .. " Answer: Okay, "animal, book, train, picture, color... " On a later test or interview: Now, be careful, only tell me what you really remember. Answer: Okay, "animal, book, train, picture, color... "

5.

External Suggestion (Loftus, Miller, & Burns, 1978) True Memory: dog, book, train, picture, color ... Misinformation Phase: How many words did you see on the list that contained the words "animal, book, train, picture, color. .. "? False Memory: On a later test or interview: Did the list contain "animal"? Answer: Yes

6.

Repeated External Suggestion (Zaragoza & Mitchell, 1996) True Memory: dog, book, train, picture, color. .. Misinformation Phase: How many words did you see on the list that contained the words "animal, book, train, picture, color. .. "? Second Misinformation Phase: The list that had the words "animal, book, train, picture, color... " was the hardest to remember, wasn't it? False Memory: On a later test or interview: Did the list contain "animal"? Answer: Yes

7.

Coercive Leading Questions (Bruck & Ceci, 1997) True Memory: Accused was asleep when baby died. Coercion: I'll let you go home just as soon as you tell me you shook the baby. I can keep you here until you tell me. Note: The examples given here do not necessarily reflect those used in the cited papers.

502

REYNAET AL.

memories that range from internally to externally generated; Reyna & Lloyd, 1997; Reyna, Holliday, & Marche, 2002). Seemingly neutral recall prompts (e.g., Tell me what happened when you got hurt.) or recognition probes (e.g., Did the man who hurt you have a beard?) can contaminate subsequent memory reports. This contamination operates for semantically consistent suggestions but is far less evident for unrelated material. Thus, it is possible for experts who are familiar with research to identify conditions under which witnesses are likely to report false memories. In addition, forensic interviews should minimize not only leading questions and suggestions, but also recall prompts and recognition probes that provide information for children to verify. If such prompts and probes are presented, the subsequent history of responses should be tracked. For example, an initial denial that nevertheless cues gist memories, followed by acceptance of gist-consistent suggestions, should be viewed as a potential instance of memory-testing-induced false memory. Research indicates that information provided to children, even indirectly through casual interactions (e.g., with family members) or in formal interviews, can taint memory reports if it reminds children of a particular gist or interpretation of events: Questions such as Did your step-Dad ever hit you? and Did he ever yell at you? cue the gist that the stepfather is a bad man and increase the probability that he will be remembered as doing a bad thing. This effect is enhanced if the child already harbors such a belief. Fuzzy-trace theory predicts the results that we have discussed, for example, that when questions cue related gist representations, those representations are more likely to be falsely remembered in subsequent interviews for both children (Brainerd & Reyna, 1996) and adults (Payne, Elie, Blackwell, & Neuschatz, 1996). Although activation, source monitoring, and other false-memory theories might be extended to cover these findings, there are other results that are more difficult to account for, such as memoryindependence effects. Moreover, there are patterns of acceptance and rejection of false memories that appear to require opponent process assumptions found in fuzzy-trace theory, specifically, verbatim and gist representations. However, a source-monitoring perspective facilitates considering memory for sources separate from memory for information derived from those sources (contrary to the source-monitoring framework in which source judgments derive from memory for events; Johnson et al., 1993). Furthermore, source-monitoring theories alone, or in combination with an activation metaphor, provide an important perspective on judgment processes that witnesses might use to discriminate internally generated and externally experienced events (Lindsay & Johnson, 2000; Roediger et al., 2001). For example, witnesses can be asked source discrimination questions, such as Do you remember thinking that he had a knife or did you see the knife? Did you see him hit your friend or did someone tell you that? However, given the potential for contamination from mere-memory testing effects, such choices should not be offered to children who have not already made related statements and should not be offered to children too young to understand complex sentence structures. Given that spontaneous false memories are endemic to processing meaningful stimuli, it would seem that they would be pervasive and inevitable. Although meaning-based false memories may be among the most common false memories, research has shown that there are factors-within the capabilities of young children-that can be used to

17.

FALSEMEMORY

503

resist false memories. Children's ability to reject false memories by retrieving verbatim memories of experience improves with age and is greater under conditions that foster verbatim memory, such as shorter intervals between events and memory tests. Both ver~ batim and gist memory abilities improve with age. Ironically, young children's failure to connect the gist across separate instances appears to explain their superior overall mem~ ory accuracy, compared with older children and adults, in many studies using the DRM explanation is supported by the shift with age paradigm. This failure~to~connect~the~gist in the nature of intrusions in false recall, which are not predominantly semantic for younger children as they are in adults. The simple manipulation of providing the con~ nected gist theme to children prior to list presentation appreciably increases the rate of false memories. These conditions would be expected to be replicated in forensically relevant contexts when children experience multiple, meaningfully related events (e.g., repeated episodes of sexual abuse). Young children would be expected to be less likely events into memory reports that did not occur, compared to interpolate gist~consistent with older children and adults. Note that older children and adults could have been vic~ timized repeatedly and yet misreport gist~consistent details. The probative status of such misreported details is therefore open to some question. In other words, errors that are gist~ consistent do not necessarily throw into doubt either the reliability of the witness or the truth of the remembered event. These results-that young children are less prone to certain meaning~related mem~ ory illusions-are echoed in other research indicating that biased social judgments that reflect prejudices and stereotypes are less likely in young children than in older ones (e.g., Davidson, 1995; Jacobs & Potenza, 1991). Prejudices and stereotypes, unfortu~ nately, develop as children grow older and become more acculturated. Thus, younger alterna~ children would be less able to distinguish actual experience from gist~consistent tives, compared with older children, and yet be less susceptible to false memories that require spontaneously connecting the gist across separate instances. In order to know which situation applies and to judge the accuracy of children's memory reports, it is im~ portant to identify factors that promote the accessibility of verbatim and gist memories, which include the age of the child witness, the time interval since the event, the ability of the child to understand the gist of the event, whether multiple related events oc~ curred, the number and nature of recall prompts and recognition probes since the event, and whether pictures or other verbatim prompts can be introduced (without contami~ nating memory). Although scientific research cannot resolve all uncertainties about the testimony of child witnesses and victims, it can reduce uncertainty and provide a prin~ cipled basis for best practices in gathering evidence from children.

REFERENCES Ackerman, B. P. (1992). The sources of children's source errors in judging causal inferences. Journal of Experimental Child Psychology, 54, 90-119. Ackerman, B. P. (1994). Children's source errors in referential communication. ]oumal of Experimental Child Psychology, 58, 432-464.

504

REYNAET AL.

Anderson, J. R., Budiu, R., & Reder, L. M. (2001). A theory of sentence memory as a general theory of memory. Journal of Memory and Language, 45, 33 7-367. Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge: Cambridge University Press. Bjorklund, D. F. (1987). How changes in knowledge base contribute to the development of children's memory. Developmental Review, 7, 93-130. Bjorklund, D. F., & Hock, H. H. (1982). Age differences in the temporal locus of memory organization in children's recall. Journal of Experimental Child Psychology, 347-362. Brainerd, C.]., & Gordon, L. L. (1994). Development of verbatim and gist memory for numbers. Developmental Psychology, 30, 163-177. Brainerd, C.]., Holliday, R. E., & Reyna, V. F. (2004). Behavioral measurement of remembering phenomenologies: So simple a child can do it. Child Development, 75, 505-522. Brainerd, C.]., Payne, D. G., Wright, R., &Reyna, V. F. (2003). Phantom recall. Journal of Memory and Language, 48, 445-467. Brainerd, C.]., & Reyna, V. F. (1993). Memory independence and memory interference in cognitive development. Psychological Review, 100, 42-67. Brainerd, C.]., & Reyna, V. F. (1996). Mere memory testing creates false memories in children. Developmental Psychology, 32, 467-476. Brainerd, C.]., & Reyna, V. F. (1998). When things that never happened are easier to remember than things that did. Psychological Science, 9, 484-489. Brainerd, C.]., & Reyna, V. F. (2002). Fuzzy-trace theory and false memory. Current Directions in Psychological Science, 11, 164-169. Brainerd, C.]., & Reyna, V. F. (2005). The science of false memory. New York: Oxford University Press. Brained, C.]., Reyna, V. F., & Forrest, T. ]. (2002). Are young children susceptible to the falsememory illusion? Child Development, 73, 1363-1377. Brainerd, C. J., Reyna, V. F., & Kneer, R. (1995). False-recognition reversal: When similarity is distinctive. Journal of Memory and Language, 34, 157-185. Brainerd, C. J., Reyna, V. F., & Poole, D. A. (2000). Fuzzy-trace theory and false memory: Memory theory in the courtroom. In D. F. Bjorklund (Ed.), False memory creation in children and adults (pp. 93-127). Mahwah, NJ: Lawrence Erlbaum Associates. Brainerd, C. J., Reyna, V. F., Wright, R., & Mojardin, A. H. (2003). Recollection rejection: False-memory editing in children and adults. Psychological Review, 110, 762-784. Brainerd, C. J., Wright, R., Reyna, V. F., & Mojardin, A. H. (2001). Conjoint recognition and phantom recollection. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27,307-327. Bransford, J.D., & Franks, J. J. (1971). The abstraction of linguistic ideas. Cognitive Psychology, 2, 331-380. Brown, A. L., Smiley, S. S., Day, J.D., Townsend, M.A. R., & Lawton, S.C. (1977). Intrusion of a thematic idea in children's comprehension and retention of stories. Child Development, 48, 1454-1466. Bruck, M., & Ceci, S. J. (1999). The suggestibility of children's memory. Annual Review of Psychology, 50, 419-439. Davidson, D. (1995). The representativeness heuristic and the conjunction fallacy effect in children's decision making. Merrill-Palmer Quarterly, 41,328-346. Deese, J. (1959). On the prediction of occurrence of certain verbal intrusions in free recall. Journal of Experimental Psychology, 58, 17-22. Forrest, T. J. (2002). Memory errors in elementary school children. Unpublished doctoral dissertation, University of Arizona, Tucson, AZ. Gallo, D. A. (2004). Using recall to reduce false recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30, 120-128.

17. FALSE MEMORY

505

Gallo, D. A., McDermott, K. B., Percer, J. M., & Roediger, H. L. III (2001). Modality effects in false recall and false recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 339-353. Ghetti, S., & Alexander, K. W. (2004). "If it happened, I would remember it": Strategic use of event memorability in the rejection of false autobiographical events. Child Development, 75, 542-561. Ghetti, S., Qin, J., & Goodman, G. S. (2002). False memories in children and adults: Age, distinctiveness, and subjective experience. Developmental Psychology, 38, 705-718. Holliday, R. E., Reyna, V. F., & Brainerd, C. J. (2006). Developmental changes in false memories following gist cues and verbatim repetition. Manuscript submitted for publication. Holliday, R. E., Reyna, V. F., & Hayes, B. K. (2002). Memory processes underlying misinformation effects in child witnesses. Developmental Review, 22, 37-77. Howe, M. L., Cicchetti, D., Toth, S. L., & Cerrito, B. M. (2004). True and false memories in maltreated children. Child Development, 75, 1404-1417. Israel, L., & Schacter, D. L. (1997). Pictorial encoding reduces false recognition of semantic associates. Psychonomic Bulletin & Review, 4, 577-581. Jacobs, J. E., & Potenza, M. (1991). The use of judgment heuristics to make social and object decisions: A developmental perspective. Child Development, 62, 166-178. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28. Johnson, J. W., & Scholnick, E. K. (1979). Does cognitive development predict semantic integration? Child Development, 50, 73-78. Karibian, D. (2003). Using fuzzy-trace theory to predict memory for word lists and sentences in children with learning disabilities. Unpublished doctoral dissertation, University of Arizona, Tucson, AZ. Liben, L. S., & Posnansky, C. J. (1977). Inferences on inference: Effects of age, transitive ability, memory load, and lexical factors. Child Development, 48, 1490-1497. Lim, P. L. (1993). Meaning versus verbatim memory in language processing: Deriving inferential, morphological, and metaphorical gist. Unpublished doctoral dissertation, University of Arizona, Tucson, AZ. Lindsay, D. S., & Johnson, M. K. (2000). False memories and the source monitoring framework: Reply to Reyna and Lloyd (1997). Learning and Individual Differences, 12, 145-161. Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4, 19-31. Mather, M., Henkel, L. A., & Johnson, M. K. (1997). Evaluating characteristics of false memories: Remember/know judgments and memory characteristics questionnaire compared. Memory and Cognition, 25,826-837. Metzger, R. L., Warren, A R., Price, J.D., Reed, A W., Shelton, J., & Williams, D. (2004). Do children "D!R-M" like adults? False memory production in children. Unpublished manuscript, University ofTennessee, Chattanooga. McDermott, K. B. ( 1996). The persistence of false memories in list recall. Journal of Memory and Language, 35, 212-230. Miller, M. B., & Wolford, G. L. (1999). Theoretical commentary: The role of criterion shift in false memory. Psychological Review, 106, 398-405. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice Hall. Paris, S. G., & Carter, A Y. (1973). Semantic and constructive aspects of sentence memory in children. Developmental Psychology, 9, 109-113. Paris, S. G., & Mahoney, G. J. (1974). Cognitive integration in children's memory for sentences and pictures. Child Development, 45, 633-643.

506

REYNAET AL.

Payne, D. G., Elie, C. J., Blackwell, J. M., & Neuschatz, J. S. (1996). Memory illusions: Recalling, recognizing, and recollecting events that never occurred. Journal of Memory and Language, 35, 261-285. Piaget, J., & lnhelder, B. (1973). Memory and intelligence. New York: Basic Books. Pipe, M. E., Gee, S., Wilson, J. C., & Egerton, J. M. (1999). Children's recall1 or 2 years after an event. Developmental Psychology, 35, 781-789. Poole, D. A., & White, L. T. (1991). Effects of question repetition on the eyewitness testimony of children and adults. Developmental Psychology, 27, 975-986. Prawatt, R. S., & Cancelli, A. (1976). Constructive memory in conserving and nonconserving first graders. Developmental Psychology, 12,47-50. Price, J.D., Metzger, R. L., Williams, D., Phelps, N. Z., & Phelps, A.M. (2001, April). Children produce as many false memories as adults (sometimes!). Poster presented at Society for Research in Child Development, Minneapolis, MN. Reed, A. W., Mangan, B. R., Warren, A. R., Price, J.D., & Metzger, R. L. (2003, April). Age differences in false memories: Now you see them, now you don't. Poster presented at the biennial meeting of the Society for Research in Child Development, Tampa, FL. Reyna, V. F. (1995). Interference effects in memory and reasoning: A fuzzy-trace theory analysis. In F. N. Dempster & C. J. Brainerd (Eds.), Interference and inhibition in cognition (pp. 29-61). New York: Academic Press. Reyna, V. F. (1998). Fuzzy-trace theory and false memory. In M. lntons-Peterson & D. Best, Eds. Memory distortions and their prevention (pp. 15-27). Mahwah, NJ: Lawrence Erlbaum Associates. Reyna, V. F. (2000). Fuzzy-trace theory and source monitoring: An evaluation of theory and false-memory data. Learning and Individual Differences, 12, 163-175. Reyna, V. F., & Brainerd, C. J. (1993). Fuzzy memory and mathematics in the classroom. In R. Logie & G. Davies (Eds.), Everyday memory. Amsterdam: North-Holland. Reyna, V. F., & Brainerd, C. J. (1995). Fuzzy-trace theory: An interim synthesis. Learning and Individual Differences, 7, 1-75. Reyna, V. F., & Brainerd, C. J. (1998). Fuzzy-trace theory and false memory: New frontiers. Journal of Experimental Child Psychology, 71, 194-209. Reyna, V. F., Holliday, R., & Marche, T. (2002). Explaining the development of false memories. Developmental Review, 22, 436--489. Reyna, V. F., & Kiernan, B. (1994). The development of gist versus verbatim memory in sentence recognition: Effects oflexical familiarity, semantic content, encoding instructions, and retention interval. Developmental Psychology, 30, 178-191. Reyna, V. F., & Kiernan, B. (1995). Children's memory and interpretation of psychological metaphors. Metaphor and Symbolic Activity, 10, 309-331. Reyna, V. F., & Lloyd, F. (1997). Theories offalse memory in children and adults. Learning and Individual Differences, 9, 95-123. Roberts, K. P. (2002). Children's ability to distinguish between memories from multiple sources: Implications for the quality and accuracy of eyewitness statements. Developmental Review, 22, 403--435. Robinson, K. J ., & Roediger, H. L., III (1997). Associative processes in false recall and false recognition. Psychological Science, 8, 231-23 7. Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented on lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21,803-814. Roediger, H. L., III, & McDermott, K. B. (1999). False alarms about false memory. Psychological Review, 106, 406--410. Roediger, H. L., III, Watson J. M., McDermott, K. B., & Gallo, D. A. (2001). Factors that determine false recall: A multiple regression analysis. Psychonomic Bulletin & Review, 8, 385-405.

17.

FALSEMEMORY

507

Rotello, C. M., Macmillan, N. A., & Van Tassel, G. (2000). Recall-to-reject in recognition: Evidence from ROC curves. Journal of Memory and Language, 43, 6 7-88. Sameroff, A. J., & Haith, M. M. (Eds.). (1996). The five to seven year shift: The age of reason and responsibility. Chicago: University of Chicago Press. Schacter, D. L., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger and older adults: The distinctiveness heuristic. Journal of Memory and Language, 40, 1-24. Small, M. Y., & Butterworth, J. (1981). Semantic integration and the development of memory for logical inferences. Child Development, 52, 732-735. Titcomb, A. L., & Reyna, V. F. (1995). Memory interference and misinformation effects. In F. N. Dempster & C. J. Brainerd (Eds.), Interference and inhibition in cognition (pp. 263-294). San Diego: Academic Press. Toglia, M.P., Neuschatz, J. S., & Goodwin, K. A. (1999). Recall accuracy and illusory memories: When more is less. Memory, 7, 233-256. Underwood, B. J. (1965). False recognition produced by implicit verbal responses. Journal of Experimental Psychology, 70, 122-129. Zaragoza, M.S., & Mitchell, K. J. (1996). Repeated exposure to suggestion and the creation of false memories. Psychological Science, 7, 294-300. Zaragoza, M.S., Payment, K. E., & Ackil, J. K. (2001). Interviewing witnesses: Forced confabulation and confirmatory feedback increase false memories. Psychological Science, 12 (6), 4 73-4 77.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

18 A Review of Factors Affecting Jurors' Decisions in Child Sexual Abuse Cases Bette L. Bottoms University of Illinois at Chicago

Jonathan M. Golding University of Kentucky

Maggie C. Stevenson Tisha R. A. Wiley University of Illinois at Chicago

John A. Yozwiak Stanford University School of Medicine

Child maltreatment is one of the most significant problems society faces today (Myers, 2004; Vieth, 2006). Each year in the United States, there are around 3 million reported cases, of which approximately 1 million are substantiated. An indeterminably large number of cases go undisclosed and unreported Oones & Finkelhor, 2001). Although child sexual abuse (CSA) constitutes only around 10% of all reported child maltreatment cases, it accounts for the majority of all sexual abuse cases handled by the American legal system. For example, 67% of all sexual assault cases reported to law enforcement agencies in the year 2000 were CSA cases (Snyder, 2000). CSA also accounts for the most trials in which children testify (Goodman, Quas, Bulkley, & Shapiro, 1999). A juror's task is never easy, but it is particularly difficult when a trial focuses on alleged CSA. Given the inherent privacy and secrecy of this crime, CSA cases often lack physical evidence and corroborating witnesses, and so jurors must base their decisions largely on the testimony of alleged victims (Myers, 1998; Pennsylvania v. Ritchie, 1987; Whitcomb, Shapiro, & Stellwagen, 1985). In fact, actual jurors in such cases say that a 509

510

BOITOMS ET AL.

child's testimony is the most important evidence (Myers, Redlich, Goodman, Prizmich, & Imwinkelried, 1999). Thus, although jury research in the field of psychology and law usually finds that hard evidence (i.e., "legal" evidence) is the most influential factor in jurors' decision making (Devine, Clayton, Dunford, Seying, & Pryce, 2001), there is little hard evidence in cases of CSA. Therefore, "extralegal" factors, such as individual juror biases or child characteristics, can influence decisions. Research on perceptions of children's testimony finds that adults are often not very skilled at discerning children's actual accuracy (Goodman, Batterman-Faunce, Schaff, & Kenney, 2002; Goodman, Bottoms, Herscovici, & Shaver, 1989; Leippe & Romanczyk, 1989; but see Leippe, Manion, & Romanczyk, 1992). If children's actual accuracy is not the primary determinant of jurors' decisions, then it is of great importance to identify the factors that are influential. Social scientists have heeded this need over the past 15 years, producing a large and growing body of research on jurors' perceptions of CSA. In this chapter, we provide the first comprehensive review of that literature, summarizing the host of legal and extralegal factors that researchers have investigated, including juror, victim, defendant, and trial factors. We illustrate the value of this field of research for psychologists and for professionals within the legal system who seek to understand how jurors react to CSA victims and how they reach their verdicts in CSA cases. In most studies reviewed in this chapter, researchers employed a mock trial methodology. That is, researchers create a detailed, plausible case, often based on facts from actual cases. Such cases are presented to mock jurors, that is, research participants who play the role of jurors for the study. These mock jurors are usually jury-eligible undergraduate students receiving course credit for participation, but they are sometimes community members paid for their time. Cases are presented via summarized written case summaries or trial transcripts, elaborately produced videotaped simulated trial excerpts, or even videotaped testimony of actual CSA victims. Mock jurors are asked to render a verdict and to provide other judgments (e.g., ratings of their confidence in the verdict, perceived credibility of witnesses, perceived victim and defendant responsibility for the abuse, etc.). Researchers typically measure jurors' individual verdict preferences, but in some studies, jurors also deliberate before rendering verdicts. Ideally, but not always, studies are conducted with as much ecological validity as is practically feasible, including pattern jury instructions and realistic case stimuli built upon admissible evidence and legally appropriate charges. Even so, the circumstances of mock trials do not duplicate the experiences of jurors in actual trials. Therefore, the artificiality in methods should be taken into account when seeking to generalize from laboratory studies to actual legal situations (Diamond, 1997; Weiten & Diamond, 1979). Still, a lack of some aspects of realism is not necessarily a marker of a nongeneralizable study. On the contrary, the experimental control offered by a laboratory setting is quite important in ensuring the scientific integrity of research. Mock trial methodology allows researchers to have the control necessary for varying certain factors (e.g., victim age) while keeping all other factors constant. Without this, it would be impossible to draw definitive conclusions about the effects of a particular factor. Mock trial methodology also allows for the collection of data from many participants rather than just a few, ensuring that results will generalize to a larger segment of the population. Such control

18. FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

511

would be impossible if researchers were restricted to studying actual cases decided by actual juries, which are unique in numerous idiosyncratic ways. We return to some of these methodological issues near the end of our chapter.

GENDER

Juror Gender

A growing number of studies have revealed robust effects of juror gender, which was one of the first factors to be studied in this field. Just as women have more negative reactions than men to rape (see Borgida & Brekke, 1985, for a review), sexual harassment (e.g., O'Donohue & O'Hare, 1997), and wife battering (e.g., Schuller & Hastings, 1996), women jurors are generally more pro-child victim and more prosecution oriented than men jurors in their reactions to CSA cases. Gender effects, however, are not always consistent or pervasive in terms of specific dependent measures; that is, sometimes gender differences emerge in verdicts (e.g., Castelli, Goodman, & Ghetti, in press; Crowley, O'Callahan, & Ball, 1994; Ross et al., 1994; but see, e.g., Tubb, Wood, & Hosch, 1999), but other times gender effects emerge only in more subtle judgments, such as perceived victim and defendant credibility or responsibility. Yet on average, women are often more likely than men to render a guilty verdict and/or to favor the prosecution in terms of victim and defendant credibility (believability) judgments, victim and witness responsibility judgments, etc. (Allen & Nightingale, 1997; Bottoms, 1993; Bottoms, Davis, & Epstein, 2004; Bottoms & Goodman, 1994, Experiments 2 and 3; Bottoms, Nysse-Carris, Harris, & Tyda, 2003; Crowley et al., 1994; Duggan et al., 1989; Gabora, Spanos, &Joab, 1993; Golding, Sanchez, & Sego, 1999; Golding, Stewart, Yozwiak, Djadali, & Sanchez, 2000; Haegerich & Bottoms, 2000; lsquith, Levine, & Scheiner, 1993; Kovera, Levy, Borgida, &Penrod, 1994; McCauley &Parker, 2001; Nightingale, 1993; Quas, Bottoms, Haegerich, & Nysse-Carris, 2002; Redlich, Myers, Goodman, & Qin, 2002; Ross et al., 1994; Scheiner, 1988; Schmidt & Brigham, 1996; Swim, Borgida, & McCoy, 1993; Yozwiak, Golding, & Marsil, 2004; see Schutte & Hosch, 1997, for a review). Such gender differences also manifest in some (Back & Lips, 1998; Broussard & Wagner, 1988; O'Donohue, Elliott, Nickerson, & Valentine, 1992; O'Donohue & O'Hare, 1997; Waterman & Foss-Goodman, 1984) but not all (O'Donohue, Smith, &Schewe, 1998; Maynard & Wiederman, 1997) studies in which participants do not pose as jurors, but instead read brief vignettes describing uncontested CSA situations and make judgments about the seriousness of the abuse and blameworthiness of the victim and perpetrator. The generalizability of these findings to actual cases is bolstered by the fact that gender differences have also been shown in studies employing arguably the most realism possible in an experimental setting. Orcutt et al. (2001; see also Goodman et al., 1998) used an elaborate mock trial paradigm involving jury-eligible community members as participants and trials that were held in an actual courtroom, with well-trained actors playing the parts of the judges and attorneys, and actual child witnesses (7- to 9-year-old boys or girls) providing live testimony about a real play event they had experienced

512

BOTTOMS ET AL.

3 weeks earlier at a university laboratory. The jurors' task in each trial was to determine whether the defendant (a "babysitter") had made a movie of himself, placing stickers on a child's exposed body parts (e.g., stomach) during the previous play event. This event was not CSA, of course, but the event was perhaps as close a simulation of CSA as is ethically possible. Results revealed that prior to deliberation, men jurors were more likely than women to endorse pro-defense attitudes, have more empathy for the defendant, view the child as more suggestible, and were less likely to convict (although they did not differ from women in terms of child empathy or ratings of the child's accuracy). Women are also more pro-victim than men when sexual abuse accusations are made in the context of especially controversial cases, such as when the victim is a teenaged juvenile on trial for murdering her alleged sexual abuser (Haegerich & Bottoms, 2000), when claims arise in the context of bizarre satanic ritual abuse allegations (Bottoms, Diviak, & Davis, 1997), and when allegations are based on recovered memories (Clark & Nightingale, 1997; ForsterLee, Horwitz, Ho, ForsterLee, & McGovern, 1999; Golding, Sanchez, et al., 1999; Golding, Sego, Sanchez, & Hasemann, 1995). In nearly all of the studies we have just reviewed, jurors rendered individual case judgments without deliberating as a group. Findings from the few studies that included deliberations suggest that despite pervasive gender differences in individual jurors' judgments, gender differences might dissipate following deliberations (Goodman et al., 1998). For example, in studies by Crowley et al. (1994) and lsquith et al. (1993), women were somewhat more likely to change their votes from pre-deliberation to post-deliberation than were men, becoming more lenient after deliberation than before deliberation. Bottoms, Schmidt, and Epstein (1998) tested two theories that could explain why this verdict shift occurred for women: (a) male normative influence, or the idea that during deliberations men influence women to change their verdicts more than women influence men to change theirs; and (b) the leniency bias, or the tendency for factions of pro-acquittal jurors (in this case, generally men) to be more influential than equal-sized factions of pro-conviction jurors (in this case generally women). This leniency bias is observed across other types of cases and is thought to be prompted by the reasonable doubt standard of proof-it is easier to convince fellow jurors to have a doubt about guilt than to convince them there is no doubt in a case (MacCoun & Kerr, 1988). In the study by Bottoms et al., undergraduate jurors viewed a videotaped simulated CSA trial in which a 13-year-old girl claimed that her stepfather had fondled her (from Duggan et al., 1989). Women were more pro-prosecution than men before, but not after, group deliberation. The researchers constructed four types of juries based on gender and pre-deliberation verdict preference: (a) half male jurors favoring guilt and half female jurors favoring acquittal, (b) half male favoring acquittal and half female favoring guilt, (c) all male split equally on verdict, and (d) all female split equally on verdict. There were no postdeliberation differences in verdicts among the groups, suggesting that the shift was due to a leniency bias rather than disproportionate male influence during deliberation, which would have resulted in relatively more convictions in group a and more acquittals in group b. In support of this theory, coding of deliberation transcripts revealed that men and women jurors did not differ in terms of persuasive behaviors (e.g., the number of

18. FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

513

words spoken, turns taken, interruptions, etc.), unlike findings from studies of deliberations in other kinds of criminal cases (Hastie, Penrod, & Pennington, 1983).

Explaining Juror Gender Differences Why do men and women, on average, have different reactions to CSA? Gender is, of course, merely a proxy variable for psychological variables and life experiences that happen to cluster or correlate with societal constructions of gender. That is, men and women differ in attitudes, feelings, and attributional tendencies relevant to the crime of CSA, and those differences translate into different case judgments. For example, women are more likely than men to make stable and internal attributions of causality and blame to CSA offenders (Beling, Hudson, & Ward, 2001). This might explain some of the differences in case judgments such as verdicts, but this theory has not been tested directly. Furthermore, experimental and survey research has shown that compared with men, women react more negatively to CSA, perceiving it to be more serious (e.g., Attias & Goodwin, 1985; Corder & Whiteside, 1988; Finlayson & Koocher, 1991; Kelly & Tarran, 1984; Kovera, Borgida, Gresham, Swim, & Gray, 1993). Men are also more likely to endorse what might be termed "CSA myths," such as, ''l\. child who does not display signs of distress probably has not been a victim of sexual assault" (Morison & Greene, 1992), just as men are more likely to endorse adult rape myths, such as, "If a woman doesn't physically fight back, you can't really say that it was rape" (Burt & Albin, 1981). In fact, similar gender differences in case judgments have been found in reactions to adult rape (Brekke & Borgida, 1988). Deitz, Blackwell, Daley, and Bentley (1982) found that these differences are explained by gender differences in empathy for victims and attitudes toward women, rape, and sexuality. Using the adult rape literature as a guide, Bottoms (1993) theorized that gender differences in response to CSA cases might be explained by gender differences in empathy for child victims and in attitudes toward CSA, children's general believability, women, and sexuality. She provided preliminary evidence from an undergraduate mock juror sample that (a) persons who have empathy for children, who consider children to be generally believable, who react negatively to adult/child sexuality, who hold pro-feminist views, and who hold conservative attitudes toward sex are most likely to make pro-child victim and anti-defendant judgments in CSA cases; (b) more women fit that profile than men; and (c) gender differences in case judgments are explained by the individual differences in child victim empathy and in attitudes, especially empathy and attitudes toward children's general believability. Bottoms, Wiley, and Stevenson (2005) recently replicated these findings with a community sample of mock jurors. It will be important to determine other attitudinal constructs that help explain gender differences, and to understand how men and women come to have such distinct beliefs and reactions to CSA. This could be linked to women's disproportionately higher rate of sexual victimization, which probably leads to a heightened tendency to identify with other victims and to be personally concerned with issues of abuse (Bottoms, 1993; Haegerich & Bottoms, 2000). This gender difference might also be linked to the internalization of societal gender

514

BOTTOMS ET AL.

roles, which demand that women be more caring, empathic, and child-oriented than men (e.g., Gilligan, 1982; Barnett & Sinisi, 1990, but see Maynard & Wiederman, 1997).

Interactions of Juror Gender with Victim and Defendant Gender On the one hand, a number of studies have shown that the effects of juror gender prevail regardless of whether the child victim is a boy or girl (Back & Lips, 1998; Bottoms & Goodman, 1994, Experiment 2; Crowley et al., 1994; Haegerich & Bottoms, 2000; lsquith et al., 1993). On the other hand, several investigations have found statistical interactions of juror gender with victim and/or defendant gender. First, some studies found that women jurors favor women perpetrators and girl victims, and men jurors favor men perpetrators and boy victims (e.g., ForsterLee et al., 1999). Second, some studies showed that on some case judgments, men jurors, but not women, are affected by victim gender (e.g., Broussard & Wagner, 1988; Clark &Nightingale, 1997; Quas et al., 2002). Third, there are studies in which women's judgments were unaffected by manipulations of defendant gender, but the same was not true for men (O'Donohue et al., 1998; Quas et al., 2002; Smith et al., 1997). For example, in the Quas et al. (2002) study, men (but not women) jurors were less likely to convict women defendants when the victim was a boy compared with when the victim was a girl. It is possible that jurors-at least men-find sexual encounters between women and boys more permissible than other forms of abuse, a point we return to below.

Victim Gender Historically, sexual victimization was considered a crime perpetrated on women and girls (Sepler, 1990). It has been argued that society has generally failed to recognize boys as true victims who are as harmed as girls by sexual abuse (Eisenberg, Owens, & Dewey, 1987; Finkelhor, 1984). Some also argue that boys are thought to be more able than girls to resist abuse and, therefore, to be at least partially responsible for abuse they endure (Rogers & Terry, 1984). It is not surprising, then, that boys have been deterred from reporting abuse (Bolton, Morris, & MacEachron, 1989; Urquiza & Keating, 1990), a fact vividly exemplified by the current Catholic Church abuse scandals, in which most victims were boys who often failed to report their abuse for decades (Bottoms, Shaver, Goodman, & Qin, 1995; John Jay College Research Team, 2004). Since the 1980s, and especially since the Catholic Church scandal, however, there has been a growing awareness that boys are sexually victimized, and efforts have been made to encourage them to report these crimes (e.g., Finkelhor, 1983; Sepler, 1990). Nonetheless, negative attitudes toward boy victims might still exist (Hunter, 1990). For instance, Donnelly and Kenyon (1996) reported that some law enforcement officers and rape crisis providers endorse myths that boys are infrequently sexually abused and that boy victims are weak, effeminate, and might even secretly enjoy sexual assault. Research on the influence of victim gender has only partially supported the idea of jurors being more skeptical of CSA cases involving boys than cases involving girls.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

515

Haegerich and Bottoms (2000) observed that mock jurors were more willing to believe abuse allegations made by a 15-year-old girl than those made by a boy in a case in which the victim claimed to have killed her/his father as a self-defensive reaction to years of sexual abuse. In another study, however, undergraduate jurors were less likely to vote guilty when a woman as opposed to a man 23-year-old victim claimed to have recovered a formerly repressed memory of childhood abuse (Clark & Nightingale, 1997). Perhaps jurors are more impressed with adult men than women who break their silence in the face of potential stigmatization. In a study of the perceived accuracy of children portrayed in a forensic investigation (Goodman et al., 2002), undergraduates found boys to be more accurate and believable than girls, but social workers' judgments were uninfluenced by victim gender. Even so, mock trial experiments have revealed few if any significant effects of victim gender on CSA case judgments, including studies by Bottoms and Goodman (1994, Experiment 2) involving 6-, 10-, and 14-year-old victims; by Crowley et al. (1994) involving 6-, 9-, and 12-year-old victims; and by Isquith et al. (1993) involving 13- and 15-year-old victims. Similarly, victim gender failed to influence attributions of responsibility for abuse in non-mock trial vignette studies (e.g., Back & Lips, 1998; O'Donohue et al., 1992; Waterman & Foss-Goodman, 1984). These results are consistent with Myers and colleagues' (1999) finding that most actual jurors (61% of 248 jurors from 42 real trials involving 5- to 18-year-old alleged victims) said that victim gender was not important in their decision-making. (Yet those jurors participated in cases in which 89% of victims were girls. Since the prototypical victim is a girl, jurors in cases involving girls probably would not think about issues of victim gender.) The results are also generally consistent with findings from studies of factors that influence actual prosecutorial decisions (Cross, DeVos, & Whitcomb, 1994; Brewer, Rowe, & Brewer, 1997; but see MacMurray, 1989, who found that cases involving boys were more likely to be prosecuted than cases involving girls).

Defendant Gender There are relatively few studies investigating the effect of defendant gender on jurors' decisions in sexual abuse cases, probably because people assume that women rarely perpetrate child sexual abuse. This assumption, however, might be based more on social attitudes and lack of reporting than the actual prevalence of woman-perpetrated sexual assault (Bolton et al., 1989; Finkelhor & Russell, 1984; Finkelhor, Hotaling, Lewis, & Smith, 1990; Finkelhor, Williams, & Bums, 1988; Hetherton, 1999). Some research suggests that allegations of CSA involving women defendants are less likely to result in formallegal action than allegations involving men defendants (Williams & Farrell, 1990), and that women offenders (with boy victims) are less likely to be prosecuted than men offenders (Finkelhor, 1983; but see Brewer et al., 1997; Cross et al., 1994). Hetherton and Beardsall (1998) also reported that, among men and women police officers and social workers, registration and incarceration of sex offenders were rated as more appropriate for men than for women. Freeman, Levine, and Doueck (1996), however, found no association between perpetrator gender and social services caseworker activity, amount

516

BOTTOMS ET AL.

of services received, and likelihood of substantiation in actual child maltreatment cases. This might be explained by the fact that Freeman et al. included physical abuse and neglect cases in their sample. That is, the differences in perceptions of men and women CSA perpetrators might not generalize to other types of abuse cases. With regard to perceptions in court, research has shown that sexual assault perpetrated by a woman tends to be perceived less negatively than assault by a man. A vignette study by Finkelhor and Redfield (1984), for example, revealed that jurors considered adult-child sexual relations to be less abusive when perpetrated by women than by men (but see Broussard & Wagner, 1988). Also, O'Donohue et al. (1998; see also Smith et al., 1997) presented undergraduates and educators with vignettes in which a child was abused by either a man (the child's uncle) or a woman (the child's aunt). Men jurors, less so than women, perceived the child as less credible when the perpetrator was a woman versus a man. O'Donohue et al. suggested plausibility as one possible explanation; that is, women-perpetrated abuse is less plausible than man-perpetrated abuse. Finally, some research has found that victim gender interacts with defendant gender to influence jurors' judgments. Specifically, just as perpetrators of same-gender sexual abuse are viewed as more psychologically disturbed than perpetrators of opposite-gender sexual abuse (Drugge, 1992), same-gender CSA is sometimes rated as more abusive than opposite-gender interactions, particularly when victims are teenagers. For example, in a vignette study by Maynard and Wiederman (1997), more blame was assigned to perpetrators of same-gender abuse than opposite-gender abuse when a victim was 15 years old (but not when the victim was 7 years old). Similarly, Dollar, Perry, Fromuth, and Holt (2004) found that people had more negative reactions to the abuse of a 16-yearold boy when it was perpetrated by a man versus a woman. Theoretically, such reactions could be driven by anti-gay biases or by social norms that encourage turning a blind eye toward sexual relations between adolescent boys and older women (Quas et al., 2002). Regarding the latter, Trivelpiece (1990; see also Hetherton, 1999; Hunter, 1990) suggests that some Americans believe that sexual relations with older women are permissible for adolescent boys. Even some researchers have suggested that some woman/boy sexual relations are not necessarily abusive (Rind, Tromovitch, & Bauserman, 1998), a claim rejected by many psychologists (Ondersma et al., 2001). Illustrating that such assumptions manifest in mock trial cases, Quas et al. (2002; see also Smith et al., 1997; Waterman & Foss-Goodman, 1984) found that men jurors were most lenient toward women defendants accused of abusing boys as compared with other victim/defendant gender combinations.

VICTIM, DEFENDANT, AND JUROR RACE

Effects associated with race and ethnicity are increasingly becoming of interest in the field of psychology and law. In the United States, minorities have suffered many injustices relative to the white majority population. There are widely documented examples of racial and ethnic stereotypes and unfavorable evaluations of minorities in various legally and non-legally relevant situations (e.g., Brigham, 1972; Devine, 1989; Landrine, 1985; Niemann, Jennings, Rozelle, Baxster, & Sullivan, 1994; Weinberg & Williams,

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

517

1988; Wilson, 1996; Wyatt, 1982, 1985, 1990). It is important to identify conditions under which race influences justice, so that social scientists can discover ways that soci~ ety and the legal system can anticipate and avoid miscarriages of justice. With regard to CSA, it is important to keep in mind that African American children are sexually abused as often as white children (Finkelhor & Baron, 1986; Wyatt, 1985), if not more often (Epstein & Bottoms, 1998; Tzeng & Schwarzin, 1990; but see Linholm & Willey, 1986). Children of Hispanic origin are also abused at rates comparable to those found for white children (Epstein & Bottoms; Linholm & Willey, 1986). Negative stereotypes about Hispanic Americans (Wilson, 1996; Deaux & Ethier, 1998) and African Americans include a sexual component: that, compared with whites, minorities are more sexually promiscuous (Powell, Wyatt, & Bass, 1983; Weinberg & Williams, 1988; Wyatt, 1982), more sexually experienced (Wyatt, 1982), and learn about and engage in pre~ marital sex at earlier ages (Wyatt, 1982). These beliefs might cause jurors (especially racist jurors) to take sexual abuse reports made by minority children less seriously than reports made by white children. In addition, stereotyped beliefs that African American defendants are more promiscuous, sexually perverse, hostile, and criminal~like than whites (e.g., Devine, 1989) might lead to greater perceptions of minority defendants' guilt, although this could interact with victim race. Finally, juror race might also affect decisions in CSA cases because of in~group favoritism (e.g., Tajfel, 1982). Despite the importance of understanding effects of race in legal settings, very little research has been conducted on race and CSA. In a sample of actual CSA cases in day~ care settings, Williams and Farrell (1990) found that perpetrators accused of assaulting white girls were arrested and convicted more frequently than those accused of assault~ ing African American girls. Yet two other studies of actual CSA cases found no relation of victim ethnicity and prosecutorial decisions (Brewer et al., 1997; Cross et al., 1994). There have been two laboratory studies investigating race and CSA case decisions. Foley and Chamblin (1982) asked African American and white mock jurors to listen to an audiotaped description of a trial in which an African American or white man was ac~ cused of sexual battery of an 11~year~old African American or white child. Victim and defendant race interacted to influence guilt judgments, but the interaction depended on the jurors' race. White jurors assigned higher guilt ratings to an African American defen~ dant accused of assaulting a white child than to defendants in the other conditions. African American jurors assigned higher guilt ratings when the defendant's race matched the victim's, regardless of the specific defendant~victim race combination. All race ef~ fects, however, were eliminated when mock jurors' perceptions of victims' and defen~ dants' socioeconomic status were statistically controlled. Bottoms et al. (2004) used a mock~trial, written case summary methodology to ex~ amine the influence of race on undergraduate jurors' case decisions. In Experiment 1, undergraduate mock jurors judged 12~ and 16~year~old African American and Hispanic girl victims to be more responsible than white victims for their sexual abuse allegedly perpetrated by a white teacher, but there were no race effects for guilt or credibility judg~ ments. Also, jurors who were of the same race as the victim rated the defendant to be significantly less credible and had more confidence in his guilt than did jurors who were of a different race (though there were no such effects on victim credibility or victim or defendant responsibility judgments). In Experiment 2, white jurors made judgments

518

BOTTOMS ET AL.

about a case in which both a defendant's and a 12-year-old victim's race was varied (African American or white). Jurors assigned more guilt to defendants in cases involving victims and defendants of the same race relative to cases involving different races, which the authors attributed to laypeople's general belief that different-race CSA is less common and thus less plausible than same-race abuse (supported by their findings in Experiment 3). Finally, contrary to the researchers' expectations, there was no significant tendency for high-prejudiced mock jurors to make more racially biased judgments than low-prejudiced mock jurors.

OTHER JUROR FACTORS

We have reviewed many studies investigating the influence of juror individual differences on CSA judgments, including juror gender, juror race, and juror attitudes. There are likely to be a host of other juror individual difference factors linked to case judgments, only some of which have been investigated to date. For example, Vidmar (1997) suggested that in CSA cases, some jurors hold a "generic prejudice" that arises from negative reactions to this type of crime. According to Vidmar, generic prejudice inhibits jurors' ability to be impartial and biases their judgments against any defendant accused of a particular crime. Vidmar (1997) examined transcripts and outcomes of challenge for cause (similar to the American process of voir dire) in 25 Canadian trials involving CSA. On average, 36% of 849 jurors indicated that they would not be able to be impartial in their decisions. Bottoms (1993) hypothesized that experiential factors, such as a personal history of abuse, would also bias case judgments. She found preliminary evidence that, compared with nonvictims, people who have been sexually victimized themselves as adults or as children hold stronger pro-victim attitudes and make more pro-victim case judgments in CSA cases. Similarly, Waterman and Foss-Goodman (1984) found that participants who said they had been molested attributed less fault to a victim in a vignette CSA study. Although one might predict a relation between judgments and parenthood or experience with children, Goodman et al. (1998) found no relations between judgments and demographic characteristics such as parental status and degree of contact jurors had with children. In another study, however, Goodman et al. (2002) found differences in case judgments between undergraduates and social workers, and between those who had and had not been abused. Specifically, when asked to watch a videotaped mock forensic interview of a child discussing a past event, undergraduates were more likely than social workers to believe that abuse had occurred, and among social workers, those who were themselves victimized or who were close to a victim were more likely than others to think abuse had occurred.

OTHER VICTIM CHARACTERISTICS

Myers and colleagues' (1999) survey of actual jurors revealed that 53% stated that they considered child-victim characteristics to be the most important evidence in the case

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

519

(20% said defendant factors were). Our review has illustrated the potential influence of victim race and gender on CSA case judgments, but other factors have also received much attention in the literature, particularly victim age.

Victim Age Victim age is a powerful predictor of juror judgments through its influence on perceptions of competence and trustworthiness (central components of credibility). Building on theories first asserted by Goodman, Golding, and Haith (1984; Goodman, Golding, Helgeson, Haith, & Michelli, 1987), Bottoms and Goodman (1994) theorized that, compared with older children and adults, young children are generally perceived to be low on competence (cognitive ability, resistance to suggestion), yet high on trustworthiness (honesty, sincerity, innocence) (see also Leippe, Brigham, Cousins, & Romanczyk, 1989; Leippe & Romanczyk, 1987; Yarmey &Jones, 1983). They suggested that the relative salience of these dimensions in a particular case would determine a child witness's credibility. In cases highlighting witness competence, such as cases in which children are bystander witnesses and must remember many details accurately, younger children are perceived as less credible than older victims (e.g., Goodman et al., 1984, 1987; Leippe & Romanczyk, 198 7, Experiment 2). In cases highlighting sincerity and honesty as opposed to cognitive competence, with CSA cases being a central example, child witnesses are generally viewed as no less credible than adults (e.g., Ross, Miller, & Moran, 1987). Bottoms and Goodman predicted that jurors would attribute credibility to young CSA victims because jurors view these victims as honest and trustworthy but also sexually naive, lacking the knowledge and cognitive capacity to fabricate sexual encounters. Also, young children would be considered unlikely to possess capacities that lower adult rape victims' credibility, such as previous sexual experience, sexual provocativeness, and ability to give meaningful consent (e.g., Borgida & Brekke, 1985; Brownmiller, 1975). Increasing age would be accompanied by decreasing perceptions of sexual naivete and honesty, increasing perceptions of cognitive capacity, and, in turn, increased suspicions regarding a child's capacity and propensity to lie about sex acts and decreased perceived credibility. Many studies varying victim age support this theory, finding that in cases involving younger CSA victims (i.e., younger than about 12 or 13 years; Bottoms & Goodman, 1994; Schiener, 1988), jurors perceive child witnesses more favorably and/or are more likely to vote guilty than in cases involving older child or adult victims (e.g., Bottoms & Goodman, 1994; Bottoms et al., 2004; Duggan et al., 1989; Gabora et al., 1993; Golding, Alexander, et al., 1999; Golding, Sanchez, &Sego, 1997; Golding, Sanchez, &Sego, 1999; Goodman et al., 2002; Nightingale, 1993; but see Allen & Nightingale, 1997; Crowley et al., 1994; Golding, Fryman, Marsil, & Yozwiak, 2003; and McCauley &Parker, 2001). To illustrate, Nightingale (1993) asked undergraduates to read a trial summary involving a girl being assaulted by a stranger on her way home from school. Convictions and perpetrator blame for the incident were more likely when the case involved a 6-yearold as compared with a 14-year-old. Vignette studies reveal similar results (Back & Lips, 1998; Collings & Payne, 1991; Finkelhor, 1984; Finkelhor & Redfield, 1984;Maynard & Wiederman, 1997; Waterman & Foss-Goodman, 1984). For example, Back and Lips

520

BOITOMS ET AL.

(1998) found that a 13-year-old child was seen as more responsible than a 6-year-old child for his/her abuse perpetrated by a male neighbor. Furthermore, the parents of the younger child were seen as more responsible than the parents of the older child. What impact does victim age have in actual cases? Regarding prosecutorial decisions, although Bradshaw and Marks (1990) found no relation of prosecution tendencies and victim age in 121 Texas cases, Finkelhor (1983) found that prosecutions were most likely in cases with children aged 7 to 12 years. Others, however, have found that cases involving older child victims are more likely to be prosecuted than cases involving younger child victims (Brewer et al., 1997; Cross et al., 1994; Gray, 1993). This pattern would result if prosecuting attorneys believe that older children will make better witnesses than younger children (Cross et al., 1994). In fact, some very young children are too afraid and traumatized to testify without accommodations, as discussed by Goodman and colleagues (this volume). Yet when the crime is CSA, and the witness can testify, this strategy might be wrong, for all the reasons we have outlined above. Myers and colleagues (1999) found few significant differences in real jurors' reported perceptions of the credibility of child witnesses as a function of child age, except that as age increased from 5 to 18 years, perceptions of the child's suggestibility decreased and perceptions of an adult hearsay witness who corroborated the child's testimony increased. Should this finding give us pause in generalizing laboratory findings to the courtroom? Not necessarily. Myers and colleagues (1999) suggested that once a case makes it to trial, the evidence might be so compelling and the children so well prepared as to negate the age/credibility relationship that typically emerges in laboratory experiments. In fact, in their study, 92% of the trials resulted in conviction. Studies of real cases could mask effects predicted by theories about perceived competence and sexual naivete; that is, attorneys could be reluctant to pursue cases in which the victim is older and circumstances could suggest that she/he had a role in the relationship. Certainly, prosecutors rarely take to trial cases in which older teenagers are said to be romantically involved with perpetrators, even when the perpetrators are much older and state laws define the relationships as abuse.

Victim Disability Children with intellectual disabilities are especially likely to be sexually abused (Friedrich & Boriskin, 1978; Westcott &Jones, 1999), yet prosecutors are sometimes reluctant to take these cases to court, fearing that jurors will be biased against disabled witnesses. In contrast, Bottoms et al. (2003) theorized that jurors would find them even more credible than children of average intelligence because, like young children, they would be deemed trustworthy and honest, yet lacking in the level of cognitive competence necessary for fabricating false charges (see also Yozwiak, 2003). To test this theory, Bottoms et al. (2003) conducted a mock trial study in which undergraduates watched videotaped excerpts from an actual trial. When the 16-year-old sexual assault victim was portrayed as "mildly mentally retarded" instead of as "having average intelligence," jurors were more likely to vote guilty and more likely to consider the victim to be believable, honest, and incapable of fabricating the sexual abuse accusation. In addition, jurors who had

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

521

more liberal views toward persons with disabilities were more likely than others to make pro-prosecution judgments on measures of guilt as well as victim understanding of, belief in, and likelihood of having fabricated the charge. Similarly, in non-mock-trial vignette studies, Podell, Kastner, and Kastner (1994) and Podell, Kastner, and Kastner (1996) found that CSA victims labeled as mentally retarded were not disadvantaged compared with a nondisabled victim, at least in terms of perceived responsibility for a forced sexual encounter. On the basis of these results, Bottoms and colleagues argue that dismissing the abuse allegations of children and adults with disabilities and assuming that they cannot be effective witnesses is an injustice.

Victim Behavior and Demeanor In deciding CSA cases, jurors might use aspects of a child's behavior during the abuse incident or during courtroom testimony as a cue. Regarding child behavior, two vignette studies have shown that jurors (especially men) attribute more responsibility for abuse to victims who do not actively resist a sexual abuse encounter compared with those who do resist (Broussard & Wagner, 1988; Collings & Payne, 1991). More research has investigated the impact of child demeanor during testimony, probably because it is so widely assumed that emotional demeanor of witnesses is a useful cue to determining their credibility (in fact, a statement to this effect is even a part of some jury instructions). Myers et al. (1999) found that actual jurors agree with this assumption and that their verdicts were correlated with the degree to which they perceived the child's facial expressions, gestures and movements, eye contact, nervousness, and manner of speaking to be important. Furthermore, 16% of those jurors indicated that crying was the most important emotion displayed by the child, 14% indicated that nervousness was most important, and 14% indicated that embarrassment was most important. These are all emotions that conventional wisdom suggests are appropriate or even expected for victimized children. There have been several experimental demonstrations of the importance of a child's emotional reactions. Regan and Baker (1998) found that child demeanor at the moment when the child encounters the defendant in court affects the perceived credibility of the child's testimony. When a child was portrayed in this case summary as having cried (versus staying calm) upon encountering the defendant, undergraduate mock jurors judged the child to be more honest, credible, accurate, and reliable. Jurors were also more likely to believe that the child who cried had been sexually abused and were more likely to find the defendant guilty. Perhaps, however, victims need to show some, but not too much, emotion. In an extension of the work by Regan and Baker, Golding et al. (2003) manipulated the child witness's crying behavior by showing undergraduate jurors a drawing of a calm, teary, or hysterically crying child. Participants who rated the defendant as relatively guilty were more likely to convict if they saw the victim depicted as teary compared with calm or hysterical. Although jurors might expect an abused child to be upset when testifying, in reality, by the time actual child victims testify about their abuse, much time has passed and many children have had time to heal emotionally. Many have also participated in pretrial court

522

BOTTOMS ET AL.

preparation programs designed to help them cope with the stress of testifying. It is an unfortunate irony, then, that programs designed to aid children in giving their testimony might well have the effect of reducing their credibility in the eyes of the jury. Kovera and her colleagues found that expert testimony explaining these facts can help remedy this problem. Specifically, Kovera, Gresham, Borgida, Gray, and Regan (1997) showed undergraduate jurors a videotaped trial simulation in which an 8-year-old girl accused her stepfather of sexual assault. She appeared to be either prepared and calm or unprepared and emotional. There was either no expert testimony or an expert who (a) described common symptoms displayed by CSA victims, but did not specifically relate his testimony to the victim ("standard testimony"); (b) described common symptoms displayed by CSA victims and specifically discussed the ways in which the victim's symptoms matched those of a typical CSA victim ("concrete testimony"); or (c) summarized the research findings after providing the standard testimony ("repetitive standard testimony"). Overall, testimony matched to the specific victim improved the child's perceived credibility and increased defendant conviction rates compared with simulated trials in which no expert testimony was provided. When the child appeared prepared and the expert witness gave standard testimony, conviction rates increased. When the testimony was linked to the specific victim and the child was unprepared, conviction rates increased. That is, the concrete testimony appears to have sensitized jurors to expect a CSA victim to act a certain way. Thus, the behaviors of the unprepared child were perceived to be more consistent with the concrete testimony than the behaviors of the prepared child. For the prepared child, this perceived inconsistency seems to have decreased guilty verdicts.

CHARACTERISTICS OF ABUSE ALLEGATIONS

Repression and Delayed Reporting Characteristics of the abuse allegations or the context of reporting can affect jurors' perceptions of child abuse cases. For example, researchers have studied jurors' reactions to repressed memory allegations, a topic that has had a high profile in recent years (Lindsay & Read, 1995). In these studies, disclosures are typically made by adults who delayed reporting their childhood abuse for years, either because they kept it a secret or because they repressed, then recovered the memories. In the first studies of this kind, Loftus (1993) and Loftus, Weingardt, and Hoffman (1993) found that jurors (especially men in the latter study) were more skeptical of victims who claimed to have repressed their memories as compared with victims who always remembered their abuse but kept it a secret (see also ForsterLee et al., 1999; Key, Warren, & Ross, 1996). This skepticism might reflect skeptical coverage of these cases in the media and popular press books (e.g., Pendergrast, 1995). Golding et al. (1995; Golding, Sego, & Sanchez, 1999; Stewart, Whiteside, & Golding, 2000) extended this research by including a control condition in which the victim reported the CSA immediately (see also Golding, Warren, & Ross, 1997). In this study, there was no difference in the believability of the alleged victims who re-

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

523

pressed versus kept the abuse secret (but see Clark & Nightingale, 1997), but both were less believable than the child who testified immediately. Golding, Sanchez, et al. (1999) investigated the impact of length of reporting delay in a study in which mock jurors read a trial summary in which a girl was allegedly abused at the age of 6 years, but re, ported it at age 7, 26, or 45 years. The longer the reporting delay (regardless of repres, sion), the less believable the victim was perceived to be. In a study by Bottoms et al. (1997), undergraduate mock jurors rendered judgments about a case involving childhood sexual abuse allegations made by either a 5,year,old child or a 30,year-old adult survivor who had delayed in reporting abuse that happened when she was 5 years old. Allegations of repression were not involved, but the presence or absence of bizarre satanic ritual abuse allegations (like those in the McMartin Day Care case) was varied. Delay did not affect jurors' judgments. Although jurors were sig, nificantly less likely to believe the satanic case details than other details of the cases, they were just as likely to vote guilty and to believe the victim in satanic as in non,satanic cases. That is, jurors appeared to set aside their skepticism about satanic ritual details, making judgments about the core issue of CSA based on their perceptions of the non, satanic case details about the sexual abuse. Furthermore, there were marked individual differences in decisions: When the allegations involved satanic ritual abuse, religious jurors were more likely than less religious jurors to believe the victim.

Disclosure Characteristics The ways in which children disclose abuse vary (Pipe, Lamb, Orbach, & Cedarborg, in press), and disclosure or report characteristics might influence jurors' perceptions of child credibility. Yozwiak et al. (2004) explored this issue. Jury-eligible community members read a case summary in which a 6-year,old girl was picked up from school by a male neighbor, who then allegedly drove her to a wooden area and forced her to have sexual intercourse with him. The child provided either full and complete disclosure of the de, tails of the incident during two interviews, or she was incomplete during a first interview but later included additional details of the event during a subsequent interview. Ju, rors were most likely to believe the victim and vote guilty when there was full disclosure during both interviews. This is particularly interesting, given that inconsistency sometimes marks children's true disclosures of abuse (Pipe et al., in press).

l)'pe of Abuse Unfortunately, sexual abuse is not the only form of child maltreatment. In fact, as we noted earlier, it accounts for only about 10% of all child abuse reports in the United States. Other forms of abuse are far more common, including neglect (the most common form of child maltreatment), physical abuse, and psychological or emotional abuse. Yet, they receive far less attention in many domains of research, including this one. Of all trials involving child testimony, sexual abuse cases are the most common, but physical

524

BOTTOMS ET AL.

abuse cases are the second most common, and the third is domestic abuse cases (Goodman et al., 1999). We know of only two vignette studies, and no mock-jury studies, investigating adults' reactions to other forms of child abuse. Muller, Caldwell, and Hunter (1993) asked undergraduates to read vignettes describing an interaction between a parent and an 8-year-old child that resulted in the parent harming the child. Men were more likely than women to blame the child for the abuse, illustrating that gender differences generalize to perceptions of other forms of abuse. Thus, although gender differences might be partly explained by attitudes related to sexuality and CSA (Bottoms, 1993), they might also be explained by more generalized attitudes about children and abuse that have not yet been investigated. In another study, Dukes and Kean (1989) examined participants' perceptions of cases of child psychological abuse, physical abuse, and neglect. The degree to which the victim provoked the abuse and the nature of the abuse were varied. Whereas physical abuse was perceived as equally severe for both very young (3-year-old) and older (10-year-old) victims, psychological abuse was perceived as more severe for the older victim, and neglect was perceived as more severe for the younger victim. In cases involving psychological and physical abuse, but not neglect, the child was considered to be more at fault when he or she directly "provoked" the abuse (see also Broussard & Wagner, 1988). In addition, the abuse was seen as less severe when the older child directly "provoked" the abuse compared with when the younger child did so.

COURTROOM AND TRIAL FACTORS Courtroom Accommodations (Testimony Medium)

In some circumstances, courts allow special accommodations for child witnesses, including preparation (training) of children for the courtroom experience and allowing children's in-court testimony to be replaced by testimony given via closed-circuit television (CCTV), videotaped pretrial forensic interviews, or a hearsay witness (which might be less likely now in the wake of Crawford v. Washington, 2004). On the one hand, it is argued that such accommodations are needed to protect children, because testifying in court is a stressful and traumatic experience (for discussion, see Bottoms, Repucci, Nysse-Carris, & Tweed, 2002; Goodman et al., 1991, this volume). On the other hand, however, courts assume that face-to-face confrontation compels witnesses to tell the truth and promotes accurate fact-finding by jurors who will have nonverbal cues to use to assess witness credibility (Coy v. Iowa, 1988; Maryland v. Craig, 1990). Less expensive innovations (e.g., preparation of the child for testifying, the presence of a supportive adult) are preferred by prosecutors and used more often than innovations such as videotaped depositions, CCTV testimony, and barriers that separate the child from the defendant's view, which prosecutors fear will hurt rather than help their case in the eyes of the jury (Goodman et al., 1999). Are their intuitions correct? Although research addressing this

18. FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

525

question has been reviewed in depth by Kovera and Borgida (1996) and McAuliff and Kovera (2002), our chapter would be incomplete without a brief summary of this literature. In general, findings from this work are somewhat mixed (e.g., Davies & Noon, 1993; Ross et al., 1994) but seem to be converging to indicate that although children's actual accuracy might sometimes increase when they testify via alternative modes like CCTV, prosecutors might be right: special accommodations sometimes negatively influence jurors' perceptions of children's competence. For example, in the first published study addressing this issue (Swim et al., 1993), undergraduate mock jurors viewed a videotaped trial simulation based on an actual trial in which an 8-year-old girl claimed to have been sexually abused by her stepfather. The child testified in court or via videotape. Many effects of testimony medium were only marginally significant and inconsistent, with some suggesting that the videotaped testimony produced more guilty verdicts, whereas other analyses suggested the opposite. The generalizability of these results is limited by the fact that both conditions were delivered via videotape to the participant jurors. That is, participants watched a videotape of a child testifying in court or a videotape of a videotape of a child's testimony. The same problem is inherent in studies by Eaton, Ball, and O'Callaghan (2001) and Lindsay, Ross, Lea, and Carr (1995). In a study that is more generalizable given its elaborately conducted mock-trial methodology, Goodman et al. (1998) explored the accuracy and perceived credibility of child testimony by first having 5- to 6-year-old and 8- to 9-year-old children participate in a play session with a male confederate, and then, about 2 weeks later, testify about the play session in an actual courthouse in front of a community-member mock jury. Compared with children who testified live in court, children who testified via CCTV were viewed by jurors as less credible, despite the fact that they were actually more accurate. A similar study by Orcutt, Goodman, and Tobey (2001) included conditions in which children were asked to lie about their experiences. Jurors were no more likely to detect children's deceit when testimony was delivered in open court versus via CCTY, but jurors were less likely to convict when testimony was delivered via CCTV versus live in the courtroom. Furthermore, children who testified via CCTV were perceived to be less accurate, less attractive and intelligent, less honest, and more likely to have made up their story than children who testified in the courtroom. Several studies have addressed another procedural innovation meant to reduce the need for children to testify in court: a special hearsay exception for child witnesses, which allows witnesses to testify regarding what the child said to them about abuse allegations. Some of these studies have shown that hearsay testimony is believed. For example, in a study by Golding, Sanchez, and Sego (1997; see also Golding, Alexander, & Stewart, 1999; Warren, Nunez, Keeney, Buck, & Smith, 2002), undergraduate mock jurors considered a written trial summary in which a 6- or 14-year-old girl accused her uncle of sexual abuse. The victim either did or did not testify, and, in addition, hearsay testimony was or was not given. Jurors were more likely to believe the victim's claims when there was a hearsay witness, regardless of whether the victim testified.

526

BOTTOMS ET AL.

In contrast, other studies have shown that hearsay testimony might not be as believable as testimony from the child him- or herself. Redlich et al. (2002; see also Ross, Lindsay, & Marsil, 1999) compared two forms of hearsay: the videotape of an actual CSA victim's forensic interview versus the testimony of a police officer who repeated what the child said during the forensic interview. There was no direct effect of testimony format on undergraduate jurors' perceptions of defendant guilt or victim credibility, but participants in the videotape condition were more likely than participants who heard the police officer to believe that the child had fully disclosed. This, in turn, led to increased perceptions of child credibility and increased ratings of defendant guilt.

Attorney Tactics Defense and prosecuting attorneys play an important role in determining how cases are presented. Can attorneys shape jurors' reactions to cases? Consider the jury that is exposed to prosecutors who are adept at creating a very compelling empathic atmosphere toward a child, or to defense attorneys who ask jurors to place themselves in the shoes of a defendant, as trial lawyers have instinctively done for years (Archer, Foushee, Davis, & Aderman, 1979; Belli, 1956; Hamlin, 1985). Haegerich and Bottoms (2000) investigated the impact of attorney-induced empathy on jurors' verdicts in a case involving a teenager being tried as an adult for murdering her father. The child defendant (whose gender was varied) claimed she had acted in self-defense because the father had tried to sexually abuse her. Empathy for the victim/defendant was manipulated in the case transcript via the defense attorney's opening and closing statements to the jury. In an empathy-prime condition, the attorney asked jurors to take the perspective of the child defendant and think of how they would be thinking and feeling if they were her. Compared with jurors who received no such instructions, jurors primed with empathy were more lenient in guilt judgments, considered the defendant to be less responsible for the murder, and were more likely to accept the defendant's claim ofCSA as a mitigating factor in the killing. Compared with men, women mock jurors were more lenient toward the child defendant overall and were more affected by the empathy induction. (Similarly, compared with men jurors, women jurors are less likely to find battered women who kill their abusive husbands to be responsible and guilty for the crime [Schuller & Hastings, 1996; Schuller, Smith, & Olson, 1994].) Thus, through opening and closing arguments, attorneys can successfully induce juror empathy for a teenaged child victim of sexual assault and, in turn, influence verdicts.

OTHER EVIDENTIARY AND CASE FACTORS

We began our chapter by noting that jury research generally finds that the strength of legal evidence presented in a case trumps all other factors in influencing jurors' verdicts (Devine et al., 2001). Because legal evidence is usually lacking in CSA cases, however, we have focused on the impact of other factors on juror decisions. Even so, some sexual

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

527

abuse cases do involve hard evidence, and it is important to understand the impact of various forms of such legal evidence on decisions.

corroborating Medical Evidence Bradshaw and Marks (1990) analyzed archival data from 360 actual prosecution files for CSA cases. Of the factors studied, the presence of medical evidence was the strongest predictor of case outcomes, nearly doubling the likelihood of prosecution and conviction (perpetrator confession was also an important predictor). Golding et al. (2000) experimentally investigated the impact of the presence or absence of DNA evidence on case judgments. Undergraduate mock jurors read a case summary in which a 6-year-old girl accused her neighbor of forcing her to have sexual intercourse with him. DNA evidence increased both jurors' positive perceptions of the child's credibility and the likelihood of conviction. We know of no other studies investigating the impact of this factor.

corroborating Eyewitness Testimony For much of the twentieth century in many countries, corroboration was often required before a child could testify in sexual assault cases. Despite relaxation of this requirement, attorneys might still be hesitant to bring cases to trial when there is little physical evidence, when there are no corroborating witnesses, or when those corroborating witnesses are other children. In fact, Brewer et al. (1997) found that the strongest predictor of prosecutorial decisions in a sample of actual cases was the number of child victims involved: Multiple victim cases were more likely to be pursued by prosecutors than cases involving only one victim. Furthermore, Myers et al. (1999) found that actual jurors in child abuse cases frequently cited corroboration as important in their decision making. Although corroboration augments pro-prosecution judgments in cases involving adult witnesses (e.g., Leippe, 1985), only a few studies have considered corroboration in cases involving child witnesses. Bottoms and Goodman (1994, Experiment 3; see also Nightingale, 1993) provided jurors with the actual videotaped testimony of either one or two boys testifying about sexual abuse by a teacher. Corroboration of a 9-year-old victim's testimony by a 14-year-old's independent testimony led to increased credibility for the 9-year-old, but the corroboration did not affect guilt judgments. This was especially true when the defendant was portrayed favorably, as an upstanding citizen with no prior record. When the defendant was portrayed unfavorably (having had similar complaints made against him), however, the child was perceived as very credible, regardless of corroboration. The authors concluded that although corroboration from another child witness could positively affect a child's perceived credibility, it does not guarantee a conviction in a sexual assault case. They cautioned, however, that the 14-year-old corroborating witness in their case was generally less credible than the younger victim, perhaps because of his age. Had he been more credible himself, verdicts might also have been influenced by the corroboration. In fact, Duggan et al. (1989) found that corroborated testimony resulted in more guilty verdicts and higher ratings of a child victim's

528

BOTTOMS ET AL.

memory abilities. In their study, guilty verdicts were rendered most often when corroboration was provided by a 9-year-old in comparison with an adult witness.

Defendant Criminal History Under some circumstances, information about a defendant's past criminal acts may be admitted at trial (Federal Rules of Evidence 404b; Kaplan, 1985). In cases involving adult witnesses, such information causes jurors to be more conviction prone (Clary & Shaffer, 1985; Kalven & Zeisel, 1966) and discredits much of a defendant's testimony (Shaffer, 1985), especially when past crimes are similar to the crime in question (Sealy & Cornish, 1973). Negative character testimony is especially biasing if other forms of evidence are lacking or unconvincing (Kaplan, 1985), as is often true in CSA cases. Logically, prosecutors in CSA cases should seek to admit defendants' past records of similar crimes, which jurors might take as evidence of enduring sexual preference. In fact, Goodman et al. (1999) found that prosecutors believe that evidence of a defendant's previous sexual acts increases guilty verdicts. Testing this assumption in a mocktrial study, Bottoms and Goodman (1994, Experiment 3) found that an unfavorable defendant characterization including such past acts was damaging to the defense, positively influencing jurors' perceptions of the victims and increasing the likelihood of guilty verdicts. The authors pointed out that disclosure of even unfounded past acts or other negative character details might be damaging to a defendant for a number of reasons. Once past acts are revealed, negative information about a defendant is more likely to be expected, becomes more salient, and therefore is more likely to be believed (Hamilton & Fallot, 1974). Negative information might also influence judgments by coloring all other evidence in a case (Penrod & Cutler, 1987) and by increasing the likelihood that jurors will infer that the defendant committed the crime in question because it is consistent with his past behavior and might indicate a stable "child molesting disposition" (Kelley, 1967; McArthur, 1972).

Expert Testimony Experts can testify in CSA trials about a variety of topics (see Kovera & Borgida, 1996, for a review). Research on this issue has typically focused on expert testimony presented as part of the prosecution's case (i.e., in support of the child victim). For example, Gabora et al. (1993; see also Crowley et al., 1994) investigated the impact of expert testimony about general psychological issues relating to CSA, such as delayed reporting. The expert discussed these issues with or without making specific reference to the case in question. Both expert conditions led mock jurors to become less accepting of misconceptions regarding CSA than mock jurors who heard no expert testimony. In addition, mock jurors presented with expert testimony referring to the actual case rendered more guilty verdicts than those who heard this testimony without such reference. Kovera et al. (1994) also studied the influence of expert testimony on judgments. Specifically, an expert testified that she believed a child witness was credible based on

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

529

(a) the content of therapy sessions with the child, (b) her opinion that the child's be, haviors and emotions matched a "syndrome" of characteristics typical of CSA victims, or (c) the child's use of an anatomically detailed doll during an interview. Overall, the presence of any expert's testimony led to increased conviction rates compared with no expert testimony, and among the three types of expert testimony, the syndromal testi, mony was rated as least credible. Additional research by Kovera and Borgida (1996) examined how expert testimony can affect jurors' expectations of a CSA victim (e.g., assumptions such as a nervous child in court must be lying). In this study, jurors watched a simulated videotaped CSA trial in which the child acted less prepared (e.g., the child looked down, fidgeted) or more prepared (e.g., the child made eye contact, sat still). In addition, the jurors heard no expert testimony, the syndromal testimony discussed above, or testimony in which the expert linked psychological research with trial evidence (i.e., the hypothetical con, clition). Finally, in both expert conditions the expert explained that a CSA victim's fear, guilt, and lack of knowledge about sexuality often cause the child to appear confused and hesitant when describing sexual abuse. Thus, a less prepared child's behavior in the courtroom was congruent with the expert's description of typical CSA victims' behav, ior. Results showed that jurors were more likely to make pro,prosecution judgments in the hypothetical condition when the child was less prepared than when he or she was more prepared. Kovera and Borgida argued that in the hypothetical condition, jurors used the concrete link in the expert's testimony to understand the child's less,prepared courtroom behavior. Conversely, jurors were more likely to make pro,prosecution judg, ments in the syndromal condition when the child was more prepared than when he or she was less prepared. Jurors appeared to use this testimony to bolster the testimony of the child who was seen as more credible from the start. With regard to the impact of defense experts in CSA cases, there has been only minimal research. For example, Stewart et al. (2000) examined expert testimony in re, pressed memory CSA cases using experts for both the plaintiff and defense. Unlike in earlier studies (Griffith, Libkuman, & Poole, 1998; Sugarman & Boney,McCoy, 1997; see also Buck & Warren, in preparation), the experts in the study by Stewart et al. re, ceived direct examination and cross,examination. In Experiment 1, there were two ex, pert conditions involving repressed memories of CSA. In one condition, experts for the plaintiff and defense testified about general memory processes and repression. In the other condition, these experts described the memory techniques used by the plaintiff's psychotherapist. Neither expert condition influenced decisions for the plaintiff, al, though when the expert described general memory processes (as opposed to memory techniques), the plaintiff was believed more. The lack of expert influence led to a sec, ond experiment (using only the plaintiff and defense experts describing memory tech, niques), in which the impact of each expert was examined after the direct examination and cross,examination of each witness. Results generally supported Stewart and col, leagues' prediction that although each expert had an influence on jurors during direct examination, the influence of each was canceled out during their respective cross, examinations.

530

BOTTOMS ET AL.

Method of Child Interviewing Finally, the way in which a child is questioned, either in court or during pretrial interviews, might affect how credible jurors perceive their testimony to be. In a study by Castelli et al. (in press) using a written case summary, a 4- or 7-year-old child accused a teacher of sexual abuse. When the child (especially the 4-year-old) was interviewed in a highly suggestive rather than a less suggestive manner, mock jurors were less likely to convict the defendant and less likely to perceive the child as credible. Similar findings were revealed in a study by Tubb et al. (1999). Undergraduate mock jurors read a trial transcript based on a real case in which a 9-year-old boy was sexually abused by his uncle. The child was depicted as being questioned in one of three ways: suggestively (e.g., "Has anybody touched you on your private parts?"), directly (e.g., "Where were you at?"), or in an open-ended manner ("Start at the beginning and tell me what happened."). Mock jurors were less likely to convict defendants when the child had been questioned in a suggestive manner as opposed to a direct or open-ended approach. Suggestive interviewing also led to lower perceived child accuracy, confidence, credibility, honesty, knowledge, persuasiveness, competence, and intelligence. Finally, Bottoms, Quas, and Davis (in press) reported preliminary findings from a study by Bottoms, Rudnicki, and Nysse-Carris (2004) indicating that socially supportive interviewing techniques, which promote children's accuracy and lower their suggestibility (e.g., Davis & Bottoms, 2002), have the potential to disadvantage child witnesses in terms of their perceived credibility. Undergraduates watched videotaped portions of a mock forensic interview in which 7- to 8-year-old children were interviewed in either a warm, supportive manner, or a more neutral (even cold) manner. Participants rated children interviewed in a nonsupportive manner as accurate more often than children interviewed in a supportive manner, even though children were actually more accurate in the supportive than in the nonsupportive condition. Replication is clearly needed, but this initial research provides an important step toward understanding how interview style can affect the perceptions that adults form when evaluating children's testimony.

METHODOLOGICAL CONSIDERATIONS

In the laboratory, one must always sacrifice some realism for experimental control. As a result, jury studies cannot fully duplicate an actual trial or the experience of serving on a real jury (Diamond, 1997; Weiten & Diamond, 1979). In many of the studies we reviewed, the mock jurors were not representative of community members who would actually serve as jurors, they were not exposed to live witnesses or as much testimony as they would be in an actual trial, they knew their judgments would not have a direct impact on a defendant's life, and they rendered individual verdicts without deliberating about the case as a group. To increase the generalizability of results in this field of research, researchers should try to address such concerns as much as possible and strive for increased ecological validity and mundane realism in their methods.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

531

Even so, differences in actual and laboratory conditions are not cause for dismissing experimental findings. First, consider the criticism that although many studies we re~ viewed used community samples or diverse undergraduate samples, others used college freshmen, a group that is usually young and of relatively high SES, and who therefore might not make judgments similar to community members. In fact, however, a recent meta~analysis by Bornstein (1999) revealed few differences in the mock~trial judgments of undergraduates and community members across many types of cases (see also Cutler, Penrod, & Dexter, 1990; Lind & Walker, 1979). Similar results were revealed by the three studies we know of that compared undergraduates and community members' reac~ tions to CSA. Specifically, Crowley et al. (1994) examined differences in the reactions of Australian undergraduate and community mock jurors to a videotaped simulation of an incest trial. They found no differences between undergraduate and community mock jurors in verdict preference, beliefs about the child's memory capacity, and beliefs about the child's ability to resist suggestion. In an unpublished dissertation by lsquith (1988) replicating the Duggan et al. (1989) studies, although some judgments differed, undergraduates and community members rendered similar verdicts. Finally, Bottoms, Wiley, and Stevenson (2005) found few differences between undergraduate and com~ munity member mock jurors' reactions to written CSA case summaries (those used in the Bottoms, 1993, studies). Thus, evidence thus far points to few key differences be~ tween community and undergraduate samples. Second, the realism of materials is also a methodological concern of importance. Although a few studies employed videotapes of actual child victim testimony (Bottoms & Goodman, 1994; Bottoms, Nysse~Carris, Harris, & Tyda, 2003) or elaborate videotaped trial simulations (Goodman et al., 1998; Orcutt et al., 2001), jurors usually simply read a written transcript. Although no one could argue that the more realistic the method, the more generalizable the product, research thus far reveals few differences between mock jurors' decisions in studies using written materials versus more elaborate videotaped tes~ timony (Scheiner, 1988; see also Goodman et al., 1984, 1987, for comparisons of mate~ rials in a non~ sexual~ abuse child witness case). Also, using videotaped trial stimuli to ex~ amine the effects of some variables such as victim age or defendant race could have That is, actors of various ages or races could differ in aspects other than costly trade~offs. strictly those attributes (e.g., attractiveness). Using written trial summaries or trial tran~ scripts allows for the experimental control necessary to draw cause~and~effect conclu~ sions about such variables, before moving on to more elaborately staged (and therefore research to pursue promising patterns of results. very expensive and time~consuming) Finally, another methodological issue of importance is that in many studies we re~ viewed, mock jurors did not deliberate before reaching a verdict, as would a real jury. Group discussion could cause jurors' extralegal biases to be publicly identified and atten~ uated (Shaw & Skolnick, 1995), or, through a process of attitude polarization (e.g., Mas~ covici & Zavalloni, 1969; Myers & Bishop, 1970), deliberation might actually strengthen those biases. Even so, research suggests that jury verdicts are often representative of the individual jurors' verdicts prior to deliberation, with deliberation only bringing forth what most jurors favor before discussion (Haegerich & Bottoms, 2004; Kalven & Ziesel, 1966; Sandys & Dillehay, 1995).

532

BOTTOMS ET AL.

PRACTICAL APPLICATIONS OF THE RESEARCH

Should the results of the literature we have reviewed be applied in our legal system? Yes. As noted by Goodman, Levine, and Melton (1992), "Psychology bears a social responsibility to provide the best available evidence on important questions of legal policy whenever it can do so (APA, 1991, Principal F). It should proceed with caution, but it should not be disabled by a requirement for perfect evidence" (p. 249). When we can be reasonably sure of a finding, after examining the study's methods and perhaps after finding converging evidence, researchers have a duty to bring it to the attention of professionals "in the trenches": social workers, attorneys, judges, legislators, the public, and others (Grisso & Melton, 1987; Reppucci, 1985). We feel that we have partly attended to this duty by writing this chapter, which we hope will reach a number of professionals who could benefit from the research. But to what end? What are the specific implications of this literature for policy and law? Attorneys and professionals from trial consulting firms express interest in research on juror individual differences, because they would like to use the findings during voir dire to identify the "best" jurors for their side in particular cases. The advisability of using the work we have reviewed for what is known as scientific or strategic jury selection is debatable (Davis & Sobus, 2004; Strier, 1999). Aside from ethical dilemmas associated with the unequal availability of jury selection services to all parties in cases due to cost, there are other problems: research reveals trends across a large number of experimental participants, rather than differences that can be pinpointed to one particular prospective juror. Consider the robust finding that women are more likely than men to be pro-prosecution in CSA cases. A significant difference in an experiment could reflect the fact that, say, 60% of women compared with only 40% of men voted pro-prosecution. Practically, then, there is no guarantee that any one particular woman will vote pro-prosecution. Therefore, picking a juror based solely on gender, which is of course not even legal (Batson v. Kentucky, 1986), might only produce "practical gains" after a large number of trials. Perhaps we paint too bleak a picture. Obviously, the research teaches us that the odds of any woman being pro-prosecution are better than the odds of any man being pro-prosecution. Also, consider research investigating the link between attitudes and judgments. To the extent that research comes closer to identifying attitudes that can be measured and shown to predict the behavior of jurors, and to the extent that trial attorneys can use these same attitude measures during voir dire, more accurate predictability might become a reality. For example, some psychologists have advocated screening jurors in rape cases based on pretrial biases such as an inability to empathize with victims (Deitz et al., 1982). Similarly, information from the line of research we have reviewed might be useful during voir dire in identifying individual prospective jurors holding biases that might jeopardize impartial jury composition in CSA cases. Importantly, we refer here to jurors who might be particularly biased against believing defendants as well as those who might be biased against believing child witnesses. Benefits of ensuring fairer juries by the exclusion of or education of biased jurors include the protection

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

533

of children from revictimization by unconvicted perpetrators, as well as the protection of innocent defendants from unjust conviction. This literature has other important implications. For example, the discovery of factors that influence case judgments could indicate the need for education of jurors or judges through expert testimony, or modification of jury instructions to include warnings about the potential influence of individual biases. Jurors might be educated about their own biases and cautioned against approaching trial evidence from a biased perspective. Jurors might be warned, for example, that teenage victims are sometimes considered responsible for sexual abuse because of perceived sexual maturity, but that this should not affect their judgments. That is, a juror should determine if sexual contact occurred, not whether consent was granted, because children are considered legally incapable of giving consent to sexual activity. In fact, Isquith et al. (1993, Experiment 1) examined the impact of such judicial instructions. Mock jurors read a vignette describing the sexual abuse of a 13-year-old girl. When jurors were instructed that children of that age legally cannot provide consent and that any sexual contact is the responsibility of the adult, jurors were significantly more likely to convict than if they were not given these instructions. These results were not replicated in a second study using a videotaped trial simulation of the same case even though perceptions of the child's responsibility were related to verdicts when jurors were not given these instructions. Understanding the factors that lead to tendencies in juror judgments and perceptions can help investigators, prosecutors, and defense attorneys think in a more informed manner about the evidence needed to build a case that will appeal to a jury. For example, if months or years have passed between when a child is abused and the time he or she testifies in court, the child might not appear visibly upset when recounting the story of her abuse. Knowing from research that this might lead some jurors to doubt the child's credibility in the absence of remedying expert testimony (Kovera et al., 1997), a prosecutor could obtain an expert witness to instruct jurors about children's expected emotional reactions and how people's assessments of credibility are influenced by emotional displays. As another example, if unfavorable information about a defendant's past activities will be admitted in a case, research (Bottoms & Goodman, 1994) suggests that the defense attorney might want to prepare an argument or obtain an expert to warn jurors that such evidence can influence jurors' assessments of a defendant's credibility. We could describe many other practical implications of the studies we have reviewed, but instead, we hope these examples illustrate how legal professionals themselves can find practical benefits in the research, benefits that will help ensure fair trials and justice for all parties involved in child sexual abuse cases.

CONCLUSIONS AND FUTURE DIRECTIONS

We have summarized a large body of research addressing many factors that influence jurors' perceptions of children's credibility in CSA cases. For many of the factors we reviewed, more research is needed, and we have noted some specific areas of need. But

534

BOTTOMS ET AL.

many other influential factors are likely to exist and should be the focus of future research in this growing subfield within the larger field of psychology and law. We believe it would be interesting, for example, to determine how CSA case decisions are affected by (a) the relationship between victims and perpetrators (e.g., trusted family members versus strangers), (b) contextual factors such as whether incest allegations arise during custody disputes between parents, (c) jurors' beliefs about children's suggestibility or cognitive abilities (especially if there is expert testimony about these factors), (d) defense tactics including the use of research findings to shape his or her presentation of case facts or choice of expert witnesses, and (e) defendant factors such as intellectual disability or mental illness. We also hope researchers will conduct research on perceptions of other forms of child maltreatment such as physical abuse and neglect and that researchers will investigate not only jurors' reactions to these cases, but also the reactions of other adults, such as social service professionals and police officers. These are just a few of our own thoughts-but there are myriad possibilities. For researchers who wish to conduct research in this field, we offer our own observation that partnerships with legal professionals provide an invaluable source of excellent ideas for studies of practical and theoretical importance. In conclusion, understanding how people perceive child victims and make decisions in child sexual abuse cases is important if courtrooms are to be places of fairness and justice. Researchers have discovered a great deal about factors that influence jurors' perceptions and judgments, and the future trajectory of this line of research appears quite promising. Continued attention to conducting scientifically sound research with ecologically valid techniques will ensure that future research on these issues will yield generalizable results of importance to law and policy.

REFERENCES Allen, L.A., & Nightingale, N. N. (1997). Gender differences in perception and verdict in relation to uncorroborated testimony by a child victim. Journal of Offender Rehabilitation, 24, 101-116. Archer, R. L., Foushee, H., Davis, M. H., & Aderman, D. (1979). Emotional empathy in a courtroom simulation: A person-situation interaction. Journal of Applied Social Psychology, 9, 275-291. Attias, R., & Goodwin, J. (1985). Knowledge and management strategies in incest cases: A survey of physicians, psychologists and family counselors. Child Abuse and Neglect, 9, 527-533. Back, S., & Lips, H. M. (1998). Child sexual abuse: Victim age, victim gender and observer gender as factors contributing to attributions of responsibility. Child Abuse & Neglect, 22, 1239-1252. Barnett, M.A., & Sinisi, C. S. (1990). The initial validation of a liking of children scale. Journal of Personality Assessment, 55, 161-167. Batson v. Kentucky, 476 US 79 (1986). Belli, M. (1956). Ready for the plaintiff. New York: Holt. Beling, J., Hudson, S. M., & Ward, T. (2001). Female and male undergraduate attributions for sexual offending against children. Journal of Child Sexual Abuse, 10, 61-82. Bolton, J. F. G., Morris, L.A., & MacEachron, A. E. (1989). Males at risk: The other side of child sexual abuse. Newbury Park, CA: Sage.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

535

Borgida, E., & Brekke, N. (1985). Psycholegal research on rape trials. In A. Burgess (Ed.), Rape and sexual assault: A research handbook (pp. 313-34 2). New York: Garland. Bornstein, B. H. (1999). The ecological validity of jury simulations: Is the jury still out? Law and Human Behavior, 23, 75-91. Bottoms, B. L. (1993). Individual differences in perceptions of child sexual assault victims. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving testimony (pp. 229-261). New York: Guilford. Bottoms, B. L., Davis, S. L., & Epstein, M. A. (2004). Effects of victim and defendant race on jurors' decisions in child sexual abuse cases. Journal of Applied Social Psychology, 34, 1-33. Bottoms, B. L., Diviak, K. R., & Davis, S. L. (1997). Jurors' reactions to ritual abuse allegations. Child Abuse and Neglect, 21, 845-859. Bottoms, B. L., & Goodman, G. S. (1994). Perceptions of children's credibility in sexual assault cases. Journal of Applied Social Psychology, 24, 702-732. Bottoms, B. L., Nysse-Carris, K., Harris, L., & Tyda, K. (2003). Juror's perceptions of adolescent sexual abuse victims who have intellectual disabilities. Law & Human Behavior, 27, 205-227. Bottoms, B. L., Quas, J. A., & Davis, S. L. (in press). The influence of interviewer-provided social support on children's suggestibility, memory, and disclosures. In M. E. Pipe, M. Lamb, Y. Orbach, & A. C. Cedarborg (Eds.), Child sexual abuse: Disclosure, delay and denial. Mahwah, NJ: Lawrence Erlbaum Associates. Bottoms, B. L., Reppucci, N.D., Tweed, J., &Nysse-Carris, K. (2002). Children, psychology, and law: Reflections on past and future contributions to science and policy. In J. R. P. Ogloff & R. Roesch (Eds.), Taking psychology and law into the twenty-first century (pp. 61-117). Kluwer Academic/Plenum. Bottoms, B. L., Rudnicki, A. G., & Nysse-Carris, K. L. (2004, Mar.). The influence of socially supportive interviewing on adults' perceptions of children's report accuracy. Presentation at the biennial meeting of the American Psychology/Law Society, Scottsdale, AZ. Bottoms, B. L., Schmidt, E., & Epstein, M. (1998). Effects of juror gender and verdict preference on persuasion during deliberations in child abuse trials. Paper presented at the Biennial meeting of the American Psychology-Law Society, Redondo Beach, CA. Bottoms, B. L., Shaver, P.R., Goodman, G. S., &Qin, J. J. (1995). In the name of God: A profile of religion-related child abuse. Journal of Social Issues, 51, 85-111. Bottoms, B. L., Wiley, T., & Stevenson, M. (2005, Mar.) Gender differences in community jurors' child sexual abuse case judgments. Presentation at the annual meeting of the American Psychology/Law Society, La Jolla, CA. Bradshaw, T. L., & Marks, A. E. (1990). Beyond a reasonable doubt: Factors that influence the legal disposition of child sexual abuse cases. Crime & Delinquency, 36, 276-285. Brekke, N., & Borgida, E. (1988). Expert psychological testimony in rape trials: A social-cognitive analysis. Journal of Personality and Social Psychology, 55, 372-386. Brewer, K. D., Rowe, D. M., & Brewer, D. D. (1997). Factors related to prosecution of child sexual abuse cases. Journal of Child Sexual Abuse, 6, 91-111. Brigham, J. C. (1972). Racial stereotypes: Measurement variables and the stereotype-attribute relationship. Journal of Applied Social Psychology, 2, 63-76. Broussard, S.D., & Wagner, W. G. (1988). Child sexual abuse: Who is to blame? Child Abuse & Neglect, 12, 563-569. Brownmiller, S. (1975). Against our will: Men, women and rape. New York: Simon & Schuster. Buck, J., & Warren, A. (in preparation). Effects of expert testimony in recovered memory cases. Unpublished manuscript. Burt, M. R., & Albin, R. S. (1981). Rape myths, rape definitions, and probability of conviction. Journal of Applied Social Psychology, 11, 212-230. Castelli, P., Goodman, G., & Ghetti, S. (in press). Effects of age and leading questions on children's credibility. Journal of Applied Social Psychology.

536

BOTTOMS ET AL.

Clark, H. L., & Nightingale, N. N. (1997). When jurors consider recovered memory cases: Effects of victim and juror gender. Journal of Offender Rehabilitation, 25, 87-104. Clary, E. G., & Shaffer D. R. (1985). Another look at the impact of juror sentiments toward defendants on juridic decisions. Journal of Social Psychology, 125, 63 7-651. Collings, S. J., & Payne, M. F. (1991). Attribution of causal and moral responsibility to victims of father-daughter incest: An exploratory examination of five factors. Child Abuse & Neglect, I5, 513-521. Corder, B. F., & Whiteside, R. (1988). A survey of jurors' perception of issues related to child sexual abuse. American Journal of Forensic Psychology, 6, 37-43. Coy v. Iowa, 487 U.S. 1012 (1988). Crawford v. Washington, 514 U.S._ (2004). Cross, T. P., DeVos, E., & Whitcomb, D. (1994). Prosecution of child sexual abuse: Which cases are accepted? Child Abuse and Neglect, 18, 663-677. Crowley, M. J., O'Callaghan, M., & Ball, P. J. (1994). The juridical impact of psychological expert testimony in a simulated child sexual abuse trial. Law & Human Behavior, I8, 89-105. Cutler, B. L., Penrod, S.D., & Dexter, H. R. (1990). Juror sensitivity to eyewitness identification evidence. Law & Human Behavior, I4, 185-191. Davies, G., & Noon, E. (1993). Video links: Their impact on child witness trials. Issues in Criminological and Legal Psychology, 20, 22-26. Davis, S. L., & Bottoms, B. L. (2002). Effects of social support on children's eyewitness reports: A test of the underlying mechanism. Law and Human Behavior, 26, 185-215. Davis, S. L., & Sobus, M.S. (2004, June). Strategic jury selection: Limiting mistakes in a difficult process. For the defense, 46, 12-15, 62. Deaux, K., & Ethier, K. A. (1998). Negotiating social identity. InJ. K. Swim &C. Stangor (Eds.), Prejudice: The target's perspective (pp. 301-323). San Diego: Academic Press. Deitz, S. R., Blackwell, K. T., Daley, P. C., & Bentley, B. J. (1982). Measurement of empathy toward rape victims and rapists. Journal of Personality and Social Psychology, 43, 3 72-384. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5-18. Devine, D. J., Clayton, L. D., Dunford, B. B., Seying, R., & Pryce, J. (2001). Jury decision making: 45 years of empirical research on deliberating groups. Psychology, Public Policy, and Law, 7, 622-727. Diamond, S. S. (1997). Illuminations and shadows from jury simulations. Law and Human Behavior, 21,561-571. Dollar, K. M., Perry, A. R., Fromouth, M. E., & Holt, A. R. (2004). Influence of gender roles on perceptions of teacher/adolescent student sexual relations. Sex Roles, 50, 91-100. Donnelly, D. A., & Kenyon, S. (1996). "Honey, we don't do men": Gender stereotypes and the provision of services to sexually assaulted males. Journal of Interpersonal Violence, 11, 441-448. Drugge, J. E. (1992). Perceptions of child sexual assault: The effects of victim and offender characteristics and behavior. Journal of Offender Rehabilitation, 18, 141-165. Duggan, L. M., Aubrey, M., Doherty, E., Isquith, P. K., Levine, M., & Scheiner, J. (1989). The credibility of children as witnesses in a simulated child sex abuse trial. InS. J. Ceci, D. F. Ross, &M.P. Toglia (Eds.), Perspectives on the child witness (pp. 71-99). New York: SpringerVerlag. Dukes, R. L., & Kean, R. B. (1989). An experimental study of gender and situation in the perception and reportage of child abuse. Child Abuse & Neglect, 13,351-360. Eaton, T., Ball, P., & O'Callaghan, M. (2001). Child-witness and defendant credibility: Child evidence presentation mode and judicial instructions. Journal of Applied Social Psychology, 3I, 1845-1858. Eisenberg, N., Owens, R. G., & Dewey, M. E. (1987). Attitudes of health professionals to child sexual abuse and incest. Child Abuse and Neglect, II, 109-116.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

537

Epstein, M. A., & Bottoms, B. L. (1998). Memories of childhood sexual abuse: A survey of young adults. Child Abuse and Neglect, 22, 1217-1238. Federal Rules of Evidence 401, 404a, 404b, 608b. (1984) Title 28, USCA. Finkelhor, D. (1983). Removing the child-Prosecuting the offender in cases of sexual abuse: Evidence from the national reporting system for child abuse and neglect. Child Abuse and Neglect, 7, 195-205. Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York: Free Press. Finkelhor, D., & Baron, L. (1986). Risk factors for child sexual abuse. Journal of Interpersonal Violence, I , 43-71. Finkelhor, D., Hotaling, G., Lewis, I. A., & Smith, C. (1990). Sexual abuse in a national sample of adult men and women: Prevalence, characteristics, and risk responsibility. Child Abuse and Neglect, 14, 19-28. Finkelhor, D., & Redfield, D. (1984). How the public defines sexual abuse. In D. Finkelhor (Ed.), Child sexual abuse: New theory and research (pp. 107-133). New York: Free Press. Finkelhor, D., & Russell, D. (1984). Women as perpetrators: Review of the evidence. In D. Finkelhor (Ed.), Child sexual abuse: New theory and research (pp. 171-187). New York: Free Press. Finkelhor, D., Williams, L. M., & Burns, L. (1988). Nursery crimes: Sexual abuse in day care. Newbury Park, CA: Sage. Finlayson, L. M., & Koocher, G. P. (1991). Professional judgment and child abuse reporting in sexual abuse cases. Professional Psychology: Research and Practice, 22, 464-472. Foley, L. A., & Chamblin, M. H. (1982). The effect of race and personality on mock jurors' decisions. The]oumal of Psychology, 112,47-51. ForsterLee, R., Horowitz, I., Ho, R., ForsterLee, L., & McGovern, A. (1999). Community members' perceptions of evidence: The effects of gender in a recovered memory civil trial. ]oumal of Applied Psychology, 84, 484-495. Freeman, J. B., Levine, M., & Doueck, H. J. (1996). Child age and caseworker attention in child protective services investigations. Child Abuse and Neglect, 20, 907-920. Friedrich, W. N., & Boriskin, J. A. (1978). Primary prevention of child abuse: Focus on the special child. Hospital and Community Psychiatry, 29, 248-251. Gabora, N. J ., Spanos, N. P., & Joab, A. (1993). The effects of complainant age and expert psychological testimony in a simulated child sexual abuse trial. Law and Human Behavior, 17, 103-119. Gilligan, C. (1982). In a different voice: Psychological theory and women's development. Cambridge, MA: Harvard University Press. Golding, J. M., Alexander, M. C., & Stewart, T. L. (1999). The effect of hearsay witness age in a child sexual assault trial. Psychology, Public Policy, & Law, 5, 420-438. Golding, J. M., Fryman, H. M., Marsil, D. F., & Yozwiak, J. A. (2003). Big girls don't cry: The effect of child witness demeanor on juror decisions in a child sexual abuse trial. Child Abuse & Neglect, 27, 1311-1321. Golding, J. M., Sanchez, R. P., & Sego, S. A. (1997). The believability of hearsay testimony in a child sexual assault trial. Law & Human Behavior, 21, 299-3 25. Golding, J. M., Sanchez, R. P., & Sego, S. A. (1999). Brief research report: Age factors affecting the believability of repressed memories of child sexual assault. Law & Human Behavior, 23, 257-268. Golding, J. M., Sego, S. A., & Sanchez, R. P. (1999). The effect of multiple childhood sexual assaults on mock-jurors' perceptions of repressed memories. Behavioral Sciences and the Law, 17, 483-493. Golding, J. M., Sego, S. A., Sanchez, R. P., & Hasemann, D. (1995). The believability of repressed memories. Law and Human Behavior, 19, 569-592. Golding, J. M., Stewart, T. L., Yozwiak, J. A., Djdali, Y., &Sanchez, R. P. (2000). The impact of DNA evidence in a child sexual assault trial. Child Maltreatment, 5, 373-383.

538

BOTTOMS ET AL.

Golding, J. M., Warren, A. R., & Ross, D. F. (1997). On legal validity, internal validity, and ecological validity: Comment on Washy and Brody. Law and Human Behavior, 21, 693-695. Goodman, G. S., Aman, C. J., & Hirschman, J. (1987). Child sexual and physical abuse: Children's testimony. InS. J. Ceci, M.P. Toglia, & D. F. Ross (Eds.), Children's eyewitness memory (pp. 1-23). New York: Springer-Verlag. Goodman, G. S., Batterman-Faunce, J. M., Schaaf, J. M., & Kenney, R. (2002). Nearly 4 years after an event: Children's eyewitness memory and adults' perceptions of children's accuracy. Child Abuse & Neglect, 26,849-884. Goodman, G. S., Bottoms, B. L., Herscovici, B. B., & Shaver, P.R. (1989). Determinants of the child victims' perceived credibility. InS. J. Ceci, D. F. Ross, & M.P. Toglia (Eds.), Perspectives on children's testimony (pp. 1-22). New York: Springer-Verlag. Goodman, G. S., Golding, J. M., & Haith, M. M. (1984). Jurors' reactions to child witnesses. Journal of Social Issues, 40, 139-156. Goodman, G. S., Golding, J. M., Helgeson, V. S., Haith, M. M., & Michelli, J. (1987). When a child takes the stand: Jurors' perceptions of children's eyewitness testimony. Law and Human Behavior, 11, 27-40. Goodman, G. S., Levine, M., & Melton, G. B. (1992). The best evidence produces the best law. Law and Human Behavior, 16, 244-251. Goodman, G. S., Levine, M., Melton, G. B., & Ogden, D. (1991). Craig vs. Maryland. Amicus brief to the U.S. Supreme Court on behalf of the American Psychological Association. Law and Human Behavior, 15, 13-30. Goodman, G. S., Quas, J. A., Bulkley, J., & Shapiro, C. (1999). Innovations for child witnesses: A national survey. Psychology, Public Policy, and Law, 5, 255-281. Goodman, G. S., Tobey, A. E., Batterman-Faunce, J. M., Orcutt, H., Thomas, S., Shapiro, C., et al. (1998). Face-to-face confrontation: Effects of closed-circuit technology on children's eyewitness testimony and jurors' decisions. Law & Human Behavior, 22, 165-203. Gray, E. (1993). Unequal justice. New York: Free Press. Griffith, J. D., Libkuman, T. M., & Poole, D. A. (1998). Repressed memories: The effects of expert testimony on mock juror's decision making. American Journal of Forensic Psychology, 16, 5-23. Grisso, T., & Melton, G. B. (1987). Getting child development research to legal practitioners: Which way to the trenches? In G. B. Melton (Ed.), Reforming the law: Impact of child development research (pp. 146-176). New York: Guilford. Haegerich, T. M., & Bottoms, B. L. (2000). Empathy and jurors' decisions in patricide trials involving child sexual assault allegations. Law & Human Behavior, 24, 421-448. Haegerich, T., & Bottoms, B. L. (2004, Mar.) Jurors' perceptions of juvenile defendants: The effect of stereotypes pre- and post-deliberation. Presentation at the biennial meeting of the American Psychology/Law Society, Austin, TX. Hamilton, D. L., & Fallot, R. D. (1974). Information salience as a weighting factor in impression formation. Journal of Personality and Social Psychology, 30, 444-448. Hamlin, S. (1985). What makes juries listen: A communications expert looks at the trial. Englewood Cliffs, NJ: Prentice Hall. Hastie, R., Penrod, S.D., & Pennington, N.H. (1983) Inside the jury. Cambridge, MA: Harvard University Press. Hetherton, J. (1999). The idealization of women: Its role in the minimization of child sexual abuse by females. Child Abuse and Neglect, 23, 161-174. Hetherton, J., & Beardsall, L. (1998). Decisions and attitudes concerning child sexual abuse: Does the gender of the perpetrator make a difference to child protection professionals? Child Abuse and Neglect, 22, 1265-1283. Hunter, M. (1990). Abused boys: The neglected victims of sexual abuse. Lexington, MA: Lexington Books.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

539

Isquith, P. (1988, April). Students and community members as jurors. In M. Levine (Chair), Simulated jury research on a child as a witness. Symposium conducted at the meeting of the Eastern Psychological Association, Buffalo, NY. Isquith, P. K., Levine, M., & Scheiner, J. (1993). Blaming the child: Attribution of responsibility to victims of child sexual abuse. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving testimony (pp. 203-228). New York: Guilford. John Jay College Research Team. (2004). The nature and scope of sexual abuse of minors by Catholic priests and deacons in the United States, 1950-2002. Washington, DC: United States Conference of Catholic Bishops. Jones, L., & Finkelhor, D. (2001). The decline in child sexual abuse cases. Office of Juvenile Justice and Delinquency Prevention Bulletin (NCJ 184741). Kalven, H., & Ziesel, H. (1966). The American jury. Boston: Little, Brown and Co. Kaplan, M. (1985). Character testimony. InS. Kassin & L. Wrightsman (Eds.), The psychology of evidence and trial procedure. Beverly Hills, CA: Sage. Kelley, H. H. (1967). Attribution theory in social psychology. Nebraska Symposium on Motivation, 15, 192-238. Kelly, R. J., & Tarran, M. J. (1984, April). Negative homosexuality bias in reactions to adult-child sex. Paper presented at the meeting of the Western Psychological Association, Los Angeles. Key, H. G., Warren, A. R., & Ross, D. F. (1996). Perceptions of repressed memories: A reappraisal. Law and Human Behavior, 20, 555-563. Kovera, M. B., & Borgida, E. (1996). Children on the witness stand: The use of expert testimony and other procedural innovations in U.S. child sexual abuse trials. In B. L. Bottoms & G. S. Goodman (Eds.), International perspectives on child abuse and children's testimony: Psychological research and law (pp. 201-220). Thousand Oaks, CA: Sage. Kovera, M.A., Borgida, E., Gresham, A. W., Swim, J., & Gray, E. (1993). Do child sexual abuse experts hold partisan beliefs? A national survey of the society for traumatic stress studies. Journal of Traumatic Stress, 6, 383-404. Kovera, M. B., Gresham, A. W., Borgida, E., Gray, E., & Regan, P. C. (1997). Does expert testimony inform or influence juror decision-making. Journal of Applied Psychology, 82, 178-191. Kovera, M. B., Levy, R. ]., Borgida, E., & Penrod, S.D. (1994). Expert witnesses in child sexual abuse cases: Effects of expert testimony and cross-examination. Law and Human Behavior, 18, 653-674. Landrine, H. (1985). Race x class stereotypes of women. Sex Roles, 13, 65-75. Leippe, M. R. (1985). The influence of eyewitness nonidentifications on mock-jurors' judgments of a court case. Journal of Applied Social Psychology, 15, 656-672. Leippe, M. R., Brigham, J. C., Cousins, C., & Romanczyk, A. (1989). The opinions and practices of criminal attorneys regarding child eyewitnesses: A survey. InS. J. Ceci, D. F. Ross, & M. P. T oglia (Eds.), Perspectives on children's testimony (pp. 100-130). New York: Springer-Verlag. Leippe, M., Manion, A., & Romanczyk, A. (1992). Eyewitness persuasion: How and how well do fact finders judge the accuracy of adults' and children's memory reports? Journal of Personality and Social Psychology, 63, 181-197. Leippe, M. R., & Romanczyk, A. (1987). Children on the witness stand: A communication/ persuasion analysis of jurors' reactions to child witnesses. InS. J. Ceci, M.P. Togolia, & D. F. Ross (Eds.), Children's eyewitness memory (pp. 155-177). New York: Springer-Verlag. Lieppe, M. R., & Romanczyk, A. (1989). Reactions to child (vs. adult) eyewitnesses: The influence of jurors' preconceptions and witness behavior. Law and Human Behavior, 13, 103-132. Lind, E. A., & Walker, L. (1979). Theory testing, theory development, and laboratory research on legal issues. Law and Human Behavior, 3, 5-19. Lindsay, D. S., & Read, J.D. (1995). "Memory work" and recovered memories of childhood sexual abuse: Scientific evidence and public, professional, and personal issues. Psychology, Public Policy, and Law, 4, 846-908.

540

BOTTOMS ET AL.

Lindsay, R. C., Ross, D. F., Lea, J. A., & Carr, C. (1995). What's fair when a child testifies? Journal of Applied Social Psychology, 25, 870-888. Linholm, K. J., & Willey, R. (1986). Ethnic differences in child abuse and sexual abuse. Hispanic Journal of Behavioral Sciences, 8, 111-125. Loftus, E. F. (1993). The reality of repressed memories. American Psychologist, 48, 518-537. Loftus, E. F., Weingardt, K., & Hoffman, H. (1993). Sleeping memories on trial: Reactions to memories that were previously repressed. Expert Evidence, 2, 51-59. MacArthur, L. A. (1972). The how and what of why: Some determinants and consequences of causal attribution. Journal of Personality and Social Psychology, 22, 171-193. MacCoun, R. J., & Kerr, N. L. (1988). Asymmetric influence in mock jury deliberation: Jurors' bias for leniency, Journal of Personality and Social Psychology, 54, 21-33. MacMurray, B. (1989). Criminal determination for child sexual abuse. Journal of Interpersonal Violence, 4, 244-288. Maryland v Craig, 58 LW 5044 (1990). Maynard, C., & Wiederman, M. (1997). Undergraduate students' perceptions of child sexual abuse: Effects of age, sex, and gender-role attitudes. Child Abuse & Neglect, 21, 833-844. McArthur, L.A. (1972). The how and what of why: Some determinants and consequences of causal attribution. Journal of Personality and Social Psychology, 22, 171-193. McAuliff, B., & Kovera, M. B. (2002). The status of evidentiary and procedural innovations in child abuse proceedings. In B. L. Bottoms, M. B. Kovera, & B. McAuliff (Eds.), Children, social science, and the law (pp. 412-445). New York: Cambridge University Press. McCauley, M. R., & Parker, J. F. (2001). When will a child be believed? The impact of the victim's age and juror's gender on children's credibility and verdict in a sexual-abuse case. Child Abuse & Neglect, 25, 523-539. Morison, S., & Greene, E. (1992). Juror and expert knowledge of child sexual abuse. Child Abuse and Neglect, 16, 595-613. Moscovici, S., & Zavalloni, M. (1969). The group as a polarizer of attitudes. Journal of Personality and Social Psychology, 12, 125-135. Muller, R. T., Caldwell, R. A., & Hunter, J. E. (1993). Child provocativeness and gender as factors contributing to the blaming of victims of physical child abuse. Child Abuse & Neglect, 17, 249-260. Myers, D. G., & Bishop, G. D. (1970). Discussion effects on racial attitudes. Science, 169, 778779. Myers, J. E. B. (1998). Legal issues in child abuse and neglect (2nd ed.). Thousand Oaks, CA: Sage. Myers, J. E. B. (2004). A history of child protection in America. Philadelphia: Xlibris Corporation. Myers, J. E., Redlich, A., Goodman, G., Prizmich, L., & lmwinkelried, E. (1999). Juror's perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy, & Law, 5, 388-419. Niemann, Y. F., Jennings, L., Rozelle, R. M., Baxter, J. C., & Sullivan, E. (1994). Use offree responses and cluster analysis to determine stereotypes of eight groups. Personality and Social Psychology Bulletin, 20, 379-390. Nightingale, N. N. (1993). Juror reactions to child victim witnesses: factors affecting trial outcome. Law & Human Behavior, 17, 679-694. O'Donohue, W. T., Elliott, A. N., Nickerson, M., & Valentine, S. (1992). Perceived credibility of children's sexual abuse allegations: Effects of gender and sexual attitudes. Violence and Victims, 7, 147-155. O'Donohue, W., & O'Hare, E. (1997). The credibility of sexual abuse allegations: Child sexual abuse, adult rape, and sexual harassment. Journal of Psychopathology and Behavioral Assessment, 19, 273-279. O'Donohue, W., Smith, V., & Schewe, P. (1998). The credibility of child sexual abuse allegations: Perpetrator gender and subject occupational status. Sexual Abuse: Journal of Research & Treatment, 10, 17-24.

18. FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

541

Ondersma, S. J., Chaffin, M., Berliner, L., Cordon, I., Goodman, G. S., & Barnett, D. (2001). Sex with children is abuse: Comments on the Rind et al. meta-analysis controversy. Psychological Bulletin, 127, 707-714. Orcutt, H. K., Goodman, G. S., Tobey, A., Batterman-Faunce, J., Thomas, S., & Shapiro, C. (2001). Detecting deception: Factfinders' abilities to assess the truth. Law and Human Behavior, 25, 337-370. Pendergrast, M. (1995). Victims of memory. Hillsdale, VT: Upper Access Books. Penrod, S. D., & Cutler, B. L. (1987). Assessing the competence of juries. In A. K. Hess & I. B. Weiner (Eds.), Handbook offorensic psychology (pp. 293-318). Oxford, England: John Wiley &Sons. Pennsylvania v. Ritchie, 480 U.S. 39, (1987). Pipe, M.-E., Lamb, M., Orbach, Y., & Cedarborg, A. C. (Eds.) (in press). Child sexual abuse: Disclosure, delay and denial. Mahwah, NJ: Lawrence Erlbaum Associates. Podell, D. M., Kastner, J., & Kastner, S. (1994). Mental retardation and adult women's perceptions of adolescent sexual abuse. Child Abuse and Neglect, 18, 808-819. Podell, D. M., Kastner, J., & Kastner, S. (1996). Adolescents with mental retardation: Perceptions of sexual abuse. American Journal of Orthopsychiatry, 66, 103-110. Powell, G. J., Wyatt, G. E., & Bass, B. A. (1983). Mental health professionals' views of AfroAmerican family life and sexuality.Joumal of Sex and Marital Therapy, 9, 51-66. Quas, J. A., Bottoms, B. L., Haegerich, T. M., & Nysse-Carris, K. L. (2002). Effects of victim, defendant and juror gender on decisions in child sexual assault cases. Journal of Applied Social Psychology, 32, 1993-2021. Redlich, A. D., Myers, J. E., Goodman, G. S., & Qin, J. (2002). A comparison of two forms of hearsay in child sexual abuse cases. Child Maltreatment, 7, 312-328. Regan, P. C., & Baker, S. J. (1998). The impact of child witness demeanor on perceived credibility and trial outcome in sexual abuse cases. Journal of Family Violence, 13, 187-195. Reppucci, N. D. (1985). Psychology in the public interest. In A. M. Rogers & C. J. Scheirer (Eds.), The G. Stanley Hall Lecture Series (Vol. 5). Washington, DC: American Psychological Association. Rind, B., Tromovitch, P., & Bauserman, R. (1998). A meta-analytic examination of assumed properties of child sexual abuse using college samples. Psychological Bulletin, 124, 22-53. Rogers, C. M., & Terry, T. (1984). Clinical interventions with boy victims of sexual abuse. In I. R. Stuart &J. G. Greer (Eds.), Victims of sexual aggression (pp. 91-104). New York: Van Nostrand Reinhold. Ross, D. F., Hopkins, S. Hansen, E., Lindsay, R. C. L., Eslinger, T., &Hazen, K. (1994). The impact of protective shields and videotape testimony on conviction rates in trials of child sexual abuse. Law & Human Behavior, 18, 553-566. Ross, D. F., Lindsay, R. C. L., & Marsil, D. F. (1999). The impact of hearsay testimony on conviction rates in trials of child sexual abuse: Toward balancing the rights of defendants and child witnesses. Psychology, Public Policy, and Law, 5, 439-455. Ross, D. F., Miller, B. S., & Moran, P. B. (1987). The child in the eyes of the jury: Assessing mock jurors' perceptions of the child witness. InS. J. Ceci, M.P. Toglia, & D. F. Ross (Eds.), Children's eyewitness memory (pp. 142-154). New York: Springer-Verlag. Sandys, M., & Dillehay, R. C. (1995). First-ballot votes, predeliberation dispositions, and final verdicts in jury trials. Law and Human Behavior, 19, 175-195. Scheiner, J. L. (1988, April). The use of the minimalist vignette as a method for assessing the generalizability of videotape trail simulation results. In M. Levine (Chair), Simulated jury research on a child as a witness. Symposium conducted at the meeting of the Eastern Psychological Association, Buffalo, NY. Schmidt, C. W., & Brigham, J. C. (1996). Jurors' perceptions of child victim-witnesses in a simulated sexual abuse trial. Law and Human Behavior, 20, 581-606.

542

BOTTOMS ET AL.

Schuller, R. A., & Hastings, P. A. (1996). Trials of battered women who kill: The impact of alternative forms of expert evidence. Law and Human Behavior, 20, 167-187. Schuller, R. A., Smith, V. L., & Olson, J. (1994). Jurors' decisions in trials of battered women who kill: The impact of prior beliefs and expert testimony. Journal of Applied Social Psychology, 24,316-337. Schutte, J. W., & Hosch, H. M. (1997). Gender differences in sexual assault verdicts: A metaanalysis. Journal of Social Behavior & Personality, 12, 759-772. Sealy, A. P., & Cornish, W.R. (1973), Juries and the rules of evidence. Criminal Law Review, 208-223. Sepler, F. (1990). Victim advocacy and young male victims of sexual abuse: An evolutionary model. In M. Hunter (Ed.), The sexually abused male: Prevalence, impact, and treatment (Vol. 1, pp. 73-85). Lexington, MA: Lexington Books. Shaffer, D. R. (1985). The defendant's testimony. InS. Kassin & L. Wrightsman (Eds.), The psychology of evidence and trial procedure. Beverly Hills, CA: Sage. Shaw, J. 1., & Skolnick, P. (1995). Effects of prohibitive and informative judicial instructions on jury decision making. Social Behavior and Personality, 23, 319-326. Smith, H., Fromuth, M. E., & Morris, C. (1997). Effects of gender on perceptions of child sexual abuse. Journal of Child Sexual Abuse, 6, 51-63. Snyder, H. N. (2000). Sexual assault of young children as reported to law enforcement: Victim, incident, and offender characteristics. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau ofJustice Statistics. Stewart, T. L., Whiteside, S., & Golding, J. M. (2000). The effective of expert witnesses in civil trials involving repressed memories of sexual assault. American Journal of Forensic Psychology, 18, 27-62. Strier, F. (1999). Whither trial consulting? Issues and projections. Law and Human Behavior, 23, 93-116. Sugarman, D. B., & Boney-McCoy, S. (1997). Impact of expert testimony on the believability of repressed memories. Violence and Victims, 12, 115-126. Swim, J., Borgida, E., & McCoy, K. (1993). Videotaped versus in-court witness testimony: Does protecting the child witness jeopardize due process? Journal of Applied Social Psychology, 23, 603-631. Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33, 1-39. Trivelpiece, J. W. (1990). Adjusting the frame: Cinematic treatment of sexual abuse and rape of men and boys. In M. Hunter (Ed.), The sexually abused male: Prevalence, impact and treatment (Vol. 1, pp. 47-71). Lexington, MA: Lexington Books. Tubb, V., Wood, J. M., & Hosch, H. M. (1999). Effects of suggestive interviewing and indirect evidence on child credibility in a sexual abuse case. Journal of Applied Social Psychology, 29, 1111-1127. Tzeng, 0. C. S., & Schwarzin, H. J. (1990). Gender and race differences in child sexual abuse correlates. International]ournal of Interpersonal Relations, 14, 135-161. Urquiza, A. J., & Keating, L. M. (1990). The prevalence of sexual victimization of males. In M. Hunter (Ed.), The sexually abused male: Prevalence, impact, and treatment (Vol. 1). Lexington, MA: Lexington Books. Vidmar, N. (1997). Generic prejudice and the presumption of guilt in sex abuse trials. Law & Human Behavior, 21,5-25. Vieth, V. (2006). Unto the third generation: A call to end child abuse in the United States within 120 years. Journal of Aggression, Maltreatment, and Trauma. Reprinted in V. Veith, B. L. Bottoms, & A. Perona (Eds.), Ending child abuse: New techniques for investigation, prosecution, and prevention (pp. 5-54). Binghamton, NY: Haworth. Warren, A., Nunez, N., Keeney, J. M., Buck, J. A., & Smith, B. (2002). The believability of children and their interviewers' hearsay testimony: When less is more. Journal of Applied Psychology, 87, 846-857.

18.

FACTORS AFFECTING CHILD SEXUAL ABUSE CASES

543

Waterman, C. K., & Foss-Goodman, D. (1984). Child molesting: Variables relating to attribution of fault to victims, offenders, and nonparticipating parents. Journal of Sex Research, 20, 3 29-349. Weinberg, M.S., & Williams, C. J. (1988). Black sexuality: A test of two theories. The Journal of Sex Research, 25, 197-218. Weiten, W., & Diamond, S. S. (1979). A critical review of the jury simulation paradigm: The case of defendant characteristics. Law and Human Behavior, 3, 71-93. Westcott, H. L., & Jones, D. P. H. (1999). Annotation: The abuse of disabled children. Journal of Child Psychology and Psychiatry and Allied Disciplines, 40, 497-506. Whitcomb, D., Shapiro, E., & Stellwagen, L. (1985). When the victim is a child: Issues for judges and prosecutors. Washington, DC: U.S. Department of]ustice. Williams, L. M., & Farrell, R. A. (1990). Legal response to child sexual abuse in day care. Criminal Justice and Behavior, 17, 284-302. Wilson, T. C. (1996). Cohort and prejudice: Whites' attitudes toward Blacks, Hispanics, Jews, and Asians. Public Opinion Quarterly, 60, 253-274. Wyatt, G. E. (1982). Identifying stereotypes of Afro-American sexuality and their impact upon sexual behavior. In B. A. Bass, G. E. Wyatt, & G. J. Powell, (Eds.), The Afro-American family: Assessment, treatment, and research issues (pp. 333-346). New York: Grune & Stratton. Wyatt, G. E. (1985). The sexual abuse of Afro-American and White-American women in childhood. Child Abuse and Neglect, 9, 507-519. Wyatt, G. E. (1990). Sexual abuse of ethnic minority children: Identifying dimensions of victimization. Professional Psychology: Research and Practice, 21, 338-343. Yarmey, A. D., &Jones, H. P. T. (1983). Is the psychology of eyewitness identification a matter of common sense? InS. M.A. Lloyd-Bostock & B. R. Clifford (Eds.), Evaluating eyewitness evidence (pp. 13-40). Chichester, England: Wiley & Sons. Yozwiak, J. A. (2003). The testimony of a child witness with mental retardation in a sexual abuse case: To believe or not to believe? Dissertation Abstracts International: Section B: The Sciences and Engineering, 63, 10-B. Yozwiak, J. A., Golding J. M., & Marsil, D. F. (2004). The impact of type of out-of-court disclosure in a child sexual assault trial. Child Maltreatment, 9, 325-334.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

19 Children's Eyewitness Memory: Balancing Children's Needs and Defendants' Rights When Seeking the Truth Lindsay

c. Malloy

University of California, Irvine

Emilie Mitchell Stephanie Block University of California, Davis

Jodi A. Quas University of California, Irvine

Gail

s. Goodman

University of California, Davis, and University of Oslo, Norway

Judith was 14 when she testified against a stranger who had abducted and repeatedly sodomized her. When asked many years later about her legal experiences, she described how the defense attorneys tried to change her story around and twist her words. She then noted that no one was in the courtroom to help her, the victim, during this traumatic time in her life. (Quas eta!., 2005) Oscar was 10 the second time he testified against his mother, whom he had witnessed beat his younger sister to death with a shoe. His words faltered as he sat fearfully behind the witness stand, crying, and explaining, "She's gone ... She's dead. My mother killed her." (Chapman, 2002)

Children, like adults, can witness or experience brutal crimes. The eyewitness evidence that they provide may be critical for successful prosecution of serious offenses. To provide testimony, however, children must navigate the unfamiliar and unfriendly territory of the legal system. Despite any distress or fear they may harbor, child witnesses are expected to recount their experiences in a clear, cogent manner. The process of legal involvement, which typically begins with a police or child protective services investiga-

545

546

MALLOY ET AL.

tion and which not as typically but potentially ends with testifying in open court, can be overwhelming for any witness, not to mention a child. In the face of such an emotional experience-one that may follow a traumatizing crime-many have wondered if a child can provide accurate and complete eyewitness testimony, and if so, under what conditions. Because the prosecution of certain crimes (most notably child sexual abuse) depends so heavily on children's testimony, professionals have pondered ways that might bolster children's ability to recount their experiences while still maintaining fair judicial proceedings for the defendant. In the United States, this means also passing U.S. Constitutional muster. For instance, professionals have noted that, in the past, legal systems have concentrated to a large extent on protecting innocent adults from false imprisonment. Several legal doctrines afford such protection. For instance, according to the 6th Amendment of the U.S. Constitution, the accused has the right "to a speedy and public trial" and "to be confronted with the witnesses against him," including witnesses who are children. These rights, which are similar to those outlined in other national and international doctrines (e.g., United Nations' European Convention on Human Rights Article 6 [3]), are believed to reduce the likelihood of false imprisonment. Protecting innocent adults is a noble goal, and such protections are essentiaL However, starting in the 1970s, when the number of children coming in contact with the legal system began to increase, partly because of laws mandating the reporting of child maltreatment, concern was raised about the needs of child witnesses who were being thrust into the legal system. In an effort to facilitate children's legal involvement while promoting the truthseeking function of the justice system, several child-oriented forensic techniques and legal accommodations were proposed. In general, the techniques were designed to enhance children's ability to recount their experiences and reduce the distress children may endure as a result oflegal involvement. If children are so unduly distressed that they cannot answer questions about a witnessed or experienced event, determination of the truth-arguably one of the ultimate goals of the legal system-might never be realized. Yet, the child-oriented techniques have been met with controversy, both because their use may infringe upon a defendant's rights (e.g., to "confront" a witness, using U.S. Constitutional terminology) and because the techniques may not, as intended, facilitate the pursuit of justice. In this chapter, we argue that many of the proposed techniques and reforms have the potential to support the truth-seeking function of the legal system. Of course, given the importance of the issues involved, a skeptical stance is warranted, but empirical research increasingly justifies at least a subset of the reforms that have been suggested. Indeed, several countries (e.g., Israel) that have less restrictive guidelines than those outlined in the U.S. Constitution's 6th Amendment have adopted some of the procedures, although many countries follow, formally or informally, standards that nonetheless continue to limit the use of some reforms (e.g., the United Nations' European Convention on Human Rights, Article 6 [3], states that individuals accused of crimes have the right to "examine or have examined witnesses against them"). Because sufficient scientific research is still lacking for other techniques, however, our chapter stands as a call for additional research on the procedures and reforms about which too little knowledge exists. The chapter is organized as follows. We first review reasons why the special techniques and reforms, which include those used both during forensic investigations and in

19. BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

547

court, were initially proposed. We then describe a number of these reforms. We discuss not only how the reforms are designed to advance the truth-seeking function of the courts, but also why they are controversial and what empirical research suggests about their use. When possible, we offer suggestions for additional research that would foster further insight into the reforms' utility. Finally, we offer some general, tentative conclusions about the use of child-oriented techniques and reforms when children provide eyewitness testimony in legal cases. For heuristic purposes, we divide the procedures into two general categories: (1) those aimed at prompting or cuing children's memory or instructing children about the "rules" of the eyewitness memory task (e.g., forensic interview), thereby improving the accuracy and completeness of children's reports, and (2) those aimed at reducing children's distress, so that children feel comfortable enough to recount their memories. The two types of reforms are designed with the same underlying goal in mind, specifically reaching the truth (i.e., promoting accuracy in forensic interviews and increasing the court's ability to hear and determine the truth). However, the means used to achieve this goal differ based on whether the techniques are intended to prompt more complete and more accurate information, that is, to help the child retrieve and articulate information stored in memory, or intended to help children feel comfortable enough to report information that they remember but may be too scared to articulate. A second heuristic difference between the two categories of reforms concerns when the techniques are implemented. Techniques included in the first category are primarily for use during forensic interviews, whereas techniques and reforms included in the second category are primarily for use (or at least considered for use) in the courtroom when children may testify. Finally, we limit our review in two important ways. First, we focus primarily on reforms that are controversial in terms of balancing defendants' rights and children's needs and/or that have been the subject of empirical investigation. Second, we discuss legal changes proposed for use with child victim/witnesses rather than child defendants. There is a growing literature concerning legal accommodations for juvenile suspects. It is possible that reforms for child witnesses may similarly assist child defendants. However, because of space constraints (and the fact that limited research exists concerning reforms in the latter population), we do not concern ourselves with child defendants (see Grisso et al., 2003, and Steinberg & Scott, 2003).

THE NEED FOR SPECIAL TECHNIQUES AND LEGAL REFORMS FOR CHILD WITNESSES

Before we discuss specific advances in legal procedure and forensic techniques to accommodate child witnesses, it is important to address why the alterations were considered in the first place. The argument for the use of reforms makes a number of assumptions: that children's statements are a needed piece of evidence, that children have more difficulty than adults accessing memories in legal contexts, that legal involvement is distressing to children, and that this distress inhibits children's ability to remember and communicate information.

548

MALLOY ET AL.

Is Children's Eyewitness Evidence Important? The first assumption underlying the need for legal reforms is quite simply that children's statements are necessary in court and constitute an important piece of evidence. Eyewitness testimony is the centerpiece of many legal cases. As prominent memory researcher Elizabeth Loftus explains (Loftus & Ketcham, 1991), "Aside from a smoking pistol, nothing carries as much weight with a jury as the testimony of an actual witness" (p. 14). Of course, as has been demonstrated time and time again in laboratory studies and actual legal cases, eyewitnesses can be inaccurate. In fact, faulty eyewitness testimony was instrumental in leading to the original convictions in a vast majority of the first 180 wrongful convictions recently exonerated by DNA evidence (Innocence Project, 2006; para. 1). Children's eyewitness testimony can similarly constitute critical evidence. This situation is most common in cases involving allegations of child sexual abuse (Goodman, Quas, Bulkley, & Shapiro, 1999) but can occur for many other crimes as well, for example, murder, physical assault, domestic violence, and theft. In cases involving alleged child sexual abuse, there may be no physical evidence (e.g., of fondling). Moreover, children often delay reporting the crime (e.g., Goodman-Brown, Edelstein, Goodman, Jones, & Gordon, 2003; Sorensen & Snow, 1991), decreasing the likelihood that physical evidence of the abuse's occurrence exists. Even if physical evidence suggestive of abuse is available, it may not definitively identify a specific perpetrator. Thus, it is imperative that children's reports are clear, consistent, detailed, and, of most importance, truthful. When children's memory accounts are vague, inconsistent, and/or incomplete, their reports tend to be met with skepticism (e.g., Leippe, Manion, & Romanczyk, 1992; see Myers, 1992). Insofar as certain child-oriented techniques and legal reforms improve children's ability to communicate the truth while concurrently guarding against false reports and wrongful punishment of innocent adults, justice is more likely to prevail.

Do Children in the Legal Context Have More Difficulty Accessing Their Memories Than Do Adults? The potential importance of children's eyewitness statements does not necessarily demand changes in legal techniques or procedures to accommodate this group of witnesses. Instead, the reforms' necessity stems from the second, third, and fourth assumptions, which are reviewed in turn in this and the following sections. Regarding the second assumption, there is ample evidence from both laboratory and naturalistic studies that children typically have more difficulty than adults in retrieving information from memory (see Bjorklund, 2005). Young children in particular need more cuing and prompts than do adults to recount information retained in memory (e.g., Fivush, 1993; Price & Goodman, 1991). Developmental factors linked to maturation, knowledge base, and socio-emotional processes all likely underlie age-related trends in mnemonic abilities (see Pipe, Thierry, & Lamb, this volume). Children may also fail to access the appropriate memories (i.e., of the alleged crime) because they do not understand the "rules" of forensic and courtroom contexts, such as that it is fine to ask for clarification if they do not

19. BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

549

understand a question (Cordon, Saetermoe, & Goodman, 2005). Of note, children's potential difficulties accessing and recounting memories must be considered in conjunction with their increased risk, relative to adults, for memory errors and false reports of entire events following certain types of interview strategies (see Melnyk, Crossman, & Scullin, Chapter 14, this volume). More generally, given the robustness of findings concerning children's limited mnemonic abilities and concurrent heightened suggestibility, there is every reason to assume that findings from empirical studies will generalize to forensic and courtroom settings involving children.

Is Legal Involvement Stressful for Children? The third assumption is that participation in legal cases is distressing for children, an assumption supported by legal precedence and empirical evidence. Several legal doctrines, for example, deem children a special class of witnesses in need of additional protection because of children's increased vulnerabilities. California Evidence Code 765 states that, when child victim/witnesses are involved, the legal system is responsible for ensuring their emotional well-being while providing the fairest judicial process for the accused. The doctrine explicitly states that, when a witness is under 14 years of age, courts should protect the child from undue harassment or embarrassment, restrict unnecessary repetition of questions, and ensure that questions are stated in a developmentally appropriate form. The Federal Rules of Evidence 611 contains a similar, although less explicit, directive about protecting witnesses from "harassment or undue embarrassment." Underlying both doctrines is the implicit assumption that participation in legal processes is distressing to children and can cause emotional harm, an assumption confirmed by empirical research. Several studies have demonstrated that legal involvement is distressing to children and can lead to adverse short- and long-term consequences (e.g., Goodman et al., 1992; Quas et al., 2005; Sas, 1991). With regard to children's immediate experiences of distress, a number of facets of legal involvement increase children's anxiety. First, involvement exposes children to procedures and language with which they are unfamiliar (WarrenLeubecker, Tate, Hinton, & Ozbek, 1989). Legal formalities and terminology are foreign to many adults, not to mention children (Saywitz, Jaenicke, & Campara, 1990). Moreover, even children who have a basic understanding of the legal system often lack the ability to apply that understanding to a specific case (e.g., Warren-Leubecker et al., 1989). Children report being afraid that their lack of understanding will interfere with their ability to answer questions accurately in court (Sas, 1991). Finally, lack of knowledge about impending stressors is associated with increased distress (e.g., HarbeckWeber & McKee, 1995; Zeltzer, Fanurik, & LeBaron, 1989), including in legal settings (e.g., Goodman et al., 1998; Oran, 1989). Another distressing facet of legal involvement is discussion of highly personal and/or traumatic events. This distress may be especially pronounced in children who feel some degree of culpability for the alleged offense (e.g., Feiring, Taska, & Chen, 2002; Quas, Goodman, & Jones, 2003). Children are likely to feel afraid and embarrassed to disclose details during an interview with an unfamiliar adult and later in a courtroom

550

MALLOY ET AL.

full of strangers (e.g., Herman & Hirschman, 1981). Such feelings may be exacerbated when children are subjected to repeated interviews (e.g., Ghetti, Alexander, & Goodman, 2002). Third, in reported abuse and/or neglect cases, many children experience out-ofhome placement. For instance, over 20% of child victims of maltreatment in the United States are removed from their home and placed in foster care (U.S. Department of Health and Human Services, 2004). In Australia, the number of child victims who are placed in out-of-home care has increased by 45% between 1997 and 2003; over 12,000 children were placed outside the home between 2002 and 2003 (Australian Institute of Health and Welfare, 2004). Although some children experience benefits due to removal from an abusive situation, many others experience stress, anxiety, and uncertainty after loss of their family and familiar environment (Block, Oran, Goodman, & Oran, 2005). Children may further be separated from siblings and experience multiple placements, which can lead to additional confusion and anxiety (Chapman, Wall, & Barth, 2004; Kaplan, 2001; Oran, 1989; Pardeck, 1984). Fourth, children's actual in-court experiences also contribute to their distress. Testifying in and of itself is associated with increased anxiety, as measured via children's or parents' reports and direct observation (e.g., Davies & Noon, 1993; Goodman et al., 1992). Such findings are evident in studies of actual legal cases (e.g., Davies & Noon, 1991; Sas, 1991) and of mock trials (e.g., Goodman et al., 1998; Nathanson & Saywitz, 2003). One particularly distressing part of testifying is answering questions during crossexamination. As Weiss and Berg (1982) point out, although attorneys' specific focus may differ based on children's age (e.g., young children's inability to distinguish reality from fantasy versus adolescents' provocation of abuse), the primary goal of cross-examination is the same-to discredit witnesses by highlighting inconsistencies or errors in their reports. Accordingly, multiple rephrasing of questions, use of legalese, subtle to not-sosubtle intimidation, leading and possibly misleading questions, and/or witness badgering are the norm for this portion of a trial. Sas (1991) examined child victims' experiences in court. Children reported being afraid of "being screamed at" in the courtroom, an experience more likely to be associated with cross-examination than direct examination. As a final note, legal involvement, including testifying in court, is also associated with emotional harm over time, as measured in the short term (e.g., a few days to several months after participation) and long term (e.g., many years after participation). In the short term, at least among child sexual abuse victims, adverse mental health outcomes are associated with repeated interviews, testifying multiple times, harsh courtroom settings, cases lacking corroborative evidence, and children not having maternal support during their legal case (Goodman et al., 1992; Henry, 1997; Sas, 1991; see Whitcomb, 2003, for a review). Also among abuse victims, negative attitudes toward the legal system are significantly related to being older at the time of the legal case (e.g., a teenager) and being female, and having been subjected to multiple interviews (Goodman et al., 1992; Tedesco & Schnell, 1987). One longitudinal study found that several characteristics associated with increased emotional harm in the short term predict psychological problems in the long term, even into adulthood (Quas et al., 2005). Specifically, testify-

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

551

ing multiple times, especially against a defendant with whom a child victim is closely related, in addition to a not-guilty case outcome, was associated with mental health problems many years after participation in the legal system. Thus, not only are several parts of legal involvement immediately distressing to children, but some of these same components may contribute to adverse consequences that persist for many years.

Does Children's Distress Affect Their Memory and Communication? It is possible that, although children are distressed as a result of their participation in a legal case, this distress poses no threat to the legal system's truth-seeking function. If so, the case for forensic and courtroom changes would rest primarily on reducing emotional harm to children, a goal of considerable value in itself. Yet, as we discuss next, children's heightened distress may actually inhibit their memory and communication abilities. Accordingly, the utility oflegal reforms is not exclusive to helping children. Instead, the reforms, in theory, facilitate the pursuit of justice and may help protect innocent defendants. There are a number of reasons why heightened distress during children's participation in a legal case could adversely affect children's abilities to report their alleged experiences. Children may have difficulty communicating effectively because of emotional problems resulting from the crime or may need to devote attentional resources and effort to coping with their emotional distress rather than to conducting a memory search for relevant information (e.g., Quas, Bauer, & Boyce, 2004). Or, particularly in the case of child sexual abuse, children may be uncomfortable, embarrassed, or ashamed about their experiences (e.g., Feiring, Taska, & Lewis, 1999; Quas et al., 2003), decreasing their willingness to describe the experiences during forensic interviews and in court, especially in front of a number of strangers and possibly the alleged perpetrator. Of importance, potential adverse effects of stress at retrieval should not be confused with the effects of distress at the time of the alleged crime, which may instead benefit memory in important respects (e.g., Buchanan & Lovallo, 2001; see Christianson, 1992; Reisberg & Heuer, Chapter 3, this volume; McGaugh, 2003). Empirical studies confirm that stress at retrieval is negatively related to children's complete and/or accurate reporting of events (e.g., Nathanson & Saywitz, 2003; Quas & Lench, in press). For example, in one naturalistic study, Goodman et al. (1992) followed child sexual abuse victims during their involvement in criminal prosecutions; a subset of the children testified in court. Among the testifiers, those who were more fearful about testifying in front of the defendant, as evidenced by pretrial self-reports and observations during testimony, were less able to answer prosecutors' questions than those who were less fearful. In experimental studies, where it is possible to verify the accuracy of children's memory, children's free recall is less complete and their responses to direct questions are less accurate when they are questioned in a courtroom than in a more familiar location (e.g., a classroom) or less intimidating environment (e.g., small room in a

552

MALLOY ET AL.

university laboratory; e.g., Hill & Hill, 1987; Saywitz & Nathanson, 1993). Also, in an experimental study, Saywitz and Nathanson (1993) found that children who rated the legal process as stressful (which is more likely to occur when they testify in open court) provided less complete recollections of staged events than children who rated the legal process as less stressful. Anecdotally, many children who testified in a courtroom expressed their anxiety and unwillingness to testify again in the future (Hill & Hill, 1987). Another stressor associated with legal involvement, facing a defendant, even during a lineup situation, similarly adversely affects children's memory accounts. In two studies, children were less likely to identify an adult as guilty of a staged crime (e.g., stealing a book) in a live lineup versus a photo lineup (Dent, 1977; Peters, 1991). Although neither researcher tested specifically whether children's accuracy was mediated by their distress, Peters (1991) reported that many children who had failed to identify the perpetrator in the live lineup condition admitted afterward that they did not identify him out of fear. Finally, the accuracy of children's responses during cross-examination, one of the most feared parts of legal involvement, has also been examined. In a clever experiment, Zajac and Hayne (2003) examined whether the primary goal of cross-examination (to highlight inconsistencies in a witness' statements and uncover the truth) is achieved when the witness is a child. Five- and six-year-olds (half of whom were exposed to post-event misleading information) were interviewed about a visit to a police station. The interviews included both direct examination (6 weeks after the event) and cross-examination (8.5 months after the event). During the cross-examination, interviewers used several techniques observed in actual court cases to attempt to persuade children to change their original answer (regardless of the accuracy of the initial answer). In response to the cross-examination questions, over 85% of the children changed some answers that they had provided during direct testimony. The changes occurred irrespective of children's accuracy or whether they had received the misleading information before direct examination. Some changes were considered small alterations in peripheral details, whereas others represented total recantations of children's original reports. In fact, approximately one-third of children changed all of their original responses. Of importance, cross-examination was unsuccessful at discrediting inaccurate children and decreased accuracy among children who were initially correct. Again, the researchers did not measure children's distress during the interview, thus limiting empirical investigation of the mediating role of stress. It is nonetheless possible that the types of tactics used, which were typical of those in real-world cross-examinations (e.g., complex, leading, irrelevant questions; challenges to children's certainty; expressions of disbelief) increased children's anxiety and nervousness, leading them to change answers and provide inconsistent and often inaccurate responses. It should be noted, however, that in most of the experimental studies we have reviewed so far, children testified about events they had actually experienced. The question remains, therefore, as to how children would respond under the same conditions if they had not in fact experienced the alleged criminal act, but rather were reporting suggested or otherwise false memories. Additional research is clearly needed on this important topic.

19. BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

553

summary

To summarize, eyewitness accounts, including those from children, can constitute a crit~ ical piece of evidence. For children, this is especially likely in cases involving alleged child sexual abuse, for which physical evidence confirming the abuse and/or identifying the perpetrator is not always available. It is imperative that children's statements are ac~ curate and complete. Yet, children, like adults, often find the legal environment to be stressful. Facets of legal involvement that appear to be most distressing include provid~ ing testimony in open court, testifying while facing the defendant, and submitting to cross~examination. It is noteworthy that these same factors are associated with poorer eyewitness memory performance. We next discuss the extent to which legal reforms de~ signed to bolster children's memory accuracy and completeness, reduce children's dis~ tress, and/or enhance children's ability to communicate are successful in doing so and hence facilitate the legal system's overarching goal of the pursuit of justice.

TECHNIQUES AND REFORMS AIMED AT FACILITATING CHILDREN'S MEMORY

Many of the forensic and courtroom changes to accommodate children were intended specifically to improve both the accuracy and completeness of children's memory re~ ports. For the most part, these reforms were designed to be implemented during forensic investigations and thus before a case actually goes to trial, although some (e.g., truth~lie discussions, use of certain mnemonic aids) may also take place while children provide courtroom testimony. To the extent that some of the techniques aid children's memory generally, not just in forensic interviews, one may question whether such techniques should also be instituted as standard practice in courts of law. Although the reforms were meant to improve children's memory, they are not nee~ essarily prosecution oriented or designed specifically to benefit a prosecution's case. In~ stead, some of these changes were motivated by legal cases that involved highly contro~ versial allegations of child sexual abuse that may have resulted from inappropriate interviewing techniques (e.g., Garven, Wood, Malpass, & Shaw, 1998; State of New Jersey v. Margaret Kelly Michaels, 1994). The techniques and reforms we describe here vary in the amount of controversy associated with their use. At one extreme, Child Advocacy Centers have received considerable support, although there are still some controversial features of these centers. At the other extreme, forensic techniques involving mnemonic aids remain highly controversial, in large part because of concerns about children's sug~ gestibility and false reports. Child Advocacy centers (CACs)

Child Advocacy Centers (CACs; aka Multidisciplinary Interview Centers) are agencies that, within the last decade or so, have been established both in the United States and

554

MALLOY ET AL.

abroad (e.g., Norway) to help the police, child protective services, and prosecutors work together to centralize child maltreatment investigations (Jones, Finkelhor, & Kopiec 2001; Myhr, Melinder, Gabrielsen, Andreassen, & Pettersen, 2004). The centers have social workers, investigators, and medical personnel on staff so that when a child is alleged to have experienced some form of maltreatment, the necessary information (e.g., forensic interview, medical evaluation) and services can be delivered at a single location, at times during a single visit. The child is typically interviewed by a highly trained social worker, who uses a minimal number of leading questions. Interviews are also routinely videotaped, with a law enforcement officer and a deputy district attorney observing the interview from behind a one-way mirror. In many U.S. jurisdictions, if the case goes forward, the law enforcement officer appears at the preliminary hearing and describes the child's statements and possibly shows the videotaped interview. It is then determined whether the case should be bound over for trial. One controversy relevant to CACs concerns whether the interviewers should be blind to the allegations so as to avoid potentially leading children during questioning. Some centers do in fact use blind interviewers, but most do not. In some countries (e.g., Norway), defense attorneys (and even the judge) are permitted to observe behind the one-way mirror and to direct questions to the interviewer, who then poses the questions to the child (Justis-og Politidepartementet, 2004; Melinder, 2004). However, in the United States, typically the defense attorney and the judge are absent during the CAC interview. A second potentially controversial issue related but not exclusive to CACs is the videotaping of interviews. Although many academic and legal scholars have hailed videotaping interviews as a means of limiting the number of interviews to which children are exposed and ensuring that an objective record of an interview exists, some concerns have been raised that the interviews will be subjected to unwarranted levels of scrutiny and discrediting of children's statements (e.g., see Myers, 1998). To date, there has been little research on the consequences of videotaping interviews on the progression of legal cases or the believability of children's statements (but see Goodman et al., in press; Myers, Redlich, Goodman, Prizmich, & lmwinkelried, 1999). More generally, research has not tested empirically the utility of CACs. A notable exception, currently under way in the United States being conducted by Jones, Finkelhor, Cross, and colleagues, involves comparing communities with and without CACs to determine whether CACs make a difference in the kind of assistance being provided to child victims of abuse. Although preliminary results indicate that the number of interviews may not be significantly reduced in the communities with CACs versus those without them, other findings suggest that CACS are still beneficial (Jones, 2004). For instance, compared with children involved in investigations in communities without CACs, children seen at CACs are more likely to have a medical exam and to receive mental health services, and parents are more satisfied overall with the care their children receive. Thus, the CACs appear to ensure that children receive appropriate services in child maltreatment investigations. Of direct relevance to the courts, information gleaned from the medical exams and forensic interviews, conducted as soon as possible

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

555

following allegations, may provide valuable information for a given case. This informa~ tion protects the defendant from unfair charges and the child from retraumatization by the legal system. As a final note regarding CACs, many of them take advantage of research findings on forensic interviewing of children and often incorporate empirically tested techniques, including those discussed here regarding forensic interview practices. Moreover, CAC interviewers often receive specific training on research~driven methods of conducting interviews. These practices may well enhance the accuracy and completeness of chil~ dren's accounts, thereby benefiting the truth~seeking function of the legal system.

Truth-Lie Discussions discussions (TLDs) have been included in some form during in~court testi~ mony for centuries. However, the form they have taken has changed in recent years to better accommodate children's limited cognitive and language abilities. Moroever, such discussions have been incorporated into standardized forensic interviews, such as those conducted at CACs. TLDs include at least one of the following components: (1) discussion of the dif er~ ence between truth and lies, (2) explanation of the implications of lying, and/or (3) af~ firmation that the child witness will provide truthful statements. The necessity ofTLDs stems from competency requirements for the admittance of eyewitness testimony. Com~ petency requirements for a child include several components, including that the child understands the difference between the truth and a lie, and that the child affirms to tell the truth. (The child having an "independent," that is, uncoached, memory for the event and the ability to have perceived the event can also be issues of competence to testify.) Although the oath, in formal language, is generally not asked of children in court, developmental versions of the oath have been routinely administered in legal set~ tings (see Lyon, 2000, for a review). The legal system assumes that, when it encourages children to tell the truth, highlights that lies are bad, and elicits promises from children, they will provide only true information, children will be more accurate and honest in their accounts. As part of competence examinations to testify in court, traditionally, the judge would often ask the child witness to define the words truth and lie. However, defining such words can be a difficult task for a child. As researchers began to study children's abilities to pass competence examinations, they found that children's understanding of the difference between truth and lies, and of the obligatory nature of a promise to tell the truth, is affected by the types of questions children are asked. By age 4, children can successfully identify honest and dishonest statements as the truth and lies, respectively, and indicate that lies are wrong (e.g., Bussey, 1992; Haugaard, Reppucci, Laird, & Nau~ ful, 1991). However, when asked to define the truth or a lie, children perform much more poorly, and even 7~year-olds have considerable difficulty. Lyon and Saywitz (1999) extended these prior investigations by examining maltreated children's understanding of Truth~lie

556

MALLOY ET AL.

the truth and lies. Despite delays in linguistic abilities (which are typical of maltreated children relative to nonmaltreated children), by age 5 years, the maltreated children could identify truth and lies and knew that lies were morally wrong. Investigations of children's understanding of the obligatory nature of an affirmation to tell the truth typically involve eliciting promises to tell the truth or promises not to lie. Studies reveal that although children younger than age 9 years find it difficult to distinguish between promising an action and executing the promised action, indicating an incomplete understanding of the term (Astington, 1988), children as young as 4 and 5 years of age appreciate that a promise is an obligation to behave in a specified manner (Lyon, 2002; Lyon & Saywitz, 1999). Specifically, when asked which story character would engage in a particular way, even younger children selected the character who said either "promise" or "will" rather than "might" (the children did not appear to differentiate between promise and will, however). Thus, when asked developmentally appropriate questions, children's comprehension is noteworthy. Based on the research, it has become fairly standard practice in CACs for children to be asked truth-lie questions in developmentally appropriate, concrete ways (e.g., What do you call this? [as the interviewer points to a table]. If I called this a chair, would that be the truth or a lie?). As mentioned, one of the reasons for administering the oath is that it is believed to compel a witness to be truthful. Researchers have thus examined whether the oath's administration and children's ability to articulate the truth-lie distinction and agree to be truthful predict their memory accuracy and/or honesty. Results have been mixed (e.g., Goodman, Bottoms, Schwartz-Kenney, & Rudy, 1991; Huffman, Warren, & Larson, 1999), although a trend appears to suggest that the administration of developmentally appropriate, elaborate TLDs, as opposed to no discussion of the importance of honesty, is effective in improving accuracy and/or honesty, even though children's comprehension or performance during the actual TLDs may be unrelated to accuracy (Huffman, Warren, & Larson, 1999; London & Nunez, 2002). Recently, Lyon and colleagues investigated whether eliciting a promise to tell the truth increased the accuracy of children's reports of a transgression that either occurred or did not occur (Lyon & Dorado, 2005; Lyon, Malloy, Talwar, & Quas, 2004). The studies were conducted with maltreated children, who either engaged in a minor mock transgression (playing with a forbidden toy) or did not. Afterward, some children were instructed to lie and either deny actual play or falsify having played. Findings revealed that the developmentally appropriate version of the oath increased children's true reports of the transgression, and that the oath's administration did not adversely affect false reports of the transgression among children who had not engaged in the act. Finally, the oath was effective in reducing false reports among children coached to lie and claim, falsely, that the transgression occurred. Thus, again, although children's actual performance on TLDs does not necessarily predict accuracy, the mere discussion of the importance of telling the truth along with an elicitation of a promise to tell the truth does improve children's honesty, in terms of both disclosing true transgressions and minimizing intentionally false allegations of transgressions. Insofar as children's honesty is enhanced by the simple inclusion of child-friendly TLDs, so is the truthseeking function of the courts.

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

557

Child-Oriented Instructions Structured forensic interviews, such as those administered at CACs, now also typically include special instructions to children, including (1) it is important not to guess, (2) it is fine to say "I don't know," (3) it is fine to say "I don't understand" if the child does not understand a question, (4) the interviewer was not present during the event in question and thus doesn't know what happened, and (5) the interviewer might need to repeat questions so that the interviewer can make sure she understands what happened. His~ torically, such instructions were not typical, but fortunately, researchers have shown that child~oriented instructions of this nature improve children's eyewitness memory perfor~ mance, including in quite young children (e.g., 4~year~olds). One wonders if such in~ structions might also benefit children in the courtroom. Moreover, researchers have examined the effects of such instructions specifically in reducing children's errors. For instance, encouraging responses of "I don't know" (Mulder & Vrij, 1996; Waterman, Blades, & Spencer, 2001), providing warnings about false sug~ gestions (e.g., "I may trick you"; Warren, Hulse~Trotter, & Tubbs, 1991), and providing explanations concerning the interviewer's lack of knowledge (Mulder & Vrij, 1996) have been shown to reduce children's inaccuracies. Cord6n et al. (2005) found that giv~ ing young children (3.5~ to 6~year~olds) three instructions ("It is fine to say 'I don't know'," "I can't help you," "I may try to trick you") led to a smaller proportion of incor~ rect responses, even when repetitive and accusatory interview styles were used. How~ ever, it was necessary to provide all three rules and no distractions to ensure positive ef~ fects. It is possible that such instructions could similarly help children maintain greater promoting the eyewitness accuracy during trials, for instance, during cross~examination, truth~ seeking function of the courts.

Mnemonic Aids Mnemonic aids have been proposed and utilized as a means of enhancing the amount of information children provide about events. Given the fact that children often need cues to retrieve information from memory, mnemonic aids are potentially useful. Moreover, mnemonic aids may help children recount experiences that are embarrassing or other~ wise difficult to describe. Concerns that certain mnemonic aids, such as anatomically detailed dolls, increase children's suggestibility have made their use particularly contro~ versial. As discussed next, researchers have responded by studying the conditions under which specific mnemonic aids, such as having children draw (e.g., with crayons) their experiences and providing children with anatomical human figure drawings or other relevant props, are beneficial versus detrimental to children's accuracy. Studies indeed suggest that having children draw particular events that they have experienced or witnessed can facilitate children's memory in both clinical (Drucker et al., 1997, as cited in Gross & Hayne, 1999) and experimental (e.g., Butler, Gross, & Hayne, 1995) situations. For example, when children are asked to draw what they experienced while recounting an event, the amount of accurate information children provide in their free~recall narratives and in response to direct questions increases (Gross & Hayne, 1999).

558

MALLOY ET AL.

Of note, however, drawings do not appear to enhance memory accuracy among young (3- to 4-year-old) children, although drawings may not significantly increase errors regardless of age, at least when children are not exposed to misleading suggestions. Finally, the benefits of drawing have been reported even over delays as long as 1 year after a target event's occurrence (Butler et al., 1995; Gross & Hayne, 1999). Drawings may be beneficial for a number of reasons. Children may feel more comfortable following a directive to "draw and tell" what happened rather than just to verbally recount what occurred. Specifically, the increased comfort of engaging in an activity that children enjoy may allow them to focus longer on the memory task instead of becoming distracted or wishing to terminate the interview. Alternatively, the increased duration of the interview may allow children to reminisce longer and therefore recall more information about the event (Roediger & Thorpe, 1978). Finally, drawing may allow children to form concrete retrieval cues that facilitate their ability to recall a given event. However, and of importance, Bruck, Melnyk, and Ceci (2000) found that drawing increases false reports when young children are asked to draw falsely suggested information instead of actual experiences. In their study, 3- to 6-year-olds played with a magician who performed tricks. During two subsequent suggestive interviews, children were given both true and false reminders about the play experience. Half of the children were asked to draw these reminders, and the others were simply asked to answer questions about the reminders. In a final interview, children were asked free recall and recognition questions. Similar to previous findings, drawing increased the amount of true information. However, drawing also increased the inclusion of false information, especially that which had been suggested to children during the prior interviews. These findings have led to concerns about the use of drawing during forensic interviews with children. False information provided by a child may well jeopardize not only the court's ability to determine the truth but also a defendant's and the child's own future. Indeed, many CACs avoid having children draw during interviews because of these concerns as well as additional fears that, if interviews are videotaped while children are drawing, children will appear less credible. Another type of mnemonic aid often used now during forensic interviews involves providing children with schematic drawings of adults, typically with anatomically correct features (rather than allowing children to draw freely). These drawings are believed to facilitate children's ability or willingness to describe personal touch and are believed to be less suggestive to children than anatomical dolls, which may promote exploratory behavior and play on the part of children (Everson & Boat, 2002). Findings concerning the utility of anatomical drawings are mixed. Aldridge and colleagues, for example, found that unclothed, gender-neutral body outline drawings elicited forensically pertinent information even after a trained interviewer (using a highly structured interview protocol, the NICHD protocol; see Lamb, Orbach, Warren, Esplin, & Hershkowitz, Chapter 15, this volume) thought that all possible information had been obtained (Aldridge et al., 2004). The benefits were particularly strong among children aged 4 to 7 years: 27% of the total details reported were elicited by the human figure drawings. Of note, because the children in the study were being questioned in actual sexual abuse cases, the accuracy of the details could not be determined. Moreover, the details were obtained mostly through

19. BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

559

questions that focused on recognition rather than recall memory. Such questions are associated with increased error rates. The researchers thus suggest that, although the drawings may have elicited more information, they should be introduced as late as possible in an investigation. Steward and colleagues reported limitations regarding the use of anatomical drawings in a study of children's memory for medical procedures that included or did not include genital touch (Steward et al., 1996). Children were more likely to report true genital touch if they were shown the anatomical pictures rather than just verbally asked. However, children who did not receive genital touch during the exam also were more likely to report that they were touched when shown the pictures. Thus, as highlighted in other studies that paired mnemonic aids with highly leading questions about body touch (e.g., Bruck, Ceci, Francouer, & Renick, 1995; Saywitz, Goodman, Nicholas, & Moan, 1991), although the drawings can lead to more accurate memory in cases where touching had actually occurred, the drawings may increase children's susceptibility to false reports when no touching took place. A final type of mnemonic aid worth mentioning briefly concerns the use of props other than those that are anatomical. In some actual legal cases, young children have profited from using props to recount traumatic experiences Qones & Krugman, 1986). However, research suggests that the use of props with young children (e.g., 3-year-olds) is risky (Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1997; Gordon et al., 1993; Pipe, Gee, & Wilson, 1993) and that to the extent props are useful with young children, the props need to be restricted to those that symbolize things known to have been present during the alleged event (e.g., Price & Goodman, 1990). Older children, alternatively, do appear to profit from the use of props (Goodman et al., 1997). Because of the event-specific nature of these types of props and because of concerns about children's suggestibility, they do not appear to be regularly used at most forensic interview centers or in court when children testify. Of interest, however, for many years, courts have permitted the admittance of prop-assisted eyewitness testimony from children.

Innovative Interviewing Methods Several structured interview protocols and guidelines have been developed to help forensic interviewers question children in the least leading manner possible. Such protocols promote the use of open-ended questions, prompt children for narrative detail, and avoid leading questions. Others provide explicit training instructions to teach children how to remember events, answer open-ended questions, or discriminate among memories derived from actual experience and memories derived from external influences. Three examples of such interviewing methods include the NICHD protocol, narrative elaboration, and source-monitoring training. The NICHD protocol, which is perhaps the most well-known and widely tested interview protocol, holds considerable promise as a means of enhancing children's memory. However, because it is described in detail in another chapter (see Lamb et al., this volume), we focus here on the other two examples. Narrative elaboration largely involves teaching children the categories of information (e.g., actions, people) that are important to report. Although even young preschool-

560

MALLOY ET AL.

age children can recount, at times with considerable detail, previous events, they may not know what is important to tell. One anecdotal example, obtained from a study concerning children's memory of a medical procedure that involved urethral catheterization (Goodman et al., 1997), was a child who, when asked about the procedure, proceeded to describe how his mother had been required to leave before the procedure was over. Although this feature of the event was clearly important to the child, it was not a description of the medical procedure. Much more direct probes had to be asked to obtain information from this child about the target event. Narrative elaboration is designed to help cue children to provide information that is important to others. The technique involves teaching children a method that the children can use to guide their retrieval and communication of information about an event. Specifically, children are given cue cards representing the participants, setting, actions, and conversation/affective states relevant to an experience to remind the children to include this information in their narrative reports. Saywitz and Snyder (1996) tested the efficacy of narrative elaboration training on 7- to 8-year-olds' and 10- to 11-year-olds' memory of a laboratory event. Children who received narrative elaboration training, in comparison with a control group, provided significantly more correct units of information on both free and cued recall tasks. It is important to note that the increase in the amount of information recalled occurred without a simultaneous increase in the number of false reports. The authors explained that the training may have helped children "bridge the gap" between their own level of cognitive functioning and the level of comprehension and detail expected in an adult courtroom situation. Bowen and Howie (2002) found similar benefits of narrative elaboration training on 4- to 6-year-old children's memory of a witnessed event. The benefits were evident in the amount of correct information children freely recalled about the event and again with no significant increase in errors. These studies demonstrate the potential that narrative elaboration has in improving children's testimony and hence better safeguarding the rights of all involved parties. Perhaps facets of empirically validated structured interviews, such as narrative elaboration, could be incorporated into direct examination in court. Yet, not all innovative interviewing methods and training procedures improve children's recounting of events. Poole and Lindsay (2002) engaged 4- to 8-year-olds in a source-monitoring training procedure, designed to help children distinguish between events that they actually experienced versus those that they only heard described, a task young children are often unsuccessful in accomplishing. Following training, the children were questioned about a completely separate experience, which involved participation in a series of science demonstrations. The children had been exposed, after the science demonstrations, to false suggestions about their experiences. Older children who had taken part in the source-monitoring training were less likely to report the false suggestions than were older children who had not taken part in the training. Source-monitoring training, however, did not reduce younger children's willingness to claim that the false suggestions had taken place. The finding oflack of benefits to the younger children in particular is reminiscent of similar results from studies examining the efficacy of certain other forensic interview techniques, such as ones intended to help children identify

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

561

perpetrators in photo lineups, which appear to help older but not younger children (e.g., see Goodman et al., 1991).

Summary In summary, several legal reforms have been designed to improve the accuracy and completeness of children's eyewitness memory reports. Among those discussed here, the use of Child Advocacy Centers to coordinate forensic investigations, engaging children in developmentally appropriate truth-lie discussions, child-oriented instructions, and certain structured interviewing methods appear to hold the most promise in terms of enhancing children's memory. Of importance, the precise benefits may vary, depending on the child's age. Thus, continued research is needed to determine when and how such reforms improve memory across childhood. Mnemonic aids at times improve children's accuracy but also may at times increase errors, including false reports of body touch when the aids are paired with highly leading questions. Finally, most reforms we reviewed here are employed before a case proceeds to trial. Accordingly, the reforms do not directly infringe on a defendant's rights in court. Yet, insofar as the reforms have the potential to affect children's eyewitness reports, the defendant's rights to a fair trial and the court's ability to determine the truth may be affected. As such, continued insight into the reforms' utility and limitations remains critical. Techniques that are eventually found, through sufficient scientific research, to increase the accuracy and completeness of children's eyewitness memory might conceivably be candidates for use directly in court, so that the truth can be known.

TECHNIQUES AND REFORMS AIMED AT REDUCING CHILDREN'S DISTRESS

Several of the more controversial legal reforms for child witnesses were designed not to improve children's memory per se but instead to reduce children's distress so that they can better communicate in legal settings. For the most part, these reforms are implemented in court and involve reducing or eliminating aspects of court hearings that children find most distressing: facing the defendant, testifying in open court, not understanding the legal process, and not having a supportive individual available when testifying. Because these reforms can affect the actual trial proceedings, their potential to infringe on a defendant's right to face witnesses against him or her is obvious, and considerable scholarly debate and case precedence have focused on the legality of the reforms. It is worth noting, however, that the reforms are most controversial at trial, where defendants' rights are strictly enforced. Particularly in the United States, the use of reforms may be more possible in preliminary hearings because defendants do not have full constitutional rights at these hearings. In other countries, the use of and controversies regarding reforms also vary, depending on the type and phase of a trial. For instance, in England and Scotland, videotaped testimony can, at times, be presented in lieu of a

562

MALLOY ET AL.

child's live testimony during evidence-in-chief (i.e., direct examination), even though the child may still be required to testify live (although possibly via closed-circuit television) during cross-examination.

Videotaped Forensic Interviews and Hearsay Testimony Two forms of hearsay testimony, presenting a videotaped forensic interview at trial in place of a child's live testimony (Bernstein & Claman, 1986; Myers, Cordon, Ghetti, & Goodman, 2002) and having a "declarant" (e.g., the child's mother, a police officer, a child protective service worker) repeat in court what a child said earlier (out of court) about the alleged crime (Myers, 1996), have been proposed for use with child witnesses. A number of benefits may be associated with videotaped testimony presented via hearsay. First, Henry (1999) found that children who had their disclosures videotaped did not have to endure as many interviews (a source of considerable distress for children), nor were they asked to testify as frequently (another source of distress for children) as were children whose disclosures were not videotaped. Second, because videotaped interviews typically occur long before trials, it has been argued that this method results in more complete reports (Bernstein & Claman, 1986). That is, when children have to wait for weeks, months, or more before testifying, the delay may decrease the amount and accuracy of information recalled because memory representations fade with time. Third, when concerns have been raised about how a child was questioned during a forensic interview, the fact finders themselves have the opportunity to view the child's disclosure during the interview. By viewing the original interview, fact finders may be in a better position to evaluate the interviewer's questions and suggestions and the child's answers; this is a procedure that would theoretically enhance the fact finders' decisionmaking abilities. Indeed, the defense bar has long called for videotaping of forensic interviews with children, and thus presenting the videotape at trial seems to be potentially less problematic in terms of defendants' rights than some other reforms (e.g., a hearsay witness testifying, closed-circuit television testimony). Even though the courts have been open to the second form of hearsay testimony (i.e., having an individual recount a child's out-of-court statements in court) under some circumstances, the use of this reform is much more controversial, at least from a defense attorney perspective, than is the admittance of videotaped hearsay. In particular, the courts have recognized that, in sexual abuse cases, a child's initial disclosure of abuse to a parent, teacher, or other trusted adult may provide a more detailed account of the alleged incident than does the child's testimony in court months or even years later. As described by Whitcomb (1992), "In cases of child sexual abuse, the child's out-of-court statements may be the most compelling evidence available. Indeed, hearsay may be the only evidence, since child sexual abuse frequently occurs in the absence of other witnesses or physical trauma to the child, and the child may be found incompetent or otherwise unavailable as a witness" (p. 85). Many people, including parents, police officers, physicians, forensic interviewers, and social workers, may be asked to testify regarding

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

563

the information a child has told them about suspected abuse. The admittance of these statements may benefit children by reducing or eliminating the need for them to testify. Moreover, because children's initial utterances had occurred possibly prior to multiple interviews and preparation by legal professionals, the utterances may be more accurate and hence facilitate justice. Despite potential benefits, a number of concerns have been raised about hearsay testimony. One is that that the child is not present in the courtroom during the trial and is not available for cross-examination, a process often relied upon heavily by opposing attorneys to support their case. For instance, a defense attorney might be inhibited from sufficiently pointing out inconsistencies between the child's testimony and the child's prior statements that may indicate less than completely truthful or accurate testimony. Although cross-examination itself may not enhance children's accuracy, as discussed earlier (see Zajac & Hayne, 2003), it may be a critical means for defense attorneys to mount a vigorous defense for their client. A second critical concern, which is specific to the second form of hearsay (i.e., when another individual recounts a child's out-of-court statements), is whether the reporting individual is accurate in her/his description of a child's statements. In one of the first studies on adults' accuracy in recounting conversations with young children, Bruck, Ceci, and Francoeur (1999) found that mothers who interviewed their preschool children were particularly good at recalling the gist of their children's statements (88% accuracy). However, mothers often misreported utterances as spontaneous when in fact they had been the result of prompted questioning. Even when interviewers are trained professionals, information may be lost when interviewers recount children's experiences. Warren and Woodall (1999), for example, found that, although experienced interviewers were accurate most of the time when recalling the gist of children's statements, they rarely included verbatim questions and answers in their accounts. The third concern regarding the admittance of hearsay testimony focuses on fact finders' ability to discern the truth. Specifically, by not viewing the witness, concerns have been raised that fact finders lack adequate information to discern truthful from inaccurate testimony. Research comparing jurors' decisions when presented with hearsay witnesses versus children's live testimony has produced mixed results. Some studies have found that, compared with having a child testify, having an adult hearsay witness testify resulted in greater believability of the child's claims, greater perceived guilt of the defendant, and/or greater likelihood of the defendant's conviction (Golding, Sanchez, & Sego, 1997; Ross, Lindsay, & Marsil, 1999; Warren, Nunez, Kenney, Buck, & Smith, 2002). However, other studies indicate that children's in-court testimony may be more important in predicting defendant guilt than an adult's presentation of children's outof-court statements, in both actual sexual abuse cases and mock trial studies (Goodman et al., 2004; Myers et al., 1999). Finally, Warren et al. (2002) found that jurors are more likely to believe false information when provided a "gist" account of an interview with a child than when they are provided with the exact words of that child. When another individual recounts a child's statements, that individual may well provide a gist of the child's story, hence leading to potential difficulties discerning the truth on the part of jurors.

564

MALLOY ET AL.

A final controversy regarding the use of hearsay stems from a recent U.S. Supreme Court ruling (Crawford v. Washington, 2004). The Supreme Court case did not itself concern children. Instead, it concerned assault and attempted murder and whether statements a woman made to police were admissible under one of the hearsay exceptions. The majority ruled that, because the statements were "testimonial," they were not admissible. Unfortunately, the court did not explicitly define what types of statements are considered "testimonial," although courts seem to be interpreting the term to mean any interview conducted for forensic purposes. In general, the need to define precisely testimonial hearsay is now being hotly debated in both federal and state courts (e.g., People v. Cervantes, 2004). In fact, the Supreme Court is set to issue two post-Crawford decisions this year that will elucidate the meaning of testimonial hearsay. It is unclear at present how these decisions will affect hearsay exceptions involving child witnesses. However, from a psychological perspective, the logic behind the court's distinction between testimonial and non testimonial hearsay does not seem justified. Generally, in light of empirical research supporting the use of hearsay, as a means of reducing children's distress, increasing the accuracy of their testimony, and facilitating jurors' fact-finding abilities, there is reason to remain supportive of the potential utility of admitting children's out-of-court statements as evidence, whether they are testimonial or not. In fact, if the "testimonial" statements are recorded on videotape, as would be most likely to occur during a forensic interview rather than during a nontestimonial setting, one can particularly question the court's logic because the ruling could preclude admittance of the videotaped forensic interview. In closing, a number of countries, such as England, regularly permit children's direct examination to be in the form of a videotaped forensic interview that is shown in court. In the United States, in contrast, videotaped forensic interviews are used in only about 17% of child sexual abuse trials. Moreover, the videotapes are almost always included in conjunction with rather than instead of having the child provide live testimony (Myers et al., 1999). Given that children may be more likely to be believed if they testify live (Goodman et al., 2004), putting children on the stand may actually be critical to the prosecution, even if it is distressing to children. To date, there is no clear, convincing evidence that admitting only children's videotaped testimony, having another individual recount children's statements, or having children testify live in open court systematically affects the truth-seeking function of the courts. Nonetheless, given the benefits of obtaining a videotape of forensic interviews, in terms of both preserving children's initial reports and highlighting the tactics employed in the interview, we can tentatively conclude that presenting videotaped forensic interviews in court represents one reform that is likely to assist (although not guarantee) the courts' pursuit of justice. Of course, the final outcomes depend on what fact finders do with the information and, specifically, how accurately judges and juries can evaluate live testimony, hearsay, and other evidence.

Closed-Circuit Television One of the most highly researched legal reforms proposed for use with children is closedcircuit television (CCTV). This technology permits children to give testimony and sub-

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

565

mit to cross,examination (thereby partially protecting defendants' right to confronta, tion) in an area outside of the courtroom. Testimony is then viewed, from the court, room, by the judge, jurors, defendant, and others, on a television monitor. Unlike the presentation of hearsay evidence, CCTV allows children to be cross,examined. Yet, CCTV remains controversial, especially in counties like the United States, that are con, stitutionally committed to a defendant's rights to face,to,face confrontation of wit, nesses. For instance, one claim is that not viewing a witness in entirety and live inhibits fact finders' ability to assess the veracity of the witness's statements. Furthermore, face, to,face confrontation, in theory, compels a witness to tell the truth. By testifying via CCTY, such pressure is eliminated, which could increase the potential for inaccuracies and falsehood (concerns that are also raised about presenting videotaped forensic inter, views in court in place of the child). Controversy in the United States is further evident in that even the Supreme Court has weighed in on the use of CCT\1. In Maryland v. Craig (1990), the majority specified that CCTV can be used if the child will be so trau, matized by facing the accused in court that the child could not reasonably communi, cate. Some courts, however, have interpreted the Maryland v. Craig ruling to apply only in child sexual abuse cases, highlighting continued debate about the conditions under which the reform can be implemented. As mentioned, considerable research has focused on the use of CCT\1. The research has addressed two related questions. The first is whether CCTV reduces children's dis, tress and facilitates their ability to communicate, and the second is whether CCTV af, fects fact finders' ability to reach the truth. A number of studies reveal in fact that via CCTV is effective at both reducing anxiety and increasing accuracy when children tes, tify. According to observer ratings, children who testify via CCTV appear more relaxed and are able to provide more information than children whose testimony does not in, valve this reform (e.g., Davies & Noon, 1991, 1993). Legal professionals (e.g., judges, court clerks, barristers) also perceive reduced stress and enhanced memory performance when child witnesses testify via CCTV and rate children who testify via CCTV as more confident, consistent, and happy than children who testify in a typical open court set, ting (Davis & Noon, 1993). Similarly, in a national survey in New Zealand, Pipe and Henaghan (1996) found that, even though defense attorneys, prosecutors, judges, and social workers differed in their perceptions of the fairness of CCTV, all reported that allowing children to testify via CCTV would reduce their distress relative to testifying live. Finally, in a mock,trial study, Goodman et al. (1998) found that children who testi, fied in open court were more anxious about testifying than were children who testified via CCTY, although such was the case only when the defendant was actually guilty of the mock crime. A trend in the study by Goodman et al. further suggested that children asked to testify in open court were more likely to refuse to testify than those children permitted to testify via CCTV. Only one study to date has found that some (not all) chil, dren prefer to testify in open court (Cashmore, 1992). Because it is critical to safeguard innocent individuals from wrongful prosecution, the effects of CCTV on children's accuracy and fact finders' decision,making must also be considered. Goodman et al. (1998) and Orcutt, Goodman, Tobey, Batterman,Faunce, and Thomas (2001) tested the effects of CCTV versus open,court testimony on chil,

566

MALLOY ET AL.

dren's accuracy and/or deception about a mock crime. Results demonstrated the potential efficacy of this reform for the truth,seeking function of the courts generally and trials specifically: children who testified via CCTV rather than in open court were more accu, rate when answering direct questions, and younger children were also less suggestible when questioned in the CCTV condition. Furthermore, CCTV did not bias jurors against the defendant. In fact, jurors were no better at ascertaining whether children were telling the truth or lying when they testified in open court versus through CCTY. The main bias so far may actually be a sometimes diminished credibility, in the eyes of jurors, of children who testify via CCTV (Orcutt et al., 2001). Despite research supporting CCTV as a means of reducing children's distress and increasing children's accuracy, which ultimately could lead to fairer justice for all involved, its actual use in the United States is quite limited. The U. S. Supreme Court's decision in Maryland v. Craig specifies that CCTV can only be used if the child will be so traumatized by facing the accused in court that the child cannot reasonably communicate. Such a standard is particularly difficult to demonstrate before a child actually testifies. Additionally, as mentioned, some courts have interpreted the Maryland v. Craig ruling as applying only to child sexual abuse cases. Accordingly, in a murder case (e.g., the child witnesses a death related to domestic violence), CCTV might not be an op, tion. In contrast, CCTV is often used in England, particularly during cross,examination, because of the less restrictive guidelines in the 1988 Criminal Justice Act. However, prosecutors' intuition that children's testimony has less impact when children testify via CCTV may have some validity (Goodman et al., 1998, 1999), and this concern appears to further limit the use of CCTV in the United States.

Support Person In actual trials, unless special circumstances exist, such as when it can be demonstrated that a child will be too traumatized to communicate effectively in court, child witnesses must typically provide live in,court testimony while facing the defendant and submit, ting to direct examination and cross examination. The availability of a support person while children testify is one legal reform that may make this necessary task less distressing for children and hence improve their willingness and ability to provide accurate tes, timony. The support person is typically a victim advocate who has built rapport (it is hoped) with the child and is permitted to escort the child into court and stand next to or directly behind the child while the child is on the witness stand. The presence of a support person is one of the more frequently used, although rarely researched, in-court reforms (Goodman et al., 1992, 1999). The research that has been conducted, however, highlights the benefits of a support person on children's emotional reactions and memory performance. Hall (2001), for instance, investigated the effects of social support during investigative interviews on children's stress and well-being following the interviews. Children with at least one socially supportive individual present during the forensic interviews had significantly lower internalizing problems (e.g., anxi, ety, withdrawal) and depression ratings after the interviews. Goodman et al. (1992) found that, among children testifying in criminal cases because of sexual abuse, the

19. BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

567

presence of a supportive adult in the courtroom was positively related to children's abil, ity to answer attorneys' questions. Similarly, Sas (1991) reported that the presence of a supportive adult when children testify serves as a stress,reducing technique that con, currently increases children's self,confidence and performance. The research of both Goodman et al. (1992) and Sas (1991) focused on actual court proceedings; thus, it was not possible from these two studies to assess the effects of a supportive adult on the ac, curacy of children's statements. It will be necessary, in future studies, to conduct such an empirical test. Although the courts have generally allowed the use of support persons in court with children and do not feel that the individual interferes with a defendant's rights to a fair trial, there is some opposition to the use of such individuals. Specifically, opponents argue that children's answers may be influenced by the mere presence of a support person. For instance, a child might be more willing to provide false testimony in the presence of someone who, in effect, works for the side that calls the witness. This possibility has not been tested empirically. However, given the widespread use of support persons, including the court's endorsement of their use, additional research on this reform is needed.

Preparation The final legal reform we review involves preparing children for legal participation, both generally throughout a legal case and for testifying specifically. In a national survey of U.S. district attorney offices, prosecutors rated this technique as one of the most widely used and effective methods of reducing stress in children. They also perceived court preparation to be particularly inexpensive and easy to implement and indicated that it rarely raised constitutional challenges or concerns about defendants' rights (Goodman etal., 1999). Empirical studies confirm many of the prosecutors' beliefs. Preparation and enhanced knowledge about impending stressors reduce children's distress behaviors in a variety of contexts (e.g., medical: Chen, Zeltzer, Craske, &Katz, 1999; Quas et al., 1999). Also, in the study by Goodman et al. (1998) of the effects of testifying in open court versus via CCTV on children's memory and fact finders' decision making, even with age controlled statistically, increases in children's legal knowledge were associated with decreased anx, iety about testifying and improved memory performance while testifying. Some districts have developed and tested thorough legal preparation programs for children. The London Family Court Clinic in Ontario, Canada, for example, examined the effects of in,depth preparation programs on abused children's experiences in court and general emotional well,being (Sas, Wolfe, Gowdey, 1996; Westcott & Davies, 1993). The program included courtroom tours, role,playing, and drawing and writing exercises about coming to court. Children who received pre,court preparation were significantly better adjusted after the conclusion oflegal proceedings than were children who had not engaged in the preparation program. Of importance, Sas (1991; Sas et al., 1996) empha, sized that, in addition to providing children general information about the legal process, preparation programs should also provide children with methods to deal with the stress that they may experience during the actual courtroom proceedings. The researchers

568

MALLOY ET AL.

hired specially trained therapists to teach child witnesses stress and anxiety-reducing techniques (e.g., breathing relaxation exercises; cognitive empowerment, which included listing strengths and positive reasons for attending court; therapeutic support, which included words of encouragement and advocacy; and development of a fear hierarchy, which included listing the child's five most relevant and anxiety-provoking fears). The researchers reported that these techniques equipped children with strategies to cope with stress and anxiety due to both abuse and courtroom proceedings, thereby enabling them to provide full evidence on the stand. Although Sas and colleagues did not systematically investigate the effect of court preparation programs or anxiety-reducing strategies on children's memory performance or accuracy in court, other evidence, for instance, from studies of the effects of knowledge on children's medical procedure distress and memory, suggests that preparation is likely to benefit children's accuracy. In general, providing children (and parents) with realistic expectations and accurate information about the legal process is an important and potentially effective method of reducing at least some of their stress and therefore enhancing their memory. Moreover, parents and children frequently recognize the helpfulness of programs that prepare them with information about what to expect during the legal process (e.g., Lipovsky, Tidwell, Kilpatrick, Saunders, & Dawson, 1991; Westcott & Davies, 1993). Insofar as children (and their families) are more willing participants, children are less distressed, and children can recount the full evidence accurately and completely, the rights of both children and defendants are better protected. Summary

Legal reforms designed to reduce children's anxieties and facilitate their communication while children provide their eyewitness accounts generally appear effective in doing so. Research consistently suggests that not having to face the defendant, for example, by presenting hearsay testimony or testifying via CCTV in lieu of live testimony, is associated with less anxiety in children. In these situations, research to date suggests that children's reports often are more accurate than they are when children testify in open court, face to face with the defendant. Results concerning fact finders' abilities to reach the truth reveal that neither the provision of hearsay nor CCTV leads to any evident biases against defendants. Instead, such reforms may, at times, lead to reductions in children's perceived credibility, an important issue in and of itself. Finally, although the use of support persons and courtroom preparation programs is endorsed by the courts as not infringing upon defendants' rights, to date little empirical research has focused on whether such reforms affect fact finders' decision-making. Given that these legal reforms are widely used, research examining their utility is needed.

CONCLUSIONS

In this review, we identified a subset of legal techniques and reforms that have been proposed for use, and actually used, to facilitate children's participation in forensic investi-

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

569

gations and legal proceedings. The overarching goal of the techniques and reforms is to facilitate the pursuit of justice. They attempt to do this by helping children remember information or by reducing children's stress so that they can better communicate. Indeed, research indicates that several facets of legal involvement, especially being subjected to multiple interviews, not understanding the legal process, testifying while facing a defendant, and being subjected to cross-examination, are distressing to children. This increased distress is associated with adverse effects on children's ability to remember and communicate while trying to provide eyewitness memory reports. Of the legal techniques and reforms designed to improve children's memory, Child Advocacy Centers, truth-lie discussions, child-oriented instructions, and innovative interviewing practices appear to be the most promising in terms of enhancing children's memory performance and reducing inaccuracies in their reports, especially during forensic interviews. Mnemonic aids are more variable in that, although they may enhance children's reports of true experiences, particularly those that involve body touch, they may also increase errors, especially among young children and when used in conjunction with misleading suggestions. Legal reforms designed to reduce children's anxiety and improve communication appear to be generally effective in doing so and, of importance, do not appear (at least based on research to date) to inhibit either children's memory accuracy or, for hearsay and CCTV, the ability of the jurors to reach the truth-an ability that may be surprisingly limited in the first place (but see Ball & O'Callaghan, 2001). More research is needed, however, on whether or not, and in what ways, the techniques and legal reforms we have described affect fact finders' judgments. Of course, the use of any forensic technique or legal reform as a means of benefiting children must be weighed against the rights of defendants. Our review highlights that several techniques and reforms indeed may benefit the pursuit of justice. They do not adversely affect children's memory or unduly bias jurors for (or against) children. Insofar as children can be protected while defendants' rights are maintained, the legal system can better fulfill its truth-seeking goal.

REFERENCES Aldridge, J., Lamb, M., Sternberg, K., Orbach, Y., Esplin, P., &Bowler, L. (2004). Using a human figure drawing to elicit information from alleged victims of child sexual abuse. Journal of Consulting and Clinical Psychology, 72, 304-316. Astington, J. W. (1988). Promises: Words or deeds. First Language, 8, 259-270. Australian Institute of Health and Welfare (2004). Child Protection Australia 2002-2003. Child Welfare Series, No. 34. Canberra, ACT: Australian Institute of Health and Welfare. Ball, C., & O'Callaghan, J. (2001). Judging the accuracy of children's recall: A statement-level analysis. Journal of Experimental Psychology: Applied, 7, 331-345. Bernstein, B. E., & Claman, L. (1986). Modern technology and the child witness. Child Welfare, 65, 155-163. Bjorklund, D. (2005). Children's thinking, 4th Edition. Belmont, CA: Wadsworth. Block, S., Oran, D., Goodman, G. S., & Oran, H. (2005, April). Maltreated children in dependency court. Paper presented at the biennial meeting of the Society for Research in Child Development, Atlanta, GA.

570

MALLOY ET AL.

Bowen, C.]., & Howie, P.M. (2002). Context and cue cards in young children's testimony: A comparison of brief narrative elaboration and context reinstatement. Journal of Applied Psychowgy, 87, 1077-1085. Bruck, M., Ceci, S. ]., & Francoeur, E. (1999). The accuracy of mother's memories of conversations with their preschool children. Journal of Experimental Psychowgy: Applied, 5, 89-106. Bruck, M., Ceci, S., Francoeur, E., & Renick, A. (1995). Anatomically detailed dolls do not facilitate preschoolers' reports of a pediatric examination involving genital touch. Journal of Experimental Psychology: Applied, 1, 95-109. Bruck, M., Melnyk, L., & Ceci, S. ]. (2000). Draw it again Sam: The effect of drawing on children's suggestibility and source monitoring ability. Journal of Experimental Child Psychowgy, 77, 169-196. Buchanan, T. W., & Lovallo, W. R. (2001). Enhanced memory for emotional material following stress-level cortisol treatment in humans. Psychoneurendocrinology, 26, 307-317. Bussey, K. (1992). Lying and truthfulness: Children's definitions, standards, and evaluative reactions. Child Development, 63, 129-137. Butler, S., Gross,]., & Hayne, H. (1995). The effect of drawing on memory performance in young children. Devewpmental Psychology, 31, 597-608. Cashmore, J. (1992). The use of closed-circuit television for child witnesses in the ACT. Sydney, Australia: Australian Law Reform Committee. Chapman, A. (2002). When children testify against their parents. Cityscape. Retrieved January 4, 2005 from http://www.jrn.columbia.edu/studentwork/cityscape/ Chapman, M. V., Wall, A., & Barth, R. P. (2004). Children's voices: The perceptions of children in foster care. American Journal of Orthopsychiatry, 74, 293-304. Chen, E., Zeltzer, L. K., Craske, M.G., & Katz, E. R. (1999). Alteration of memory in the reduction of children's distress during repeated aversive medical procedures. Journal of Consulting & Clinical Psychowgy, 67, 481-490. Christianson, S. (1992). Emotional stress and eyewitness memory: A critical review. Psychowgical Bulletin, 112, 284-309. Cordon, I. M., Goodman, G. S., & Saetermoe, C. (2005). Facilitating children's accurate responses: Conversational rules and interview style. Applied Cognitive Psychowgy, 19, 249-266. Crawford v. Washington, 541 U.S. 36; 124 S. Ct. 1354 (2004). Davies, G., & Noon, E. (1991). An evaluation of the live link for child witnesses. London: Home Office. Davies, G., & Noon, E. (1993). Video links: Their impact on child witness trials. Issues in Criminological & Legal Psychowgy, 20, 22-26. Dent, H. R. (1977). Stress as a factor influencing person recognition in identification parades. Bulletin of the British Psychowgical Society, 30, 339-340. Drucker, P.M., Greco-Vigorito, C., Moore-Russell, M., Avaltroni, J., & Ryan, E. (1997, April). Drawing facilitates recall of traumatic past events in young children of substance abusers. Paper presented at the biennial meeting of the Society for Research in Child Development, Washington, DC. Everson, M. D., & Boat, B. W. (2002). The utility of anatomical dolls and drawings in child forensic interviews. In M. L. Eisen, J. A. Quas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview. (pp. 383-408). Mahwah, NJ: Lawrence Erlbaum Associates. Feiring, C., Taska, L., & Chen, K. (2002). Trying to understand why horrible things happen: Attribution, shame, and symptom development following sexual abuse. Child Maltreatment, 7, 26-41. Feiring, C., Taska, L., & Lewis, M. (1999). Age and gender differences in children's and adolescents' adaptation to sexual abuse. Child Abuse & Neglect, 23, 115-128. Fivush, R. (1993). Developmental perspectives on autobiographical recall. In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving children's testimony (pp. 1-24). New York: Guilford.

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

571

Garven, S., Wood, J. M., Malpass, R. S., & Shaw, J. S. (1998). More than suggestion: The effect of interviewing techniques from the McMartin preschool case. Journal of Applied Psychology, 83, 347-359. Ghetti, S., Alexander, K. A., & Goodman, G. S. (2002). Legal involvement in child sexual abuse cases: Consequences and interventions. International Journal of Law & Psychiatry, 25, 235-251. Golding, J. M., Sanchez, R. P., & Sego, S. A. (1997). The believability of hearsay testimony in a child sexual trial. Law & Human Behavior, 2I, 299-325. Goodman, G. S., Bottoms, B. L., Schwartz-Kenney, B. M., & Rudy, L. (1991). Children's testimony about a stressful event: Improving children's reports. Journal of Narrative & Life History, I, 69-99. Goodman, G. S., Myers, J. E. B., Qin, J. J., Quas, J. A., Castelli, P., Redlich, A., & Rogers, L. (in press). Children's out-of-court statements: Effects of truthful and deceptive hearsay on jurors' decisions. Law and Human Behavior. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1997). Children's reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge, and parental attachment. Applied Developmental Science, I, 54-75. Goodman, G. S., Quas, J. A., Bulkley, J., & Shapiro, C. (1999). Innovations for child witnesses: A national survey. Psychology, Public Policy, & Law, 5, 255-281. Goodman, G. S., Taub, E. P., Jones, D. P. H., England, P., Port, L. P., Rudy, L., & Prado, L. (1992). Testifying in criminal court: Emotional effects on child sexual assault victims. Monographs of the Society for Research in Child Development, 57 (Serial No. 229). Goodman, G. S., Tobey, A. E., Batterman-Faunce, J. M., Orcutt, H., Thomas, S., Shapiro, C., & Sachsenmaier, T. (1998). Face to face confrontation: Effects of closed-circuit technology on children's eyewitness testimony and jurors' decisions. Law & Human Behavior, 22, 165-203. Goodman-Brown, T. B., Edelstein, R. S., Goodman, G. S., Jones, D. P. H., & Gordon, D. S. (2003). Why children tell: A model of children's disclosure of sexual abuse. Child Abuse & Neglect, 27, 525-540. Gordon, B. N., Ornstein, P. A., Nida, R. E., Follmer, A., Crenshaw, M. C., & Albert, G. (1993). Does the use of dolls facilitate children's memory of visits to the doctor? Applied Cognitive Psychology, 7, 459-474. Grisso, T., Steinberg, L., Woolard, J., Cauffman, E., Scott, E., Graham, S., Lexcin, F., Reppucci, N.D., & Schwartz, R. (2003). Juveniles' competence to stand trial: A comparison of adolescents' and adults' capacities as trial defendants. Law and Human Behavior, 27, 333-363. Gross, J., & Hayne, H. (1999). Young children's recognition and description of their own and others' drawings. Developmental Science, 2, 476-489. Hall, S. R. (200 1). Assessing emotional distress in abused children during videotaped investigative interviews: The effect of risk and protective factors. Unpublished doctoral dissertation, University of California, Los Angeles. Harbeck-Weber, C., & McKee, D. H. (1995). Prevention of emotional and behavioral distress in children experiencing hospitalization and chronic illness. In M. C. Roberts (Ed.), Handbook of pediatric psychology (2nd ed., pp. 167-184). New York: Guilford. Haugaard, J. J., Reppucci, N. D., Laird, J., & Nauful, T. (1991). Children's definitions of the truth and their competency as witnesses in legal proceedings. Law & Human Behavior, IS, 253-271. Henry, J. (1997). System intervention trauma to child sexual abuse victims following disclosure. Journal oflnterpersonal Violence, I2, 499-512. Henry, J. (1999). Videotaping child disclosure interviews: Exploratory study of children's experiences and perceptions. Journal of Child Sexual Abuse, 8, 35-49. Herman, J., & Hirschman, L. (1981). Families at risk for father-daughter incest. AmericanJoumal of Psychiatry, 138, 967-970. Hill, P. E., & Hill, S.M. (1987). Videotaping children's testimony: An empirical view. Michigan Law Review, 85, 809-833.

572

MALLOY ET AL.

Huffman, M. L., Warren, A. R., & Larson, S.M. (2001). Discussing truth and lies in interviews with children: Whether, why, and, and how? In R. Bull (Ed.), Children and the law: The essential readings. Essential readings in developmental psychology (pp. 225-246). Malden, MA: Blackwell. Innocence Project. (2006). Mistaken eyewitness identification: The problem. Retrieved June 10, 2006 from http: I/www. innocenceproject.org/causes/mistakenid. php Jones, L. M. (2004, July). Service access patterns for child abuse victims across communities. Paper presented at the American Psychological Association Convention, Honolulu, HI. Jones, L. M., Finkelhor, D., & Kopiec, K. (2001). Why is sexual abuse declining? A survey of state child protection administrators. Child Abuse & Neglect, 25, 1139-1158. Jones, D.P., & Krugman, R. D. (1986). Can a three-year-old child bear witness to her sexual assault and attempted murder? Child Abuse & Neglect, 10, 253-258. Justis-og Politidepartementet. (2004). Dommeravh0r og observasjon av bam. Evaluering fra arbeidsgruppe. Rapport, 2004. Oslo: Author. Kaplan, D. L. (2001). Young maltreated children's perceptions of their placement experiences. Dissertation Abstracts International, 61, p. 2593. Leippe, M. R., Manion, A. P., & Romanczyk, A. (1992). Eyewitness persuasion: How and how well do fact finders judge the accuracy of adults' and children's memory reports? Journal of Personality & Social Psychology, 63, 181-197. Lipovsky, J. A., Tidwell, R. P., Kilpatrick, D. G., Saunders, F. F., & Dawson, V. L. (1991, August). Children as witnesses in criminal court: Examination of current practices. Paper presented at the American Psychological Association Convention, San Francisco, CA. Loftus, E., & Ketcham, K. (1991). Witness for the defense: The accused, the eyewitness, and the expert who puts memory on trial. New York: St. Martin's Press. London, K., & Nunez, N. (2002). Examining the efficacy of truth/lie in predicting and increasing the veracity of children's reports. ]oumal of Experimental Child Psychology, 83, 131-14 7. Lyon, T. D. (2000). Child witnesses and the oath: Empirical evidence. Southern California Law Review, 73, 1017-1074. Lyon, T. D. (2002). Child witnesses and the oath. In Westcott, H.L., Davies, G. M., & Bull, R. H. C. (Eds.), Children's testimony: A handbook of psychological research and forensic practice (pp. 245-260). West Sussex, England: Wiley. Lyon, T. D., & Dorado, J. (2005). The effects of the oath and reassurance on maltreated children's true and false reports of a minor transgression. Manuscript submitted for publication. Lyon, T. D., Malloy, L. C., Talwar, V., & Quas, J. A. (2004, March). Truth induction: Reducing maltreated children's lies and secrets. Paper presented at the American Psychology-Law Society Biennial Meeting, Phoenix, AZ. Lyon, T. D., & Saywitz, K. J. (1999). Young maltreated children's competence to take the oath. Applied Developmental Science, 3, 16-2 7. Maryland v. Craig, 497 U.S. 836 (1990). McGaugh, J. L. (2003). Memory and emotion: The making of lasting memories. New York: Columbia University Press. Melinder, A. (2004). Perspectives on children as witnesses. Unpublished doctoral dissertation. University of Oslo, Norway. Mulder, M. R., & Vrij, A. (1996). Explaining conversation rules to children: An intervention study to facilitate children's accurate responses. Child Abuse & Neglect, 20, 623-631. Myhr, R., Melinder, A., Gabrielsen, R., Andreassen, 0., & Pettersen, T. (2004). Interviewing children for the court: Evaluation of the regulations of judges interviews and observations. Oslo: Norwegian Ministry ofJustice and Police. Myers, J. E. B. (1992). Legal issues in child abuse and neglect. Interpersonal violence: The practice series. Thousand Oaks, CA: Sage. Myers, J. E. B. (1996). A decade of international reform to accommodate child witnesses. Criminal]ustice & Behavior, 23, 402-422.

19.

BALANCING CHILDREN'S NEEDS AND DEFENDANTS' RIGHTS

573

Myers, J. E. B. (1998). Legal issues in child abuse and neglect practice (2nd ed.). Thousand Oaks, CA: Sage. Myers, J. E. B., Cord6n, I., Ghetti, S., & Goodman, G. S. (2002). Hearsay exceptions: Adjusting the ratio of intuition to psychological science. Law and Contemporary Problems, 65, 3-46. Myers, J. E. B., Redlich, A. D., Goodman, G. S., Prizmich, L. P., & Imwinkelried, E. (1999). Jurors' perceptions of hearsay in child sexual abuse cases. Psychology, Public Policy, & Law, 5, 388-419. Nathanson, R., & Saywitz, K. J. (2003). The effects of the courtroom context on children's memory and anxiety. Journal of Psychiatry & Law, 31, 67-98. Oran, H. (1989). Abused and neglected children in juvenile court. Unpublished doctoral dissertation. University of California, Los Angeles. Orcutt, H. K., Goodman, G. S., Tobey, A. E., Batterman-Faunce, J. M., & Thomas, S. (2001). Detecting deception in children's testimony: Factfinders' abilities to reach the truth in open court and closed-circuit trials. Law & Human Behavior, 25, 339-372. Pardeck, J. T. (1984). Multiple placements of children in foster family care: An empirical analysis. Social Work, 29, 506-509. People v. Cervantes, 118 Cal. App. 4th 162, 12 Cal. Rptr. 3d 774 (2004). Peters, D. P. (1991). The influence of stress and arousal on the child witness. In J. Doris, The suggestibility of children's recollections (pp. 60-76). Washington, DC: American Psychological Association. Pipe, M., Gee, S., & Wilson, C. (1993). Cues, props, and context: Do they facilitate children's event reports? In G. S. Goodman & B. L. Bottoms (Eds.), Child victims, child witnesses: Understanding and improving testimony (pp. 25-45). New York: Guilford. Pipe, M., & Henaghan, M. (1996). Accommodating children's testimony: Legal reforms in New Zealand. Criminal Justice and Behavior, 23, 377-401. Poole, D. A., & Lindsay, S.D. (2002). Reducing child witnesses' false reports of misinformation from parents. Journal of Experimental Child Psychology, 81, 117-140. Price, D. W., & Goodman, G. S. (1990). Visiting the wizard: Children's memory for a recurring event. Child Development, 61, 664-680. Quas, J. A., Bauer, A., & Boyce, W. T. (2004). Physiological reactivity, social support, and memory in early childhood. Child Development, 75, 797-814. Quas,J. A., Goodman, G. S., Bidrose, S., Pipe, M., Craw, S., &Ahlin, D. S. (1999). Emotion and memory: Children's long-term remembering, forgetting, and suggestibility. Journal of Experimental Child Psychology, 72, 235-270. Quas, J. A., Goodman, G. S., Ghetti, S., Alexander, K. W., Edelstein, R., Redlich, A., Cordon, I. M., &Jones, D.P. H. (2005). Childhood sexual assault victims: Long-term consequences after testifying in criminal court. Monographs of the Society for Research in Child Development, 70 (Serial No. 280). Quas, J. A., Goodman, G. S., & Jones, D. P. H. (2003). Predictors of attributions of self-blame and internalizing behavior problems in sexually abused children. Journal of Child Psychology & Psychiatry, 44, 723-736. Quas, J. A., & Lench, H. (in press). Effects of interviewer support and arousal on children's memory for emotion-eliciting events. Applied Cognitive Psychology. Roediger, H. L., & Thorpe, L.A. (1978). The role of recall time in producing hypermnesia. Memory & Cognition, 6, 296-305. Ross, D. F., Lindsay, R. C. L., & Marsil, D. F. (1999). The impact of hearsay testimony on conviction rates in trials of child sexual abuse: Toward balancing the rights of defendants and child witnesses. Psychology, Public Policy, & Law, 5, 439-455. Sas, L. (1991). Reducing the system-induced trauma for child sexual abuse victims through court preparation, assessment, and follow-up. Final Report for the National Welfare Grants Division, Health, and Welfare, Ottawa, Canada.

574

MALLOY ET AL.

Sas, L. D., Wolfe, D. A., & Gowdey, D. A. (1996). Children and the courts in Canada. Criminal Justice & Behavior, 23, 338-357. Saywitz, K. J., Goodman, G. S., Nicholas, E., & Moan, S. F. (1991). Children's memories of a physical examination involving genital touch: Implications for reports of child sexual abuse. Journal of Consulting & Clinical Psychology, 59, 682-691. Saywitz, K., Jaenicke, C., & Camparo, L. (1990). Children's knowledge of legal terminology. Law & Human Behavior, 14, 523-535. Saywitz, K. J., & Nathanson, R. (1993). Children's testimony and their perceptions of stress in and out of the courtroom. Child Abuse & Neglect, 17, 613-622. Saywitz, K. J., & Snyder, L. (1996). Narrative elaboration: Test of a new procedure for interviewing children. Journal of Consulting & Clinical Psychology, 64, 1347-1357. Sorensen, T., & Snow, B. (1991). How children tell: The process of disclosure in child sexual abuse. Child Welfare, 70, 3-15. State of New Jersey v. Margaret Kelly Michaels, 136 N.J. 299; 642 A.2d 1372 (1994). Steinberg, L., & Scott, E. (2003). Less guilty by reason of adolescence: Developmental immaturity, diminished responsibility, and the juvenile death penalty. American Psychologist, 58, 1009-1018. Steward, M.S., Steward, D. S., Farquhar, L., Myers, J. E. B., Reinhart, M., Welker, J., Joye, N., Driskill, J., & Morgan, J. (1996). Interviewing young children about body touch and handling. Monographs of the Society for Research in Child Development, 61, 1-186. Tedesco, J. F., & Schnell, S. V. (1987). Children's reactions to sex abuse investigation and litigation. Child Abuse & Neglect, II, 267-272. United Nations' European Convention on Human Rights, Article 6 (3). U.S. Department of Health and Human Services (2004). Child maltreatment 2002. Washington, DC: Children's Bureau, Administration for Children. Warren, A. R., Hulse-Trotter, K., & Tubbs, E. (1991). Inducing resistance to suggestibility in children. Law and Human Behavior, 15, 273-285. Warren, A. R., Nunez, N., Keeney, J. M., Buck, J. A., & Smith, B. (2002). The believability of children and their interviewer's hearsay testimony: When less is more. Journal of Applied Psychology, 87, 846-857. Warren, A. R., & Woodall, C. E. (1999). The reliability of hearsay testimony: How well do interviewers recall their interviews with children? Psychology, Public Policy, & Law, 5, 355-371. Warren-Leubecker, A., Tate, C. S., Hinton, I. D., & Ozbek, I. N. (1989). What do children know about the legal system and when do they know it? First steps down a less traveled path in child witness research. InS. J. Ceci, D. F. Ross, & M.P. Toglia (Eds.), Perspectives on the child witness (pp. 23-46). New York: Springer-Verlag. Waterman, A. H., Blades, M., & Spencer, C. (2001). Interviwing children and adults: The effect of question format on the tendency to speculate. Applied Cognitive Psychology, 15, 521-531. Weiss, E. H., & Berg, R. F. (1982). Child victims of sexual assault: Impact of court procedures. Journal of the American Academy of Child Psychiatry, 21, 513-518. Westcott, H. L., & Davies, G. M. (1993). Children's welfare in the courtroom: Preparation and protection of the child witness. Children and Society, 7, 388-396. Whitcomb, D. (1992). Legal reforms on behalf of child witnesses: Recent developments in the American courts. In H. Dent & R. Flin (Eds.), Children as witnesses. Wiley series in the psychology of crime, policing and law (pp. 151-165). Oxford, England: John Wiley & Sons. Whitcomb, D. (2003). Legal interventions for child victims. Journal of Traumatic Stress, 16, 149-157. Zajac, R., & Hayne, H. (2003). I don't think that's what really happened: The effect of crossexamination on the accuracy of children's reports. Journal of Experimental Psychology, 9, 187-195. Zeltzer, L. K., Fanurik, D., & LeBaron, S. (1989). The cold pressor pain paradigm in children: Feasibility of an intervention model: II. Pain, 37,305-313.

IV LIFESPAN EYEWITNESS ISSUES: OLDER ADULTS

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

20 The Elderly Eyewitness: A Review and Prospectus Katrin Mueller-Johnson University of Cambridge, United Kingdom

Stephen J. Ceci Cornell University

Seniors are one of the fastest-growing population segments in many developed nations. According to the American Association of Retired Persons (MRP), it is projected that by 2050 one in five Americans will be aged 65 or older (American Association of Retired Persons, 1998; Federal Interagency Forum on Aging Related Statistics, 2000). For other countries this ratio is even higher: one in four for France and Germany, and one in three for Greece, Italy, Spain, and Japan (United Nations Population Division, 2001, 2002). Because of improved economic status and access to health care, seniors now remain healthy and active for a longer time (Rothman, 2000). As their numbers grow and they remain actively involved in their communities, seniors will become increasingly important as witnesses to crimes, accidents, and transactions. For this reason, seniors' abilities as eyewitnesses have begun to receive more attention from psycho-legal researchers in recent years. In this chapter we first consider definitional issues and discuss some criminological findings on the victimization of older adults. We then describe developmental changes related to aging that are pertinent to eyewitness situations and review research focusing on older adults' performance as eyewitnesses. We conclude that although a great deal of promising work has been done in the past decade, much more needs to be done to determine which subgroups of the elderly present reliability risks, which techniques can offset such deficits, and which questions provide the most promise for future research. The categorization of who is a senior varies between studies and contexts. Some studies put the limit at age 60, others at age 65. In the USA the age at which the social security benefit system regards people as seniors is age 65 (and steadily moving upward). In gerontology research is it common to differentiate between young-old (65-74 years), middle-old (75-84 years), and old-old seniors (over 85 years) (Atchley, 2000). In the 577

578

MUELLER-JOHNSON AND CECI

psycholegalliterature a focus on older adults with this differentiation among gradations of aging is still rare. This is unfortunate because, as we show below, cognitive performances differ along these gradations in ways that are relevant for eyewitness ability.

VICTIMIZATION IN OLDER ADULTHOOD

One of the sources of information about victimization of the elderly is the National Crime Victimization Survey (NCVS). The NCVS consists of interviews with all occupants age 12 years and older of housing units in a representative sample of 42,000 households in the United States. It asks about crime victimization (e.g., rape, sexual assault, robbery, assault) for all members of the household. Although numbers from the NVCS do not show the complete picture of victimization of older adults, it is one of the best available resources on the victimization of the elderly. A review of NCVS data for the years 1992-1997 showed that older people generally experience lower rates of crime victimization than younger people. Although seniors constituted 15% of the population at the time of the data collection, they accounted only for 7% of the victimization (Klaus, 2000). However, because the NCVS does not include any crimes against the institutionalized, such as abuse in nursing homes, or fraud, such as telemarketing scams that the elderly may be particularly affected by (Moore & Lee, 2000), the 7% victimization rate is probably an underestimate. Additional information about victimization not captured by the NCVS comes from elder mistreatment research. The National Elder Abuse Study estimated that in 1996 450,000 adults over the age of 60 experienced abuse or neglect in domestic settings, not including self-neglect (National Center on Elder Abuse, 1998). NCVS data show that the types of crimes that are most frequently perpetrated against older adults differ from those of young adults. Rates of violent crime against noninstitutionalized persons are lower for the elderly. In the period between 1992 and 1997, on average 5 in 1,000 seniors experienced a violent crime in a given year, compared with 47 in 1,000 persons aged 25-49 and 105 in 1,000 persons aged 12-24. Although these are low relative rates, in absolute terms this nevertheless translates into more than 160,000 violent crimes against aged victims in this time period. Assault and robbery were the most common violent crimes committed against older adults. Robbery seems to be a particularly characteristic crime committed against them: comparing crime types across age groups, robbery accounted for a quarter of violent crimes against persons 65 and older, whereas for victims of violent crimes aged 12-64, robbery accounted for less than an eighth (Klaus, 2000). Violent crimes are only a small fraction of the victimization of the elderly, accounting for only 6% of the crimes against this group. The largest group of crimes is by far property crimes, with around 92% in the NCVS survey. This ratio is different from that of younger age groups, in which only 72% of the crimes measured in the survey were property crimes, and 23% of the victimizations were violent crimes. Although seniors experience generally lower victimization rates than younger adults, for personal theft they have the same rates as younger persons, age 25 and up, making this a crime that they are particularly prone to experience (Klaus, 2000).

20. THE ELDERLY EYEWITNESS

579

Not only do the types of crimes that seniors suffer from differ from those experi~ enced by younger adults; the circumstances of the offence are also dissimilar. Most crimes against the elderly are likely to occur in or close to their homes and during daylight hours. Compared with younger adults, older victims are less likely to use self-protective measures, such as calling for help, running away, or attacking the offender. They experienced rates of being injured during violent crime similar to that of younger adults. One in five elderly victims of nonlethal violence in the NCVS survey were injured in the attack (Klaus, 2000). However, elderly victims who are injured by a crime are more likely to need medical treatment than younger victims who are similarly injured (Lachs et al., 2001). It has also been reported by geriatric physicians that crime victimization may accelerate functional decline in old age and may be a factor in compounding mental illness (Lachs et al., 2001). Of particular interest in the current chapter is that the elderly were more likely than younger groups to report nonlethal violence and personal theft to the police. Although still about half of violent crimes in this age group (48% in the NCVS) and more than half of thefts (56% in the NCVS) went unreported, this still results in a sizable number of elderly victims and complainants confronting the legal system. In a recent study linking data from an established longitudinal study (New Haven Established Population for Epi~ demiological Studies in Elderly, EPESE) with police records (Lachs et al., 2003), 31% of their sample of over 2300 seniors had had contact with the police during a 10-year ob~ servation window from 1985 until1995. Some of them contacted the police for multiple incidents (e.g., for a burglary in one year and a robbery in another), and some of them had had contact with the police in multiple roles, such as being the victim in one incident and a witness in another, or being an offender in one and a victim in another. Most interactions with the police (60.5%) involved seniors as victims, 13.5% seniors as complainants, 8.1% as offenders, and 4.1% as witnesses to a crime. If one considers that victims usually also give witness accounts of the crime, then the majority of police contacts with older adults, here around 65%, involve obtaining a statement about a witnessed event. Contrary to what one might expect from the literature dealing specifically with elder mistreatment, seniors who were in contact with the police in the Lachs et al. (200 1, 2003) studies were slightly younger and less functionally impaired than those in the comparison group of seniors who did not have contact with the police. Both victims and witnesses were less likely to be cognitively or visually impaired. The victim, witness, and complainant groups had a higher percentage of people in excellent health than the com~ parison group, and both victims and complainants had a social network of greater size than their comparison counterparts (Lachs et al., 2003).

WITNESSING A CRIME-THE EFFECT OF AGING

Aging brings with it a variety of perceptual and cognitive changes that may influence the performance of eyewitnesses. The ability to accurately report an event (e.g., a crime) depends on three distinct stages: the initial processing of the crime, the retention of the memory for the crime until an account is made, and the production of an account of the

580

MUELLER-JOHNSON AND CECI

crime at a later time. These correspond to what cognitive scientists refer to as the encoding, storage, and retrieval stages, respectively. We describe each of these in greater detail in the following sections.

The Encoding Stage Aging brings about several changes in the sensory systems, most prominently in vision and hearing, that can affect how well an event is encoded. Vision generally worsens with age. Average visual acuity declines beginning around age 50, and by age 85 visual acuity in some adults is around 20% of what it was when they were young adults (Papalia, Sterns, Feldman, & Camp, 2002). However, for most older adults corrected vision remains good through their mid-eighties (Schieber & Baldwin, 1996). Visual acuity can be differentiated into acuity for static objects and acuity for moving objects. The acuity for moving objects declines even more with aging than that for static objects (Kline & Scialfa, 1996). Therefore tasks like identifying a person in a moving car or reading a sign on a moving object, such as a number plate, may be more difficult for older adults than for young adults. Bad lighting conditions also affect older people more than young adults. This is caused by a weakening of the muscles that dilate the pupils to admit more light in dark conditions. The pupils of the elderly do not dilate as much as they did when they were younger, and thus more light is needed to see as welL Research suggests that by middle age adults need lighting to be one-third brighter to make up for this reduction in pupil dilation when compared with young adults (Schieber & Baldwin, 1996). Older adults' pupils are also slower to dilate or contract, so this age group experiences more difficulty coping with abrupt changes in lighting conditions, such as driving into a tunnel on a bright day or stepping out into the dark from a brightly lit house. Furthermore, the ability to discriminate background and foreground in conditions of low contrast decreases with age (Yarmey, 2001), and depth perception diminishes as well (Papalia et al., 2002). Thus, especially in low light conditions, older witnesses are likely to perceive less well than younger adults. Color perception also changes with age. Around the age of 50, the lens starts turning yellow (Stuen & Faye, 2003) and increasingly (the biggest change occurs between the ages of 60 and 80; see Anderson & Palmore, 1974) filters out green, blue, and violet light (Kline & Scialfa, 1996). Consequently, the ability to differentiate between these colors, especially between blue and green, is diminished (Kline & Scialfa, 1996), which can lead to false descriptions of objects' colors. A general decrease in the sensitivity to colors has also been reported, so that colors appear less vivid with age (Stuen & Faye, 2003). Another change that comes with aging is an increased sensitivity to glare (Papalia et al., 2002; Stuen & Faye, 2003), which can be important for very bright light conditions, but also for photo-lineup procedures. The presentation of photos in these contexts should be done in a way that minimizes glare, such as by taking special care in lighting arrangements in the room or making sure that the photos are not in sheet protectors that introduce additional glare.

20.

THE ELDERLY EYEWITNESS

581

One of the most common degenerations of the visual system is presbyopia, or agerelated farsightedness (Papalia et al., 2002). Usually this does not present a problem, as it is routinely corrected, and with glasses most people see quite well. Data based on an evaluation of the inhabitants of an entire town (the Framingham Heart Study) showed that 92% of those persons aged between 65 and 74 and 70% of those aged between 75 and 85 retained corrected visual acuity of at least 20/25 in the better eye (Kannel & Gordon, 1973). Another frequent problem in the elderly is cataracts, leaving them with clouded and blurred vision. This can be surgically corrected and is the most common operation American seniors undergo (Steinberg, Javitt, et al., 1993). Macular degeneration, glaucoma, and corneal disease are other diseases of the visual system that occur mainly in old age (Papalia et al., 2002). Aging also affects the auditory system. About 40% of persons over age 75 report that they have hearing problems of a degree that affects everyday speech communication (U.S. Congress Office ofTechnology Assessment, 1986). The first impairment of hearing is commonly presbycusis, a decrease in the ability to hear high-pitched sounds. For this reason voices of children and women are particularly difficult to hear for older adults (Papalia et al., 2002), particularly for men (Weinstein, 2003). Extremely low frequencies also become a problem in increasing age, and all of these difficulties are exacerbated when set against a noisy background: the elderly have greater difficulty perceiving what is said in a conversation when there is background noise (Slawinski, Hartel, & Kline, 1993; Tun & Wingfield, 1999; Weinstein, 2003; Working Group on Speech Understanding and Aging, 1988; Yarmey, 2001). Older adults show such impaired speech perception for situations in which there is background noise compared with younger adults, even when the ability to hear pure tones (such as in a hearing test) and words spoken in a quiet situation are equivalent across age groups (Frisina & Frisina, 1997; Gordon-Salant, 1987; Tun, 1998), or when they did not show any clinical hearing impairment for pure tones in the speech range (Tun, 1998). Older adults are also less able to recognize fast speech than younger adults (Tun, 1998), and when fast speech is paired with a noisy background, older adults are particularly challenged (Tun, 1998). A related concern is the decrease in available attentional resources with age (Anderson & Craik, 2000; Craik & Byrd, 1982; Hasher & Zacks, 1979). Episodic memory is regarded as particularly resource-demanding because events are often idiosyncratic and unpredictable, and therefore schemas and routines that usually help with the encoding of new information are of only limited or no use in the encoding or retrieval of the event (Craik, 1999). Because there are fewer available attentional resources, when they are applied to an eyewitness situation, older adults may be less able than young adults to encode the eyewitness situation in general and give overall poorer descriptions of the event. On the other hand, older adults may be more selective in the details of a situation to which they pay attention, and consequently they may encode different or fewer details of a situation than younger adults. One way they could be more selective is to pay attention to details that are self-relevant. List (1986) found in a study on adult age differences in eyewitness memory that older adults paid more attention when a perpetrator

582

MUELLER-JOHNSON AND CECI

in a crime video was older than when the perpetrator was a young person. She suggested that a scene with an older perpetrator was more self-relevant to the older participants, and this self-reference enabled more elaborate encoding of the information than in the young perpetrator condition (List, 1986). In addition to the various perceptual and attentive deficits associated with aging, there is some evidence that the elderly are less likely than younger adults to initiate spontaneously encoding strategies that facilitate remembering, such as restructuring of material into semantic clusters (Hultsch, 1969; Sanders, Murphy, Schmitt, & Walsh, 1980; Witte, Freund, & Brown-Whistler, 1993). When presented with verbal stimuli, younger subjects routinely encoded them at more semantic levels than did the elderly (Puglilsi, 1980; Simon, 1979). However, older adults can adopt more elaborate encoding strategies when experimental instructions ask them to (West & Boatwright, 1983) or after undergoing memory training (Verhaeghen, Marcoen, & Goossens, 1992). But given the research showing the less frequent spontaneous use of elaborate encoding strategies, it is thus possible that in a typical eyewitness situation older eyewitnesses do not encode as well as younger eyewitnesses.

The Storage Stage There is ample evidence that certain aspects of memory performance decline with age, although not all parts of memory are equally affected (e.g., Balota, Dolan, & Duchek, 2000). There are relatively small age differences for tasks that place few demands on attention, such as sensory memory tasks, implicit processing, or semantic priming tasks (Balota et al., 2000). However, one of the largest deficits occurs in the area oflong-term episodic memory (Balota et al., 2000), which is heavily relied upon in eyewitness statements. Much of the age difference in the ability to remember may depend on the circumstances under which the memories are encoded and retrieved rather than on difficulties during the retention stage. Research on the effect of the retention interval on age differences in memory performance relies on being able to equate encoding for young and old adults and has met mixed results, mainly due to differences in experimental design, which makes comparison of results across studies difficult (Giambra & Arenberg, 1993). From the evidence it seems that age difference in forgetting rates across retention intervals diminish when initial encoding is controlled (Giambra & Arenberg, 1993; Park, Royal, Dudley, & Morrell, 1988; Rybarczyk, Hart, & Harkins, 1987). Some age differences do nevertheless persist, especially-(for word, sentence, or picture recall) between 12-second and 3-minute intervals (Poon & Fozard, 1980). After about 3-5 minutes up to about 7 days, the two age groups seem to have comparable rates of forgetting (Giambra & Arenberg, 1993; Hulicka, 1965; Hulicka & Weiss, 1965; Mcintyre & Craik, 1987; Rybarczyk et al., 1987), and after 7 days they may have slightly higher forgetting rates (Park, Puglisi, & Smith, 1986; Park et al., 1988). Investigating age differences and retention interval for memory for performed activities, Earles and Coon (1994) contacted participants from an earlier experiment between 2 and 182 days after the experiment and asked them to describe the activities that they had performed as part of the study. The researchers predicted that recall would interact with age and retention inter-

20.

THE ELDERLY EYEWITNESS

583

val. Although recall was related to age and to the length of the retention interval, there was no interaction effect for age by retention interval, suggesting that older adults and younger adults were similarly affected by the length of the retention interval (Earles & Coon, 1994).

The Retrieval Stage Deficits in episodic memory decrease when supportive contextual information (such as cues and reinstatement of the original context) is given at encoding, retrieval, or preferably both (Craik, 2000). Recent research has built upon this finding in order to enhance older witnesses' memory by using the so-called Cognitive Interview, one component of which is context reinstatement. This research is reported in a later section of this chapter. With age, the ability to identify the source of a memory, that is, to engage in source monitoring, decreases. In general, it becomes more difficult to remember contexts, and this can lead to source misattribution (i.e., misremembering where familiar information comes from, such as direct perception, suggestion, dreams, etc.). Compared with young adults, seniors have been shown to have impaired memory when they have to decide if a certain fact was presented either visually or aurally (e.g., Mcintyre & Craik, 1987), if they have seen something on video or in a still photograph (Schacter, Koutstaal, Gross, Johnson, & Angell, 1997), or heard a statement from one person as opposed to another. This is especially evident if the sources are similar, such as when both of them are women (Ferguson, Hashtroudi, &Johnson, 1992). They also have more difficulty distinguishing if something was on a word list or if it was internally generated (i.e., imagined; Hashtroudi, Johnson, & Chrosniak, 1989; Rabinowitz, 1984; Hashtroudi, 1989, p. 48) or if they carried out an action or merely imagined doing so (Brustrom & Ober, 1996). Source misattribution may be related to a decrease in the ability to form associative connections among items or events and to use these associations later to retrieve the memory. This ability to integrate mental events has been called binding and has been shown to decrease with age (Chalfonte &Johnson, 1996). In the legal context, the source of a memory can be of importance in several ways. There are cases in which it is crucial to know whether one person made a statement or another. However, in many cases source memory will be more important in a more indirect way: source misattribution is one way in which misinformation can taint the testimony of a witness. Witnesses to a crime or an accident may learn additional information about the observed incident by reading a related article in the local newspaper or by watching coverage of it in the news. Witnesses may be questioned by interviewers who mistakenly introduce false information in some of their questions. If source monitoring is impaired, then witnesses may come to believe that the additional information stems from the event itself, as opposed to coming from the subsequent informational sources. Source misattributions are important for the explanation of false memories and susceptibility to suggestions (Belli, Lindsay, Gales, & McCarthy, 1994; Johnson, Hashtroudi, & Lindsay, 1993; Multhaup, De Leonardis, &Johnson, 1999)-topics we will return to in the section on seniors' testimony.

584

MUELLER-JOHNSON AND CECI

An effect that may be related to an impairment in source-monitoring ability is the so-called truth-effect. This term describes the effect that mere repetition of an argument tends to make it be perceived as more valid and believable than when it is presented only once. Seniors show a higher susceptibility to the truth effect than young adults (Law, Hawkins, & Craik, 1998). This finding is capitalized on in marketing and advertising but can become important in the legal setting in the context of fraud, such as telemarketing scams. As mentioned above, the elderly are especially prone to be targeted by such schemes (Moore & Lee, 2000). Aging also affects the ability to accurately recall the temporal order of events and actions. Seniors make more mistakes in the temporal ordering of actions than young adults (Krausler & Phillips, 1988; Krausler & Wiley, 1990), a deficit that may be relevant in the chronicling of a series of actions, such as whether the perpetrator first demanded money and then brandished a weapon or vice versa (they lead to substantively different criminal charges). Similarly, spatial memory declines with age. Uttl and Graf (1993) showed that seniors were less able than young adults to describe the spatial layout of museum displays recently visited. Thus, in situations where it can be crucial to describe the layout of the scene of a crime or the temporal order of events, seniors may perform less accurately than young adults.

EYEWITNESS IDENTIFICATION

Visual Lineups Person identification is one of the most studied fields of elderly witness research. Performance of seniors in both simultaneous and sequential photographic lineups has been studied quite extensively. In simultaneous photo lineups the picture of the target and that of all of the foils are presented at the same time. In sequential lineups one photo is shown at a time. In contrast to visual identification, we are not aware of any research with this age group on earwitness identification. The closest information available to date on older earwitnesses is that participants over the age of 40 performed significantly poorer on voice lineups than participants aged 20 to 40 years (Bull & Clifford, 1984). Identifying a perpetrator in a mugshot book, photo spread, video, or live lineup is, of course, a type of recognition task. We mentioned earlier that age differences are smaller for recognition tasks than for recall tasks, such as giving crime or perpetrator descriptions. Accordingly, rates of correctly identifying a target in a target-present lineup are not very different for young and older adults (Yarmey & Kent, 1980). The biggest difference in simultaneous lineups is that seniors generally have higher false-alarm rates than do younger adults, meaning that they more frequently choose a foil, both when the target is present in the lineup and when the target is absent (Memon & Bartlett, 2002; Memon & Gabbert, 2003; Searcy, Bartlett, Memon, & Swanson, 2001; Searcy, Bartlett, & Memon, 1999, 2000). This age difference in the false-alarm rate increases with a

20. THE ELDERLY EYEWITNESS

585

longer delay interval between exposure to the target and the lineup (Memon, Bartlett, Rose, & Gray, 2003). This age difference has been explained with the use of a dual-process framework (Searcy et al., 1999), one that assumes that there are two independent retrieval processes for recognition memory: one is a conscious and effortful process, and the other an automatic process, relying on familiarity Oacoby, 1999). As shown above, seniors recall fewer contextual details of an event than young people (Bartlett & Fulton, 1991; Jacoby, 1999). They also make more source-monitoring errors and encode more in a gist-like manner, rather than in a verbatim/distinctive detail way (Koutstaal & Schacter, 1997; Schacter, Norman, & Koutstaal, 1998; Tun, Wingfield, Rosen, & Blanchard, 1998). Thus, seniors are more likely to make their judgments based on familiarity rather than relying on conscious retrieval processes.

Attempts to Decrease Seniors' High False-Alarm Rates Research with young adults has shown that it is possible to decrease false identification rates in simultaneous lineups by increasing correct rejection rates through the use of prelineup questions, such as: How clear a memory do you have for the face of the criminal? How confident are you that you will be able to select the criminal if you see a photograph of him in a lineup? How confident are you that you will realize that the guilty person is not in the lineup if you are shown a lineup with only innocent people in it? (Dysart & Lindsay, 2001). A replication study and extension to a sample of older adults (Memon & Gabbert, 2003) did not show any effect of pre-lineup questions in increasing correct rejections on target-present or target-absent choices for younger or older participants. Thus it appears that this technique does not promise to offset deficits in correct rejections by older adults. Another technique shown to reduce false-alarm rates in young adults is the sequential lineup (e.g., Cutler & Penrod, 1988; Lindsay et al., 1991). One possible underlying reason for this could be that sequential lineups lead to the use of an absolute judgment strategy, whereas simultaneous lineups provide conditions in which participants use a comparative choosing strategy (Dysart & Lindsay, 2001; Kneller, Memon, & Stevenage, 2001). Another explanation is that a sequential lineup setting makes participants set their decision criterion higher for when they consider a match to have occurred between the perpetrator they observed and the faces in the lineup, than they do in a simultaneous lineup (Ebbesen & Flowe, as cited in Memon & Gabbert, 2003). Giving seniors instructions that are designed to make them adopt a strict decision criterion works to lower false alarm rates to a certain extent (Memon, Hope, Bartlett, & Bull, 2002; Multhaup et al., 1999), but it does not eliminate age differences completely. Nevertheless, for seniors, sequential testing may not be a good solution. As seniors have greater difficulty inhibiting activated but irrelevant information (e.g., Hamm & Hasher, 1992; Hartman & Hasher, 1991), they may also find it harder to ignore the information in the faces of preceding foils in a sequential lineup and thus actually increase

586

MUELLER-JOHNSON AND CECI

false choices. Memon and Gabbert (2003) showed that sequential lineups led to higher rates of correct rejections in the target-absent condition for both young and old participants. However, whereas young adults showed the expected decrease in false alarm rates in the target-present condition, seniors did not show such lower false alarm rates, but, on the contrary, actually evidenced a tendency to increase false alarms for sequential lineups, compared with simultaneous lineups. In addition, both age groups showed a decrease in the hit rates when compared with simultaneous lineups (which is a common finding; see Memon & Gabbert, 2003, table 2). Thus, although sequential lineups may be the best available lineup strategy for younger adults, for older adults sequential lineups may actually lead to worse performance than simultaneous lineups.

Biases in Person Identification One of the classic findings in eyewitness identification research is the cross-race effect: witnesses have higher accuracy rates when identifying someone from their own race than when they identify someone from a different race. Caucasians have been shown to do this at a higher rate than African Americans (see Meissner & Brigham, 2001, for a meta-analytic review). To our knowledge there has been only one study looking at the cross-race effect for the elderly (Brigham & Williamson, 1979). This study showed that African-American seniors were significantly better at identifying African-American target pictures (one male, one female) than in identifying Caucasian target pictures (one male, one female). However, Caucasian seniors did not differ significantly in their identification rates between photos of Caucasian and African-American targets. Thus, in this study a cross-race effect was shown only for the African-American seniors, not for the Caucasian seniors. In evaluating Brigham and Williamson's (1979) result, one should note that a cross-race effect for the Caucasian subsample may have been masked by very low hit rates for the white female face in the study (for both groups of participants). Looking only at the male faces, it appears that a cross-race effect may have been found also for Caucasians, but the hit rates for male and female Caucasian faces were not tested separately in the study, so this interpretation of the data remains speculative. Furthermore, the sample size for this study was extremely small, containing only 14 Caucasian seniors and 27 African-American seniors. Additional work on the cross-race effect in older adults is needed. It has been suggested that there may also be an "own-age bias," analogous to the own-race bias. Wright and Stroud (2002) showed that for target-present lineups young adults' (18-25, 18-33) and middle-aged adults' (35-55, 40-55 respectively) identification hit rates were higher for targets of their own age group. However, this was only the case for target-present lineups, not for target-absent lineups, where no own-age bias was found (Wright & Stroud, 2002). In another study comparing seniors (60-82) and students (16-33) on lineups with young and old targets, such superior performance for pictures of participants' own age group was not found (Memon & Gabbert, 2003). Hit rates for students did not differ across the two target-present lineups, and seniors were actually performing worse for the old faces lineup than for the young faces lineup. Young students showed lower false-

20. THE ELDERLY EYEWITNESS

587

alarm rates in the young face target,present lineup than in the old face target,present lineup, and seniors showed similar rates of false alarms in the two lineups. Only in the target,absent condition was an age bias found-young adults showed higher correct re, jection rates for the young face lineup than for the old face lineup, and seniors had higher correct rejection rates for the old face lineup than for the young face lineup. Even for the old face lineup, where seniors were at their best and students at their worst, how, ever, students still outperformed seniors (Memon et al., 2003)

Confidence for Identification Choices Some studies find that seniors are less confident in their identifications than younger adults (Neuschatz et al., 2005; Yarmey, 1984), whereas others do not find any age differ, ences (Adams Price, 1992; O'Rourke, Penrod, Cutler, & Stuve, 1989). Similar to well, known findings with young adults, young,old adults and old,old adults do not evidence a correlation between confidence and accuracy (Scogin, Calhoon, & D'Errico, 1994). That is, being very confident is not indicative of a high degree of accuracy, such that at any age a witness's level of sureness should not be viewed as diagnostic of identification accuracy. Furthermore, confidence can easily be influenced, for instance by confirmatory post,identification feedback. In a series of clever studies by Wells and his colleagues, par, ticipants viewed a videotaped event and then attempted to identify the suspect from a suspect,absent lineup. Instructions prior to the showing of the photospread tilted the odds in favor of making a selection, and indeed all participants made an incorrect iden, tification. Interestingly, when given positive feedback after an identification was made (i.e., the participants were told that they had chosen the correct person), compared with no feedback, people said that they were more confident in their choices and that they would be more willing to testify in court. This inflation in confidence has been shown for younger adults (e.g., Bradfield, Wells, & Olson, 2002) and was also recently extended to older adults (Neuschatz et al., 2005).

EYEWITNESS TESTIMONY

Completeness and Accuracy Older adults tend to give fewer details than young adults when describing an event they have experienced (Brimacombe, Quinton, Nance, & Garrioch, 1997), both in di, rect examination and in simulated cross,examinations (Brimacombe, Jung, Garrioch, & Allison, 2003). As a group seniors are also likely to be less accurate than young adults (Brimacombe et al., 1997; Yarmey & Kent, 1980). In a recent review of the eyewitness literature, Yarmey (2001) reported results on older witness accuracy average across three studies (Yarmey, 1982; Yarmey, Jones, & Rashid, 1984; Yarmey & Kent, 1980). Here older adults were 20% less accurate than younger adults in free recall and 13% less accurate in cued recall. Overall they were also 15% less complete when describing the perpetrator.

588

MUELLER-JOHNSON AND CECI

There are qualifications to this age difference. It seems that under certain conditions the amount of detail and accuracy of details can reach the level of young adults. List (1986) showed participants several shoplifting scenes in which the shoplifter was either a college-age female or a middle-aged female. Although younger adults gave more complete and more accurate description for the younger thief than children and seniors, for the older thief seniors' memory was as complete and accurate as that of younger adults. List suggests that to the older adults the older thief scene was particularly memorable because it contained more self-relevant information. This self-reference then permitted more elaborate encoding of the event. Brimacombe and colleagues (Brimacombe et al., 2003) investigated the possibility that the familiarity of the crime setting may also lead to an encoding and retrieval advantage for older adults. Familiar situations trigger the activation of schemas, internal packets of scripted knowledge, that, in principle, should facilitate both encoding of the events and their recollection. In their study young and old adults watched a simulated theft in either a familiar or an unfamiliar setting. Familiarity of context had no effect on the accuracy of the statements, but it had an effect on the length of the descriptions delivered for the old-old participants (mean age 78.5). Older seniors who witnessed the theft in the familiar setting (i.e., in their senior center) provided significantly longer descriptions than in the unfamiliar setting (i.e., in the university community center). For young adults and young-old seniors (mean age 67.7 years) the familiarity of the setting did not have an effect on the length of the statement. Young seniors, on average, gave longer descriptions than young adults. In general, the increase in words for seniors was not related to more objective details, but to a higher number of subjective impressions mentioned in the statement, such as "I thought it would be normal for someone to look up if a person nearby was speaking loudly, but that woman did not look up." The familiar setting condition prompted older seniors also to make more off-topic remarks than in the unfamiliar settings, which was mainly related to the fact that they were accustomed to the setting.

Suggestibility In addition to lower rates of accuracy in old age, suggestibility has been discussed as a concern facing elderly witnesses. It has been reported that in at least some situations the elderly are more vulnerable to suggestions than younger adults (Cohen & Faulkner, 1989; Karpel, Hoyer, & Toglia, 2001; Loftus, Levidow, & Duensing, 1992). Most research to date on suggestibility and the elderly has employed the standard misinformation paradigm, in which participants watch a film and receive misinformation about some details of the film a short time later (typically after 10-30-minute intervals), and then after another brief delay they are tested for acceptance of the misinformation in a questioning phase. Results of these studies on seniors have been mixed. Some researchers have reported higher rates of susceptibility to misinformation in older adults than in young adults (Cohen & Faulkner, 1989; Karpel et al., 2001; Loftus et al., 1992; Mitchell, Johnson, & Mather, 2003). However, other studies have not reported reliable age differences (Bornstein, Witt, Cherry, &Greene, 2000; Coxon & Valentine, 1997; Gabbert, Memon,

20. THE ELDERLY EYEWITNESS

589

& Allan, 2003; Searcy et al., 2000). One study has even found that young adults were more suggestible than older adults (Marche, Jordan, & Owre, 2002). Our own work (Mueller-Johnson & Ceci, 2004), using a live event involving bodily touching, showed that seniors were more susceptible than young adults for certain types of misinformation but not for other types. In this study older and younger adults received a massage and participated in a series of actions. On two separate occasions several weeks after this event, half of each age group was given misleading information about certain aspects of the event. Several types of suggestions were given: suggestions about touching (e.g., that they had been massaged on parts of their body that had not actually been touched), suggestions about the clothing of the massage therapist, and suggestions that subjects had actually performed certain actions instead of having merely imagined them. Older adults were more suggestible overall than younger adults. They were particularly influenced by the suggestions involving touching. However, for the suggestion concerning the massage therapist's clothes, only younger adults showed a significantly greater suggestibility effect than older adults. For suggestions inviting source misattributions, such as that they had performed an action when in fact they had only imagined it, older adults showed elevated rates of inaccurate answers; however, because of the elevated error rates in the control group, no suggestibility effect was detected (i.e., even in the absence of misinformation, almost half of the seniors nevertheless claimed they performed activities that they had merely imagined). These findings showed that suggestibility is affected by the type of suggestions made. For some suggestions seniors do not show higher rates of susceptibility to misinformation, and for others, including suggestions about having been touched, they do. Suggestions also brought about age differences in inaccurate answers (i.e., in answers that incorporated the suggested details). Without suggestions, and when participants successfully withstood suggestions, old and young adults showed the same confidence rate. If they incorporated suggested details in their answers, young adults' confidence decreased compared with peers who withstood the suggestions, whereas seniors' confidence remained the same, even when they incorporated erroneous suggestions (Mueller-Johnson & Ceci, 2004). Another path to the investigation of interrogative suggestibility employs the Gudjonsson Suggestibility Scale. In this paradigm participants hear a story and give a free account of what they heard, and after an interval of about 30 minutes and another free recall account, they are asked a series of 20 closed-format questions about the story. Fifteen of these questions are suggestive. After the participants have answered all questions, they are informed that they made mistakes and that it is therefore necessary to go through the questions again and a second round of the same 20 questions follows. In coding the answers, one score (Yield 1) is given for the number of times the participant acquiesced to a suggestion in round 1, and a second score (Yield 2) is given for the number of acquiescent answers in round 2. A difference score is given for changes in answers between round 1 and round 2 (Shift score), and an overall suggestibility score is derived from summing up the Yield 1 score and the Shift score. Most of the focus in research using this scale has been on young adults, and only very recently has the scale been applied to older adults. Polczyk and colleagues (Polczyk et al., 2004) found that for a Polish

590

MUELLER-JOHNSON AND CECI

sample, older adults were more suggestible on the yield scores of the scale than young adults, and had comparable shift scores. This means that older adults were more influenced by the suggestive wording of the questions than young adults but were not more influenced by the coercive social situation of being asked to answer questions again that they had seemingly answered wrongly before (Polczyk et al., 2004). These results are in contrast to our own study (Mueller-Johnson & Ceci, manuscript in preparation), in which we administered the Gudjonsson Suggestibility Scale to the same American sample of young and old adults as in our study described above and did not find any elevated rates for suggestibility in the older sample on any measures of the Gudjonsson Suggestibility Scale. More research is needed on the circumstances in which older adults may show higher suggestibility scores than young adults.

INTERVIEWING THE ELDERLY The Mindset: The Effect of Aging Stereotypes

It has been noted that elders are sometimes spoken to in what is called "patronizing speech" or "elder talk." This has been observed in the context of health care settings (Coupland, Coupland, Giles, & Henwood, 1988; Lanceley, 1985) and in community settings (Coupland et al., 1988). Although there are no studies on the incidence of patronizing speech in the context of investigative interviews, given the prevalence of patronizing speech in other situations, it is not unlikely that patronizing speech could occur in law enforcement settings. One conceptualization of elder talk is that the sender in a communication context automatically tries to adapt the message to the recipient's perceived level of understanding. Age stereotypes can trigger this adaptation of speech even when the elder does not show any problems in understanding and communicating with the sender. It has been shown that age stereotypes can be triggered by age clues in the facial features of the aged person (Hummert, 1994, 1999), such as degree of wrinkling, skin tone, hair greying, and eye droop. Hummert (1994) presented to observers photographs of older adults with different levels of facial aging cues and asked them to sort them into groups according to certain aging stereotypes identified in previous research. Faces that involved fewer aging cues were more frequently assigned to positive stereotypes, such as "perfect grandparent," "liberal matriarch/patriarch," or "John Wayne conservative," rather than negative stereotypes, such as "despondent," "shrew/curmudgeon," "severely impaired." The more facial aging cues were present, indicating higher age, the more frequently negative stereotypes were assigned. Noticing facial aging cues in a witness could therefore activate aging stereotypes in an interviewer, who may then react by resorting to patronizing speech. If interviewers engage in patronizing speech, this, in turn, can lead to the activation of aging stereotypes in the elderly person. In the mildest case, patronizing speech has also been shown to be perceived as negative and impolite and can decrease cooperation. It has also been reported, however, to decrease the aged person's self-esteem and confidence. A phenomenon called "instant aging" has been observed, in which elders who

20. THE ELDERLY EYEWITNESS

591

are talked to in a patronizing way behave, look, think, and talk "older" than control subjects (Giles, 1994). Activation of aging self-stereotypes has been shown to decrease memory performance (Hess, Auman, Colcombe, & Rahhal, 2003; Levy, 1996). In summary, patronizing speech from interviewers can decrease the quality of the witness's memory report and potentially increase his or her susceptibility for misleading information. When interviewing elders as witnesses, it is therefore important to be aware of and avoid the deleterious effects of patronizing speech.

When to Interview: Circadian Rhythms and Memory People have different times of day when they are cognitively at their best-hence the popularity of such expressions as "I am not a morning person." With advancing age, the time of day of optimal functioning shifts closer to the morning. More seniors than younger adults are "morning people" (May, Hasher, & Stoltzfus, 1993; Mecacci, Zani, Rocchetti, & Lucioli, 1986). Data reported from about 1,600 college students and 600 seniors showed that roughly 40% of the students were evening types, less than 10% were morning types, and a large group were "neither type." In contrast, about 75% of seniors were morning types and less than 3% were evening types (Yoon, May, & Hasher, 2000). Older adults are more affected in their performance by their circadian rhythm than are younger adults. They have been shown to make more mistakes on working memory tasks when tested in the late afternoon (5 P.M.) than when they are tested in the morning (9 A.M.) (West, Murphy, Armilio, Craik, & Stuss, 2002). People tested during off-peak times of day are more prone to use stereotypes when making social judgments than are people who are tested during their peak time (Bodenhausen, 1990). The effect of circadian rhythms has also been studied in the context of persuasion. The persuasion literature customarily differentiates between strong and weak arguments and shows that under optimal conditions, such as high motivation and no distraction, strong arguments are persuasive. Under low motivational conditions or under conditions of distraction, weak arguments, such as the attractiveness of the sender, are used when judgments are made about the persuasiveness of an argument. When this research paradigm was used with young and old seniors at different times of day, it was shown that distraction had no effect on seniors in the morning hours, when they were at their peak time, but did have a large impact in the afternoon, when they were off their peak. In the afternoons, weak arguments under distractions were considered to be more persuasive than in the mornings (Rahhal, as cited in Yoon et al., 2000). It has been shown that where there is distraction, high motivation can usually assist in processing arguments in a central rather than a peripheral way, thus leading to use of strong arguments rather than weak arguments. However, in off-peak conditions, it has been shown that seniors are persuaded by weak arguments, under both low and high motivation. In contrast, for students motivation still made a difference despite the fact that they were tested off-peak. If their motivation was high, students continued to be persuaded by strong rather than weak arguments, even in off-peak times (Yoon & Lee, as cited in Yoon et al., 2000). Thus, seniors are more affected than young adults by circadian rhythms in their social cognition.

592

MUELLER-JOHNSON AND CECI

An application of preferred and nonpreferred testing time to false memories using the Deese/Roediger-McDermott paradigm (Roediger & McDermott, 1995) showed that off-peak-time testing resulted in the creation of false memories (Intons-Peterson, Rocci, West, McLellan, & Hackney, 1999). In this paradigm participants are shown word lists in which all presented words on a list are associatively related to a critical word that was not presented. The critical word, for instance, may be sweet, and the words in the list could be sugar, honey, tooth, sour, candy, etc. After being shown each list, participants recalled the words that were on the list. Later they were administered a recognition test in which they decided if a word had been presented on the list or was new. Typically participants in this paradigm recall and recognize the critical nonpresented word as having been on the list, although it had not been presented. Intons-Peterson and colleagues tested young and old adults with this technique at either their preferred testing time (for seniors mainly in the mornings) or their nonpreferred testing time (for seniors mainly in the afternoon). Seniors who were tested at their nonpreferred testing time recalled and recognized more of the critical (i.e., nonpresented) words. Thus they had more false memories than participants tested at a preferred time of day. For the young adults testing time did not make a difference. We are not aware of any research that investigated older adults' circadian variations in the context of eyewitness situations. However, investigations of preferred testing time on older adults' performance in interview situations may be a promising direction for future eyewitness research on the elderly. It is possible that seniors may remember better and rely less on scripts (schemas) and stereotypes when interviewed at their preferred time. Drawing on the findings oflntons-Peterson et al. (1999) findings, researchers may find that seniors also show lower rates of false alarms in lineups and are even less suggestible when tested at their preferred time of day.

Improving Seniors' Witness Accounts: The Cognitive Interview The Cognitive Interview is an approach to enhancing witness accounts by facilitating retrieval of memories (Fisher & Geiselman, 1992). It relies mainly on two very important principles from cognitive science research (Fisher, Brennan, & McCauley, 2002): that contextual reinstatement improves memory performance by increasing the accessibility of encoded information (encoding specificity hypothesis by Tulving and Thompson, 1973) and that memories are encoded in a network of associations (multiple trace theory; see Bower, 1967), which implies that they can be triggered by multiple cues. The components of the original Cognitive Interview include mental reinstatement of the context; complete reporting, including details believed to be trivial; varied retrieval routes; and "witness-compatible questioning." Varied retrieval includes changes in retrieval order and perspective, such as starting at the beginning, then starting at the end, then reporting from the perspective of a bystander. Witness-compatible questioning includes tailoring the interview questions to the mental representation of the witness instead of using the same standardized checklist for all witnesses. The Cognitive Interview

20. THE ELDERLY EYEWITNESS

593

is also sensitive to the fact that people's resources are limited, and therefore questions are not asked while the witness is still thinking about the last question. Furthermore, when participants are asked to reinstate the context, they are told to close their eyes and concentrate on visualizing the event, to obviate competing demands on attentional resources (Fisher et al., 2002). The Cognitive Interview has been shown to produce an increase in the number of accurate details reported by witnesses, without increasing the number of inaccurate details, for both young adults and children (Koehnken, Milne, Memon, & Bull, 1999). Over the years different versions of the Cognitive Interview have been developed, such as the Enhanced Cognitive Interview (e.g., Fisher, McCauley, & Geiselman, 1994) and the Modified Cognitive Interview, which was adapted for use with older adults by Mello and Fisher (1996). In this modified version for older adults, the part in which the witness is requested to describe the events from the perspective of another person is omitted, and the opening free recall part is minimized because seniors are assumed to have difficulty with both parts. Further ways in which the Modified Cognitive Interview differs are an even slower pace of the interview and expressly telling the participants that the interview will proceed slowly, simpler wording of directions and questions, and avoiding interruption of the participant. Because the authors assumed that seniors can behave in an overly cautious way and therefore may withhold important information, a further modification is to stress that it is important to describe all memories, even if they are considered trivial. At the same time seniors are reminded not to guess. Some of these components are also part of the original Cognitive Interview, but the modifications extend these, such as going even more slowly than the already slow speed of the Cognitive Interview, or stressing even more that participants should not fabricate any information (Mello & Fisher, 1996). The evidence about the effectiveness of the Cognitive Interview in enhancing seniors' performance is mixed. Mello and Fisher (1996) interviewed young and old adults about a crime video, with the use of the standard police interview, the Cognitive Interview, or the Modified Cognitive Interview. They found a general superiority of the Cognitive Interview both for young and especially for old adults. For the older adults, the performance on the Modified Cognitive Interview did not differ from that in the original Cognitive Interview. In the standard interview no age differences were found, and in the Cognitive Interview the older participants actually remembered more details than the young adults. McMahon (2000) also tested younger and older adults for their recall of a crime video. She used either the Structured Interview or the Cognitive Interview. The Structured Interview was developed by Kohnken (1993) in an attempt to provide a control condition for the evaluation of the Cognitive Interview, and it employs steps similar to those of the Cognitive Interview, such as free narrative phase, open-ended questions, etc., but does not contain the cognitive aspects of the Cognitive Interview, such as context reinstatement or the change of order. It is almost identical with the structure outlined in the British Home Office Memorandum of Good Practice (Wright & Holliday, 2003). In McMahon's study the Cognitive Interview did not lead to better performance

594

MUELLER-JOHNSON AND CECI

than the Structured Interview. Young participants gave more correct responses than the older group, but the groups did not differ in the number of incorrect responses or confabulated responses (McMahon, 2000). A recent study compared the effect of the Enhanced Cognitive Interview, the Modified Cognitive Interview, and the Structured Interview on seniors (Wright & Holliday, 2003). Wright and Holliday (2003) showed seniors a brief video about an attempted car theft. Afterward the seniors were questioned, in the format of the Structured Interview, the Enhanced Cognitive Interview, or the Modified Cognitive Interview. Both the Enhanced Cognitive Interview and the Modified Cognitive Interview elicited more details than the Structured Interview. The Enhanced Cognitive Interview yielded 13.2% more accurate details than the Modified Cognitive Interview, which in turn obtained 21.7% more accurate details than the Structured Interview. Although there was this increase in accurate details, no increase in inaccurate details or confabulations was observed. This is especially important because an increase in accurate information would be of little value if it came at the cost of more mistakes. When trying to pin down the locus of the increase in details, Wright and Holliday (2003) showed that the increase occurred in the free recall phase of the interviews rather than in the questioning phase. This is an interesting finding because seniors do not usually do well on free recall tasks. As stated above, one reason for this may be that they are less likely to make spontaneous use of mnemonic strategies (Yarmey, 1996). The authors suggest that the Cognitive Interview and especially the Enhanced Cognitive Interview provide a contextual scaffolding that facilitates free recall (Wright & Holliday, 2003). These findings suggest that the Enhanced Cognitive Interview may particularly warrant further research with elderly participants and provide a possible way to enhance their performance in court. If training in the Cognitive Interview is not possible for frontline staff, certainly at least some of the components of the Cognitive Interview could be relatively easily implemented. Such feasible parts include the slow pace; the explicit instruction not to leave anything out, not even details considered trivial; and reinstating the context of the crime, for example, by instructing the witness to take a moment, close his or her eyes, and think back to the scene of the crime. One of the most important practical issues in everyday interviewing may be to provide sufficient time to allow elderly persons to feel that they are not rushed.

CREDIBILITY OF THE OLDER WITNESS

The effect of old age on the credibility of a witness's report has been investigated in a few studies so far. Ross, Dunning, Toglia, and Ceci (1990) examined the effect of manipulating the age of a witness (8, 21, or 74 years) while maintaining the same testimony. Student participants received information about a trial for possession of cocaine in which the target witness was the main witness for the prosecution. Then they watched the target witness give evidence on video (experiment 1) or read transcripts of their testimony (experiment 2). Participants judged whether they found the defendant guilty and rated the target witness on his accuracy of memory, confidence, forcefulness of response, cred-

20. THE ELDERLY EYEWITNESS

595

ibility, and honesty. Participants rated the elderly witness as accurate in his or her mem~ ory as younger witnesses and children, and, if anything, as slightly more accurate than younger witnesses. Elderly witnesses were also rated as more honest and trustworthy than their 21~year~old counterparts. Another study looking at older adults' credibility found contradictory results (Kwong See, Hoffman, & Wood, 2001). This study used an indirect method to investigate the ef~ feet of age on the credibility of witnesses. These researchers exposed participants to the standard misinformation paradigm, in which they watched a slide show and were pre~ sented with a narrative that contained some neutral and some misleading details, coun~ terbalanced across groups. Participants were told that this description was obtained from a female witness who was either 28 or 82 years old, and they were also shown a picture of this witness. The authors measured the believability of the witness as the size of a mis~ information index that resulted from subtracting for each person the average memory score on items on which he or she had been misled, from the average memory score for neutral items. Scores for characteristics such as the witness's competence and honesty were obtained by participant ratings, and the older witness was rated as less competent than the young, but as more honest. The size of the misinformation effect that was ob~ tained for each witness age group was not reported. Rather, the believability scores, re~ suiting from the individual misinformation indices, were used as criteria for analyses with age of witness, competence, and honesty as predictors. Witness age and honesty did not predict the believability of the witness, but perceived competence and the compe~ tence by age interaction contributed to believability. Whereas for the young witness competence predicted believability, for the old witness this was not the case. The results of this study confirm previous findings, in that the older witness was perceived to be more honest, but they diverge in that the older witness was perceived to be less compe~ tent. The older witness was not necessarily seen as less believable than the young witness (Kwong See et al., 2001). The elderly are a very heterogeneous group with a wide variety of cognitive func~ tioning levels and communicative skills. Similarly, there are a variety of stereotypes about the elderly, ranging from the positive image of the golden ager and perfect grand~ parent to the negative image of the severely impaired nursing home resident (e.g., Hum~ mert, 1999). The results of the previous two studies on elderly witnesses suggest that participants based their ratings and judgments on elderly witness capacity and stereo~ types. A negative perception of an elderly witness would be expected if this person trig~ gered stereotypes of being impaired and suffering from memory loss, but not if the wit~ ness activated positive stereotypes, such as being experienced and wise. Further research is needed to create a more differentiated picture of the credibility of "the" elderly wit~ ness, taking into account the large heterogeneity of this age group. Nunez and her col~ leagues (Nunez, McCoy, Clark, & Shaw, 1999) made a first attempt at a more dif eren~ tiated concept of elderly individuals in research on credibility and aging. They gave jurors a description of an aggravated assault. The elderly victim testified at trial and identified the accused as the perpetrator. The written description of this witness was var~ ied to elicit a senior citizen with health problems, a pleasant grandfather, or an elder statesman. According to expectations, jurors were more likely to convict the accused in

596

MUELLER-JOHNSON AND CECI

the condition of the elder statesman than in the condition of the grandfather or the health-impaired senior citizen. Interestingly, however, mock jurors did not assign different credibility ratings to these three different elderly witnesses (Nunez et al., 1999). The studies reviewed above tried to investigate older adults' credibility by keeping the witness statement itself constant and varying only the witness age. As we have shown in previous sections of the chapter, older adults' witness statements may likely differ from those of younger adults (e.g., be less complete in the amount of details and possibly less accurate). These characteristics of witness accounts could also affect perceived credibility, especially when paired with old age. Brimacombe et al. (1997) used a different method to study older adults' perceived credibility that took these potential differences into account. They first had older and younger adults give statements about a video they had watched. Then they used these videos as stimulus material and had mock jurors evaluate these statements for credibility. Seniors' statements were perceived to be less accurate and less credible than those of young adults. However, credibility was not affected by the age of the witness, but by the quality and characteristics of the statements. It has been shown for younger witnesses that they are perceived as more credible if their accounts include a high amount of peripheral details in their descriptions of the observed crime (Wells & Leippe, 1981). As seniors typically provide fewer peripheral details in their accounts, it is plausible that this is a signal to mock jurors that they are not as accurate witnesses. Another explanation for the lower perceived credibility of the seniors in this study could be specifics of seniors' use of language, rather than the amount of details. In arecent study Brimacombe et al. (Brimacombe et al., 2003) showed interviews with young adults (18-30 years), young seniors (59-74 years), and old seniors (75-88 years) to mock jurors. Prior to this they had established that the interviews with older seniors were less accurate than those of young adults and young seniors. However, despite this lower level of accuracy, older seniors were perceived to be as accurate as the younger groups. The authors explained this seemingly contradictory finding in terms of differences in the delivery of the statements. In the study by Brimacombe et al. (1997), where seniors were perceived to be less credible, seniors uttered more negative qualifiers (such as "I am not sure"), whereas in the other study by Brimacombe et al. (2003), where seniors were perceived as more credible, older seniors employed very few cognitive hedges. Such negative qualifiers have been shown to detract from the credibility or persuasive power of a statement (Erickson, Lind, Johnson, & O'Barr, 1978). Brimacombe et al. (2003) suggest that the difference in using these qualifiers may be related to the sample. Brimacombe et al. (1997) used participants who all lived in a senior residence that provided the services of an assisted living facility, whereas in Brimacombe et al. (2003) seniors were living independently in the community. These researchers argue that assisted living/institutionalized living has negative effects on self-esteem and that this lower selfesteem may have prompted these seniors to use more negative qualifiers, regardless of their confidence that this particular statement was right or wrong (Brimacombe et al., 2003). If further research supports this argument, then the living circumstances of an older eyewitness may, albeit indirectly, influence his or her perceived credibility.

20. THE ELDERLY EYEWITNESS

597

To sum up, several attempts have been made to assess older witnesses' perceived credibility. These attempts have looked at the effect of stating the witness's old age independent of the quality of the statement and at characteristics of older witnesses' statements, such as amount of detail and use of language. CONCLUSIONS

Research on elderly witnesses is a rapidly growing area. The projected shifts in demographics will make the elderly even more important as witnesses. Aging brings about some important changes in sensory perception and memory that are related to less keen perception and less detailed and accurate accounts of perceived events, at least in comparison with young adults. There have been some promising attempts to enhance seniors' performance as eyewitnesses, such as the Cognitive Interview for forensic interviews. Similarly, in the area of person identification, researchers have attempted, albeit with only limited success, to improve older adults' performance, for example, by reducing false alarm rates. It is important to remember that seniors vary greatly in how much they are affected by aging; some elders can be highly accurate until they reach an advanced age (old-old age), whereas others show sizable deficits relatively early (young-old age). More research is needed to address the heterogeneity of this group and identify individual differences related to eyewitness performance. Some research has also shown that some age differences between old adults and young adults do not occur until old-old age. Until more is learned about the role of individual differences, future research should at the very least differentiate age ranges within old age, rather than simply classifying all individuals over 65 as "elderly." Although there are already a few promising approaches to enhancing elderly witnesses' performance, such as applications of the Cognitive Interview, more research is needed to determine what other aspects of the interviewing situation can be adapted to optimize their performance. Addressing interviewer-related components, such as interviewer behavior toward seniors and identifying nonhelpful attitudes such as ageist stereotypes, is a promising possibility for future research.

REFERENCES Adams Price, C. (1992). Eyewitness memory and aging: Predictors of accuracy in recall and person recognition. Psychology and aging, 7, 602-608. American Association of Retired Persons. (1998). A profile of older Americans: 1998. Washington, DC: Author. Anderson, B., & Palmore, E. (1974). Longitudinal evaluation of ocular function. In E. Palmore (Ed.), Normal aging II, Reports from the Duke longitudinal studies, 1970-1973 (pp. 24-32). Durham, NC: Duke University Press. Anderson, N.D., & Craik, F. I. M. (2000). Memory in the aging brain. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 411-4 25). Oxford: Oxford University Press.

598

MUELLER-JOHNSON AND CECI

Atchley, R. C. (2000). Social forces and aging. An introduction to social gerontology (9th ed.). Belmont, CA: Wadsworth Thomson Learning. Balota, D. A., Dolan, P. 0., & Duchek, J. M. (2000). Memory changes in healthy older adults. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 395-409). Oxford: Oxford University Press. Bartlett, J. C., & Fulton, A. (1991). Familiarity and recognition of faces in old age. Memory and Cognition, 19(3), 229-238. Belli, R. F., Lindsay, D. S., Gales, M.S., & McCarthy, T. T. (1994). Memory impairment and source misattribution in posteven misinformation experiments with short retention intervals. Memory and Cognition, 22,40-54. Bodenhausen, G. V. (1990). Stereotypes and judgmental heuristics: Evidence of circadian variations in discrimination. Psychological Science, 1, 319-322. Bornstein, B. H., Witt, C.]., Cherry, K. E., & Greene, E. (2000). The suggestibility of older witnesses. In M. B. Rothman, B. D. Dunlop, & P. Entzel (Eds.), Elders, crime and the criminal justice system: Myth, perceptions, and reality in the 21st century (pp. 149-161). New York: Springer. Bower, G. H. (1967). A multi component theory of the memory trace. InK. W. Spence & ]. T. Spence (Eds.), Psychology ofleaming and motivation (Vol. 1, pp. 229-325). New York: Academic Press. Bradfield, A. L., Wells, G. L., & Olson, E. A. (2002). The damaging effect of confirming feedback on the relation between eyewitness certainty and identification accuracy. Journal of Applied Psychology, 87, 112-120. Brigham, J. C., & Williamson, N. L. (1979). Cross-racial recognition and age: When you're over 60, do they still "all look alike"? Personality and Social Psychology Bulletin, 5(2), 218-222. Brimacombe, C. A. E., Jung, S., Garrioch, L., & Allison, M. (2003). Perceptions of older adult eyewitnesses: Will you believe me when I'm 64? Law and Human Behavior, 27(5), 507-522. Brimacombe, C. A. E., Quinton, N., Nance, N., & Garrioch, L. (1997). Is age irrelevant? Perceptions of young and old adult eyewitnesses. Law and Human Behavior, 21, 619-634. Brustrom, ]. E., & Ober, B. A. (1996). Source memory for actions in Alzheimer's disease. Aging, Neuropsychology and Cognition, 3(1), 56-66. Bull, R. H. C., & Clifford, B. R. (1984). Earwitness voice recognition accuracy. In G. L. Wells & E. F. Loftus (Eds.), Eyewitness testimony: Psychological perspectives (pp. 92-123). Cambridge and New York: Cambridge University Press. Chalfonte, B. L., & Johnson, M. K. (1996). Feature memory and binding in young and older adults. Memory and Cognition, 24, 403-416. Cohen, G., & Faulkner, D. (1989). Age differences in source forgetting: Effects on reality monitoring and on eyewitness testimony. Psychology and Aging, 4, 10-17. Coupland, N., Coupland, J., Giles, H., & Henwood, K. (1988). Accommodating the elderly: Invoking and extending a theory. Language in Society, 17 (1) , 1-41. Coxon, P., & Valentine, T. (1997). The effect of the age of eyewitnesses on the accuracy and suggestibility of their testimony. Applied Cognitive Psychology, 11, 415-430. Craik, F. I. M. (1999). Memory, aging and survey measurement. InN. Schwarz, D. C. Park, B. Knauper, & S. Sudman (Eds.), Cognition, aging, and self-reports. Philadelphia: Taylor & Francis. Craik, F. I. M. (2000). Age-related changes in human memory. In D. C. Park & N. Schwarz (Eds.), Cognitive aging: A primer (pp. 75-92). Philadelphia: Taylor & Francis. Craik, F. I. M., & Byrd, M. (1982). Aging and cognitive deficits: The role of attentional resources. In F. I. M. Craik & S. Taub (Eds.), Aging and cognitive processes (pp. 191-221). New York: Plenum. Cutler, B. L., & Penrod, S. D. (1988). Improving the reliability of eyewitness identification: Lineup construction and presentation. Journal of Applied Psychology, 73, 281-290.

20.

THE ELDERLY EYEWITNESS

599

Dysart, J. E., & Lindsay, R. C. L. (2001). A preidentification questioning effect: Serendipitously increasing correct rejections. Law and Human Behavior, 25, 155-166. Earles, J. L., & Coon, V. E. (1994). Adult age-differences in long-term memory for performed activities. Journals of Gerontology, 49B(l), 32-34. Erickson, B., Lind, E. A., Johnson, B. C., & O'Barr, W. M. (1978). Speech style and impression formation in a court setting: The effects of "powerful" and "powerless" speech. Journal of Experimental Social Psychology, 14(3), 266-279. Federal Interagency Forum on Aging Related Statistics. (2000). Older Americans 2000-Key indicators of well-being. Washington, DC: Data Dissemination Branch of the National Center on Health Statistics. Ferguson, S. A., Hashtroudi, S., & Johnson, M. K. (1992). Age differences in using sourcerelevant cues. Psychology and Aging, 7, 443-452. Fisher, R. P., Brennan, K. H., & McCauley, M. R. (2002). The Cognitive Interview Method to Enhance Eyewitness Recall. In M. L. Eisen, J. AQuas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 265-286). Mahwah, NJ: Lawrence Erlbaum Associates. Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques for investigative interviewing-The Cognitive Interview. Springfield, IL: Charles C. Thomas. Fisher, R. P., McCauley, M. R., & Geiselman, R. E. (1994). Improving eyewitness testimony with the Cognitive Interview. In D. F. Ross, D. J. Read, & M.P. Toglia (Eds.), Adult eyewitness testimony: Current trends and developments (pp. 245-269). New York: Cambridge University Press. Frisina, D. R., & Frisina, R. D. (1997). Speech recognition in noise and presbycusis: Relations to possible neural mechanims. Hearing Research, 106, 95-104. Gabbert, F., Memon, A., & Allan, K. (2003). Memory Conformity: Can eyewitnesses influence each other's memories for an event? Applied Cognitive Psychology, 17(6), 703-714. Giambra, L. M., &Arenberg, D. (1993). Adult age differences in forgetting sentences. Psychology and Aging, 8(3), 451-462. Giles, H. (1994). Talking age and aging talk. In M. L. Hummert, J. M. Wieman, & J. F. Nussbaum (Eds.), Interpersonal communication in older adulthood: Interdisciplinary theory and research. Thousand Oaks, CA: Sage. Gordon-Salant, S. (1987). Age-related differences in speech recognition performance as a function of test format and paradigm. Ear and Hearing, 8, 277-282. Hamm, V. P., & Hasher, L. (1992). Age and the availability of inferences. Psychology and Aging, 7, 56-64. Hartman, M., & Hasher, L. (1991). Aging and suppression: memory for previously irrelevant information. Psychology and Aging, 6, 587-594. Hasher, L., & Zacks, R. T. (1979). Automatic and effortful processes in memory. Journal of Experimental Psychology: General, 108(3), 356-388. Hashtroudi, S., Johnson, M. K., & Chrosniak, L. D. (1989). Aging and source monitoring. Psychology and Aging, 4, 106-112. Hess, T., Auman, C., Colcombe, S. J., & Rahhal, T. A (2003). The impact of stereotype threat on age differences in memory performance. Journal of Gerontology, 58B(l), 3-11. Hulicka, I. M. (1965). Age differences for intentional and incidental learning and recall scores. Journal of the American Geriatrics Society, 13, 639-649. Hulicka, I. M., & Weiss, R. L. (1965). Age differences in retention as a function oflearning.Journal of Consulting Psychology, 29, 125-129. Hultsch, D. F. (1969). Adult age differences in the organization of free recall. Developmental Psychology, 1, 673-678. Hummert, M. L. (1994). Physiognomic cues to age and the activation of stereotypes of the elderly in interaction. International Journal of Aging and Human Development, 39(1), 5-19.

600

MUELLER-JOHNSON AND CECI

Hummert, M. L. (1999). A social cognitive perspective on age stereotypes. In F. B.-F. Thomas & M. Hess (Ed.), Social cognition and aging (pp. 175-196). San Diego: Academic Press. Intons-Peterson, M. J., Rocci, P., West, T., McLellan, K., & Hackney, A. (1999). Age, testing at preferred or nonprefered times (testing optimality) and false memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25(1), 23-40. Jacoby, L. ( 1999). Ironic effects of repetition: measuring age-related differences in memory. Journal of experimental Psychology: Learning, Memory, and Cognition, 25, 3-22. Johnson, M. K., Hashtroudi, S., &Lindsay, D. S. (1993). Source monitoting. Psychological Bulletin, 144,3-128. Kannel, W. B., & Gordon, T. (1973). The Framingham Heart Study: An epidemiologic investigation of cardiovascular disease. Washington, DC: U.S. Government Printing Office. Karpel, M. E., Hoyer, W. J., & Toglia, M.P. (2001). Accuracy and qualities of real and suggested memories: nonspecific age differences. Journal of Gerontology, 56B(2), 103-110. Klaus, P. A. (2000). Crimes against persons age 65 or older, 1992-1997 (No. NCJ 176352). Washington, DC: U.S. Department ofJustice, Office ofJustice Programs. Kline, D. W., & Scialfa, C. T. (1996). Visual and auditory aging. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of aging (pp. 181-203). San Diego: Academic Press. Kneller, W., Memon, A., & Stevenage, S. (2001). Simultaneous and sequential lineups: Decision processes of accurate and inaccurate eyewitnesses. Applied Cognitive Psychology, 15(6), 659-671. Koehnken, G., Milne, R., Memon, A., & Bull, R. (1999). The Cognitive Interview: A metaanalysis. Psychology, Crime and Law, 5(1-2), 3-27. Ki:ihnken, G. (1993). The Structured Interview: A step-by-step introduction. Unpublished manuscript. Koutstaal, W., & Schacter, D. L. (1997). Gist based false recognition of pictures in older and younger adults.]ournal of Memory and Language, 37,555-583. Krausler, D. H., & Phillips, P. L. (1988). Instructional variation and adult age differences in activity memory. Experimental Aging Research 14, 4. Krausler, D. H., & Wiley, J. G. (1990). Temporal memory and content memory for actions: Adult age differences in acquisition and retention. Experimental Aging Research, 16(3), 147-150. Kwong See, S. T., Hoffman, H. G., & Wood, T. L. (2001). Perceptions of an old female eyewitness: Is the older eyewitness believable. Psychology and Aging, 16(2), 346-350. Lachs, M.S., Bove, C., Wearing, M. H., Williams, C. S., Bachman, R. B., & Cooney, L. M. (2001). The clinical epidemiology of crime victimization in older adults: A multidisplinary pilot study. Journal of Elder Abuse and Neglect, 13(3), 79-90. Lachs, M.S., Kossak, A., O'Leary, J., Williams, C. S., Bachman, R. B., & Wearing, M. H. (2003). Crime, health, and police department contact with older adults: A population-based exploratory study. Paper presented at the conference of the Gerontological Society of America, November 21-25, San Diego, CA. Lanceley, A. (1985). Use of controlling language in the rehabilitation of the elderly. Journal of Advanced Nursing, 10, 125-135. Law, S., Hawkins, S. A., & Craik, F. I. M. (1998). Repetition-induced belief in the elderly: Rehabilitating age-related memory deficits. journal of Consumer Research, 25, 91-107. Levy, B. R. (1996). Improving memory in old age by implicit self-stereotyping. Journal of Personality and Social Psychology, 71, 1092-1107. Lindsay, R. C., Lea, J., Nasworthy, G., Fulford, J., Hector, J., LeVan, V., et al. (1991). Biased lineups: Sequential presentation reduces the problem. Journal of Applied Psychology, 76, 796-802. List, J. A. (1986). Age and schematic differences in the reliability of eyewitness testimony. Developmental Psychology, 22, 50-57. Loftus, E. F., Levidow, B., & Duensing, S. (1992). Who remembers best? Individual differences in memory for events that occurred in a science museum. Applied Cognitive Psychology, 6, 93-107.

20. THE ELDERLY EYEWITNESS

601

Marche, T., Jordan, J. J., & Owre, K. P. (2002). Younger adults can be more suggestible than older adults: the influence of learning differences on misinformation reporting. Canadian Journal on Aging, 21 (1), 85-93. May, C. P., Hasher, L., & Stoltzfus, E. R. (1993). Optimal time of day and the magnitude of age differences in memory. Psychological Science, 4, 326-330. Mcintyre, J. S., & Craik, F. I. M. (1987). Age differences in memory for item and source information. Canadian Journal of Psychology, 41, 175-192. McMahon, M. (2000). The effect of the Enhanced Cognitive Interview on recall and confidence in elderly adult. Psychiatry, Psychology and Law, 7, 9-32. Mecacci, L., Zani, A., Rocchetti, G., & Lucioli, R. (1986). The relationship between morningnesseveningness, aging, and personality. Personality and Individual Differences, 7, 911-913. Meissner, C. A., & Brigham, J. C. (2001). 30-years of investigating the own-race bias in memory for faces-A meta-analytic review. Psychology, Public Policy and Law, 7(1), 3-35. Mello, E. W., & Fisher, R. P. (1996). Enhancing older adult eyewitness memory with the cognitive interview. Applied Cognitive Psychology, 10(5), 403-417. Memon, A., & Bartlett, J. C. (2002). The effect of verbalisation on face recognition in young and old adults. Applied Cognitive Psychology, 16, 635-650. Memon, A., Bartlett, J., Rose, R., & Gray, C. (2003). The Aging Eyewitness: Effects of age on face, delay, and source-memory ability. Journal of Gerontology, 58B(6), 338-345. Memon, A., & Gabbert, F. (2003). Improving the identification accuracy of senior witnesses: Do prelineup questions and sequential testing help? Journal of Applied Psychology, 2, 341-347. Memon, A., Hope, L., Bartlett, J. C., & Bull, R. H. C. (2002). Eyewitness recognition errors: The effect of mugshot viewing and choosing in young and old adults. Memory and Cognition, 30, 1219-1227. Mitchell, K. J., Johnson, M. K., & Mather, M. (2003). Source monitoring and suggestibility to misinformation: Adult age-related differences. Applied Cognitive Psychology, 17, 107-119. Moore, M., & Lee, J. (2000). Vulnerable elderly consumer: Examination of an actual victim profile. Journal of Consumer Education, 18, 28-34. Mueller-Johnson, K., & Ceci, S. J. (2004). Memory and suggestibility in older adults: Live event participation and repeated interview. Applied Cognitive Psychology, 18, 1109-112 7. Mueller-Johnson, K., & Ceci, S. J. (in preparation). Two distinct types of interrogative suggestibility? Comparing the misinformation paradigm and the Gudjonsson Suggestibility Scale. Multhaup, K. S., De Leonardis, D. M., &Johnson, M. K. (1999). Source memory and eyewitness suggestibility in older adults. Journal of General Psychology, 126(1), 74-84. National Center on Elder Abuse. (1998). The National Elder Abuse Incidence Study. Washington, DC: National Center on Elder Abuse. Neuschatz, J. S., Preston, E. L., Burkett, A. D., Toglia, M.P., Lampinen, J. M., Neuschatz, J. S., et al. (2005). The effects of post-identification feedback and age on retrospective eyewitness memory. Applied Cognitive Psychology, 19(4), 435-453. Nunez, N., McCoy, M. L., Clark, H. L., & Shaw, L.A. (1999). The testimony of elderly victim/ witnesses and their impact on juror decisions: the importance of examining multiple stereotypes. Law and Human Behavior, 23(4), 413-423. O'Rourke, T. E., Penrod, S.D., Cutler, B. L., & Stuve, T. E. (1989). The external validity of eyewitness identification research: Generalizing across subject populations. Law and Human Behavior, 13, 385-395. Papalia, D. E., Sterns, H. L., Feldman, R. D., & Camp, C. J. (2002). Adult development and aging (2nd ed.). New York: McGraw-Hill. Park, D. C., Puglisi, J. T., & Smith, A. D. (1986). Memory for pictures: Does an age-related decline exist? Psychology and Aging, 1, 11-17. Park, D. C., Royal, D., Dudley, W., & Morrell, R. (1988). Forgetting of pictures over a long retention interval in young and old adults. Psychology and Aging, 3, 94-95.

602

MUELLER-JOHNSON AND CECI

Polczyk, R., Wesolowska, B., Gabarczyk, A., Minakowska, I., Supska, M., & Bomba, E. (2004). Age differences in interrogative suggestibility: A comparison between young and older adults. Applied Cognitive Psychology, 18(8), 1097-1107. Poon, L. W., & Fozard, J. L. (1980). Age and word frequency effects in continuous recognition memory. Journal of Gerontology, 35, 77--86. Puglilsi, J. T. (1980). Semantic encoding in older adults as evidenced by release from proactive inhibition. Journal of Gerontology, 35(5), 743-745. Rabinowitz, J. C. (1984). Aging and recognition failure. Journal of Gerontology, 39, 65-71. Roediger, H. L. I., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. Ross, D. F., Dunning, D., Toglia, M.P., & Ceci, S. J. (1990). The child in the eyes of the jury. Law and Human Behavior, 14(1), 5-23. Rothman, M. B., Dunlop, B. D., & Entzel, P. (2000). Introduction. In M. B. Rothman, B. D. Dunlop, & P. Entzel (Eds.), Elders, crime and the criminal justice system: Myth, perceptions, and reality in the 21st century (pp. xxix-xxxviii). New York: Springer. Rybarczyk, B. D., Hart, R. P., & Harkins, S. W. (1987). Age and forgetting rate with pictorial stimuli. Psychology and Aging, 2, 404-406. Sanders, R. E., Murphy, M.D., Schmitt, F. A., & Walsh, K. K. (1980). Age differences in free recall rehearsal strategies. Journal of Gerontology, 35, 550-558. Schacter, D. L., Koutstaal, W., Gross, M.S., Johnson, M. K., & Angell, K. E. (1997). False recollection induced by photographs: A comparison of older and younger adults. Psychology and Aging, 12(2), 203-215. Schacter, D. L., Norman, K. A., & Koutstaal, W. (1998). The cognitive neuroscience of constructive memory. Annual Review of Psychology, 49, 289-318. Schieber, F., & Baldwin, C. L. (1996). Vision, audition, and aging research. In F. BlanchardFields & T. H. Hess (Eds.), Perspectives on cognitive change in adulthood and aging (pp. 122-162). New York: McGraw-Hill. Scogin, F., Calhoon, S. K., & D'Errico, M. (1994). Eyewitness confidence and accuracy among three age cohorts. Journal of Applied Gerontology, 13, 172-184. Searcy, J. H., Bartlett, J. C., & Memon, A. (1999). Age differences in accuracy and choosing in eyewitness identification and face recognition. Memory and Cognition, 27, 538-552. Searcy, J. H., Bartlett, J. C., & Memon, A. (2000). Relationship of availability, lineup conditions and individual differences to false identification by younger and older eyewitnesses. Legal and Criminological Psychology, 5, 219-236. Searcy, J. H., Bartlett, J., Memon, A., & Swanson, K. (2001). Aging and lineup performance at long retention intervals: Effects of metamemory and context reinstatement. Journal of Applied Psychology, 86, 207-214. Simon, E. (1979). Depth and elaboration of processing in relation to age. Journal of Experimental Psychology: Human Learning and Memory, 5(2), 115-124. Slawinski, E. B., Hartel, D. M., & Kline, D. W. (1993). Self-reported hearing problems in daily life throughout adulthood. Psychology and Aging, 8(4), 552-561. Steinberg, E. P., Javitt, J. C., Sharkey, P. D., Zuckerman. A., Legro M. W., Anderson, G. F., et al. (1993). The content and cost of cataract surgery. Archives of Ophthalmology, 111, 1041-1049. Stuen, C., & Faye, E. (2003). Vision Loss: Normal and not normal changes among older adults. Generations, 27(1), 8-14. Tulving, E., & Thompson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80, 353-370. Tun, P. A. (1998). Fast noisy speech: Adults' age differences in processing rapid speech in noise. Psychology and Aging, 13, 424-434.

20.

THE ELDERLY EYEWITNESS

603

Tun, P. A., & Wingfield, A. (1999). One voice too many: Adult age differences in language processing with different types of distracting sounds. Journals of Gerontology, 54B (5), 317-32 7. Tun, P. A., Wingfield, A., Rosen, M. J., & Blanchard, L. (1998). Response latencies for false memories: Gist-based processes in normal aging. Psychology and Aging, 13 (2), 230-241. United Nations Population Division. (2001). World urbanization prospects: The 2001 revision. Retrieved 16 March, 2004, from http://esa.un.org./unpp United Nations Population Division. (2002). World population prospects: The 2002 revision. Retrieved 16 March, 2004, from http://esa.un.org/unpp U.S. Congress Office of Technology Assessment. (1986). Hearing impairment and elderly peopleA background paper (No. OTA-BP-BA-30). Washington, DC: U.S. Government Printing Office. Uttl, B., & Graf, P. (1993). Episodic spatial memory in adulthood. Psychology and Aging, 8(2), 257-273. Verhaeghen, P., Marcoen, A., & Goossens, L. (1992). Improving memory performance in the aged through mnemonic training: A meta-analytic study. Psychology and Aging, 7(2), 242-251. Weinstein, B. (2003). A primer on hearing loss in the elderly. Generations, 27(1), 15-19. Wells, G. L., & Leippe, M. R. ( 1981). How do triers of fact infer the accuracy of eyewitness identifications? Using memory for peripheral detail can be misleading. Journal of Applied Psychology, 66(6), 682-687. West, R. L., & Boatwright, L. K. (1983). Age differences in cued recall and recognition under varying encoding and retrieval conditions. Experimental Aging Research, 9, 185-189. West, R., Murphy, K. J., Armilio, M. L., Craik, F. I. M., & Stuss, D. T. (2002). Effects of times of day on age differences in working memory. Journals of Gerontology, 57B(l), 3-10. Witte, K. L., Freund, J. S., & Brown-Whistler, S. (1993). Adult age differences in free recall and category clustering. Experimental Aging Research, 19, 15-28. Working Group on Speech Understanding and Aging. (1988). Speech understanding and aging. Journal of the Acoustical Society of America, 83, 859-895. Wright, A. M., & Holliday, R. E. (2003). Interviewing elderly witnesses and victims. Forensic Update, 73, 20-24. Wright, D., & Stroud, J., N. (2002). Age differences in lineup identification accuracy: People are better with their own age. Law and Human Behavior, 26(6), 641-654. Yarmey, A. D. (1982). Eyewitness identification and stereotypes of criminals. In A. Trankell (Ed.), Reconstructing the past: The role of psychologists in criminal trials (pp. 205-225). Stockholm: Norstedt & Souers. Yarmey, A. D. (1984). Accuracy and credibility of the elderly witness. Canadian Journal on Aging, 3, 79-90. Yarmey, A. D. (1996). The elderly witness. InS. L. Sporer, R. S. Malpass, & G. Koehnken (Eds.), Psychological issues in eyewitness identification. Hillsdale, NJ: Lawrence Erlbaum Associates. Yarmey, A. D. (2001). The older eyewitness. In M. B. Rothman & P. Entzel (Eds.), Elders, crime and the criminal justice system (pp. 127-148). New York: Springer. Yarmey, A. D., Jones, H. T., & Rashid, S. (1984). Eyewitness memory of elderly and young adults. In D. J. Muller, D. E. Blackman, &A. J. Chapman (Eds.), Psychology and law (pp. 215228). Chichester, UK: John Wiley & Sons. Yarmey, A. D., & Kent, J. (1980). Eyewitness identification by elderly and young adults. Law and Human Behavior, 4, 359-371. Yoon, C., May, C. P., & Hasher, L. (2000). Aging, circadian arousal patterns, and cognition. In D. C. Park & N. Schwarz (Eds.), Cognitive aging: A primer. Philadelphia: Taylor & Francis.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

21 False Memory Susceptibility in Older Adults: Implications for the Elderly Eyewitness Donna J. LaVoie Heather K. Mertz Saint Louis University

Tracey L. Richmond Washington University School of Medicine

The laboratory study of false memory; while not a new area of investigation, has received particular attention over the last 10 years with the reintroduction of Deeses's (1959) converging associates list-learning paradigm by Roediger and McDermott (1995). In this particular paradigm (known as the Deese-Roediger-McDermott or DRM paradigm), false recall and recognition of nonpresented critical target words that are semantically related to a previously presented list of words occurs with relatively high frequency. Roediger and McDermott have reported false alarm rates of 84% for related lures, a number comparable to the hit rate (86%) for studied items; the false alarm rate for unrelated items is significantly lower at only 2%. Similar types of false recall and false recognition have been observed in a variety of paradigms employing a wide variety of materials, including categorized word lists (e.g., Smith, Ward, Tindell, Sifonis, & Wilkenfeld, 2000), sentences (e.g., Bransford & Franks, 1971), text passages (Bower, Black, & Turner, 1979), and pictures (e.g., Israel & Schacter, 1997). The common thread across all of these studies is the claim by participants that a novel word, object, or event is familiar and, therefore, has been presented earlier. This phenomenon has been noted in a wide range of individuals, including children (e.g., Ghetti, Qin, & Goodman, 2002), neurologically damaged individuals (e.g., Schacter, Curran, Galluccio, Milberg, & Bates, 1996), amnesics (e.g., Schacter, Verfaellie, Anes, & Racine, 1998), and older adults (e.g., Tun, Wingfield, Rosen, & Blanchard, 1998). This chapter focuses on this last group of individuals, drawing connections between false memory susceptibility and eyewitness memory 605

606

LAVOlE, MERTZ, RICHMOND

in older adults. To preview our conclusions based on this review, both the false memory literature and the eyewitness memory literature by and large point to an increased susceptibility in older adults relative to young adults to making memory errors, particularly in the recollection of event details. This is a disconcerting conclusion for anyone interested in the accuracy of elderly eyewitnesses, as it suggests that their testimony is more likely to be inaccurate than a younger adult's testimony would be. With U.S. Census Bureau projections (2004) estimating that 19.6% of our population will be over age 65 by the year 2030, it is very likely that this age group will find itself in contact with the legal system, either as victims of or witnesses to crime. It is important, then, to understand the conditions under which older adults' memory for events is compromised. As our review points out, although the weight of the empirical evidence indicates that older adults possess less accurate memories, there are some situations in which older adults are no more susceptible to memory errors than are younger adults, so a clear understanding of the limits of memory accuracy in older adults is necessary before concluding that in all cases they are poor eyewitnesses. The chapter begins with an overview of some selected findings regarding false memory susceptibility in older adults, in particular false memories for semantic or associative information, as these data have been predominant in recent years. We examine these findings in light of two of the major theoretical arguments for why older adults may be particularly susceptible to false and other similar illusory memories to provide some insight into the set of circumstances that make older adults vulnerable to memory errors. We then describe some of the elderly eyewitness literature that specifically addresses issues related to the accuracy of eyewitness memory. Finally, we outline some of the legal implications these findings have for the elderly eyewitness and describe situations in which the older adult may be less likely to provide accurate information about a witnessed event.

AGING AND FALSE MEMORY

A well-documented finding across the cognitive literature is an age-related reduction in episodic memory (for a review, see Balota, Dolan, & Duchek, 2000). Evidence is also accumulating from a variety of sources that in addition to reductions in episodic memory performance, an increased susceptibility to false or illusory memories occurs with advancing age. For example, in an early study by Smith (1975), older adults produced almost twice as many false alarms to synonyms of studied words on a recognition test as did young adults. Similar age differences in false recognition of semantically related lures have been noted by others (e.g., Norman & Schacter, 1997; Tun, Wingfield, Rosen, & Blanchard, 1998, Experiment 2). Despite these findings, however, there have been failures to detect age differences in false memory susceptibility (e.g., Tun et al., 1998, Experiments 1 and 3). Nonetheless, the bulk of the evidence strongly supports the conclusion that older adults are more susceptible than young adults to memory errors, illusions, and false recollections. Why? A number of theoretical arguments have been proposed, but we focus our discussion on two that we think have strong implications for the per-

21. FALSE MEMORY IN OLDER ADULTS

607

formance of an elderly eyewitness: (1) age-related failures in source monitoring and (2) an age-related reliance on gist-based processing.

SOURCE-MONITORING FAILURES

According to Johnson and colleagues (e.g., Johnson, Hashtroudi, & Lindsay, 1993), the characteristics of a memory that help determine its source (e.g., perceptual information, contextual details, or even a record of the cognitive operations involved in creating the memory) are influential in determining whether an activated memory trace is judged as being real or imagined. This latter type of source monitoring is a form of reality monitoring in which discriminations are made between memories for internally generated thoughts and memories for perceived events. Generally, perceived events are richer in sensoryperceptual characteristics and contextual details than internally generated memories, which tend to include more information about the cognitive operations used in their creation. As applied to false memories, this framework predicts that false recollections arise out of source confusion, with critical items being falsely recalled or recognized because there is a failure to identify the source of familiarity for the item as being from internal activation that resulted from associative processing, and not from external activation due to list study. A number of studies indicate that older adults have particular difficulty with source monitoring (e.g., Brown, Jones, & Davis, 1995; Henkel, Johnson, & De Leonardis, 1998), and so this framework predicts that age-related failures in source monitoring will lead to increased false memory susceptibility in older adults. To illustrate this point, consider the DRM task, widely used in laboratory investigations of false memory. The lists used in this paradigm contain varying numbers (typically 12-15 items) of associated words, and this association converges on at least a single item, the critical lure. As has been discussed by Roediger and colleagues (e.g., Roediger, McDermott, & Robinson, 1998), this critical item is likely to be highly activated as aresult of the converging association spreading from presented list items to the critical item (a process first labeled by Underwood [1965] as an implicit associative response, or IAR). The critical item, consequently, has a high probability (relative to noncritical items) of being falsely recollected. According to the source-monitoring framework, however, this is not simply because the critical item is activated, but rather because the source of the activation is difficult to ascertain, and its activation is misattributed to actual list study. Such difficulty is more plausible when the activated critical item is more similar to the actual studied list items, whose source of activation comes from actual list presentation, because the qualitative differences between the two sources are fewer than if such similarity did not exist. Although the two types of items differ greatly in terms of their sensory-perceptual qualities, they overlap in terms of their semantic qualities and hence are more confusable than if such overlap did not exist. Indeed, Schacter, Norman, and Koutstaal (1998) have argued that the high rates of false recall and recognition that occur in the DRM paradigm are partially attributable to the fact that this paradigm emphasizes the common semantic features of presented items, rather than the distinctive

LAVOlE, MERTZ, RICHMOND

608

details of individual presented items. To the extent that older adults have difficulty monitoring the distinctive perceptual qualities of list items that make them discriminable from nonstudied but highly associated (and likely activated) critical lures, a sourcemonitoring failure in older adults may produce higher rates of false memory relative to young adults. LaVoie and Faulkner (2000) used a variation of the DRM paradigm to test these basic arguments. In this study, older adults studied DRM lists, after which they were given a forced-choice recognition task for studied items. Three types of forced-choice pairings were used to assess item memory (studied items +fillers), gist memory (i.e., memory for list themes, assessed through critical lure+ filler item pairings), and false memory (studied item + critical lure pairings). These latter two types of pairings differ along one important dimension (i.e., the qualitative information activated about the critical lure that makes it distinct from other studied and nonstudied items), even though both pairings allow for the appearance of false recollections. In the item memory and gist memory pairings, studied items and critical lure items are more likely to be chosen than fillers, simply because they are activated (from either actual list item presentation or spreading activation through associative processing, respectively). To the extent that this activation is similar in young and older adults, no age differences should be observed in the proportion of studied items or critical lures chosen over filler items, as was the case (see Figure 21-1). However, in the last type of pairing (i.e., the studied+ critical lure pairs),

Recognition Type FIGURE 21-1. Age effects in memory for studied items, studied list themes (gist memory), and false memories in a forced-choice recognition task. Reliable age differences (p < .05) only exist in false memory (adapted from LaVoie & Faulkner, 2000).

21.

FALSE MEMORY IN OLDER ADULTS

609

in which both items are activated, the correct choice of the studied item can only be made if one is able to discriminate between the two sources of activation. A failure to do so will lead to more errors. We found that the only age differences observed were for this last pairing, with older adults labeling the critical lure as the studied item almost twice as often as the young adults did. LaVoie and Willoughby ( 1998) have obtained other evidence for the role of sourcemonitoring errors in false memories for words, using the DRM lists by repeating list presentations during two study phases. During the first study phase, young and older subjects were presented with a 12-item DRM list (the short list condition, containing associates of a critical nonpresented (CNP) item, e.g., bed, rest, awake, tired, etc., but not the item itself). This was immediately followed by a second study phase consisting of a longer 30-item list, which contained the CNP (e.g., sleep) for the Phase 1 short list, plus other unrelated items. After seeing both the short and long lists, subjects were then given a recognition test for list items, where in addition to making old/new judgments, they were asked to make source judgments for items they had labeled "old" by identifying the list on which they had seen the item. The important finding here was that older adults were more likely than young adults to incorrectly choose the short list as the source of the critical lure, when in fact it had only been presented in the long list (see Figure 21-2). In this case, older adults incorrectly identified the source of the critical item as a list of items that shared semantic features with the lure, suggesting that they had difficulty utilizing the contextual qualities that discriminated the critical lure from the short-list items to accurately recollect its source. Similar age differences in source confusion errors for performed, watched, or imagined events have been reported by Cohen and Faulkner (1989). More recently, Hedden and Park (2003) have obtained evidence that source confusion errors are the primary cause of retroactive interference

Item type to source

FIGURE 21-2. Source attributions for studied and critical items presented in a list of unrelated items. Note that the correct response for studied items is "Studied to Short," and for critical items "CNP to Long" is the correct response (from LaVoie & Willoughby, 1998).

610

LAVOIE, MERTZ, RICHMOND

in older adults, providing additional support for the argument that older adults experience difficulty in the appropriate use of contextual information when retrieving information. Work by Schacter, Koutstaal, Johnson, Gross, and Angell (1997), who used a different methodology with pictures instead of word lists, also implicates source-monitoring failures as a likely locus for age differences in susceptibility to false recollections. In this study, young and older adults viewed photographs of events that had been part of a previously viewed videotape (true photographs) or had not been viewed previously but depicted events similar to those seen in the videotape (false photographs). Immediately following viewing of the videotapes, subjects were shown the true and false photographs and were asked to rate the similarity of these photos to events seen in the video. Two days later, subjects were given a recognition test consisting of verbal descriptions of objects, some of which were in the video and photos, some of which were from the photos only, and some of which were new. Subjects were asked to provide two different types of responses on the recognition test. First, they were to make old/new judgments of items and judge as old only those items that had appeared in the video, forcing subjects to carefully discriminate between items presented from alternative sources. Second, the qualitative aspect of subjects' recollective experience was assessed through Remember/ Know judgments for items called old (provide a "Remember" response if specific details about the presentation can be recalled, and a "Know" response when there is a general feeling of familiarity for the item in the absence of specific details) and through subjective ratings of the characteristics of these memories, made with the use of a questionnaire. The main effect of age on false recognition judgments (in this case, labeling a new or photo-only object as old) was marginal in Experiment 1 (p = .10), but it was reliable in Experiment 2, where the procedure was modified slightly to make it more difficult to monitor the source of activation for objects presented in photos, presumably increasing the probability that older adults would make more source misattributions. In addition to this increase in source errors, older adults were significantly more likely to provide Remember responses to photo-only objects than were young adults, indicating they had a different recollective experience than did the young adults for these items. Older adults were also more likely than young adults to report what the described object looked like in the video, although it had been presented in the photo only, and to report a stronger sense of familiarity for this object. These findings highlight an important factor regarding false memory susceptibility in older adults, that is, that the qualitative differences between real and false memories may be smaller in older adults, making it more difficult for older adults to accurately recollect the source of an experience, leaving them more vulnerable to false recollections. Norman and Schacter (1997) provide additional data that speaks directly to this issue. Using a variation of the DRM paradigm, they found that older adults showed smaller differences in attribute ratings for sensory (i.e., sound) and contextual (i.e., list position) details of true and false memories than did young adults, along with higher rates of false recall and recognition of critical items. Combined, the findings described above suggest that older adults are less able than young adults to use the qualitative distinctive details about list items that support true recall and recognition. Consequently for older adults, when an activated item is retrieved, a feeling of familiarity, in the absence of details about the source of the activa-

21.

FALSE MEMORY IN OLDER ADULTS

611

tion, will lead to an increased likelihood that this item will be reported as experienced. One interpretation of these findings is that the inability by older adults to utilize perceptual details as an aid to more accurate recollection may arise from an inability to adequately encode perceptual details (for a review, see Spencer & Raz, 1995) and that increased false memory susceptibility in older adults may really reflect a perceptual encoding failure. This issue has been addressed by examination of older adults' suppression of false recognition as a consequence of increasing the perceptual distinctiveness of studied items to facilitate source monitoring. If distinct perceptual details can be encoded, then false recollection should be decreased relative to conditions where such information is less distinct. Kensinger and Schacter (1999), for example, attempted to increase the perceptual qualities of list items by repeating study lists. Whereas young adults showed decreasing false recall over trials, older adults did not. Budson, Daffner, Desikan, and Schacter (2000), however, found that older adults can suppress false recognition after five study repetitions to rates comparable of those seen in young adults, although their false recognition rates fluctuated across trials. Schacter, Israel, and Racine (1999) have also been able to obtain suppression of false recognition in young and older adults with pictorial encoding versus word encoding of list items (but it is important to note that older adults still had marginally higher false recognition rates than did the young adults, p = .089). When given warnings about the false recognition effect at study, McCabe and Smith (2002, Experiment 1) found that both young and older adults could use these warnings to effectively suppress false recognition relative to an unwarned control group, but when warnings were given after study and prior to retrieval (Experiment 2), only young adults showed suppression of false recognition. The evidence is mixed, then, on whether older adults can suppress false recognition by encouraging more distinctive encoding, as only strong manipulations designed to enhance the perceptual distinctiveness of studied items produce such suppression in this age group. Nonetheless, the observation of suppression makes it difficult to firmly conclude that older adults fail to encode the perceptual qualities that distinguish among similar activated items, but rather points to a failure to adequately use this information in the recollection of events. Further evidence supporting this argument comes from recent work by Koutstaal (2003), who notes that older adults encode the perceptual details of events that help differentiate them from similar but novel events; however, older adults are less likely than young adults to successfully use these details during recognition. In this study, participants were given three different types of memory tests for studied photographs of common objects. On each test, participants were required to make judgments about studied (same) exemplars, exemplars that were perceptually and conceptually similar to studied exemplars (different exemplars), and novel items. For the identical recognition test, participants were to make standard old/new judgments for items, but were encouraged to call an item old only if it was identical to one presented previously. For the meaning-based recognition test, participants were to make similar old/new judgments, but were told to designate an item as old if it was either identical to one seen earlier or similar to one presented earlier. Finally, participants performed a repetition priming (size judgment) task for same and different exemplars. Performance on these latter two

612

LAVOlE, MERTZ, RICHMOND

memory tests provides information about how the perceptual details of studied ob, jects, if they have been encoded, influence recognition. Specifically, if perceptual details about the studied items are available to help with recollection, then there should be greater recognition and greater priming effects for same exemplars (whose perceptual details are reinstated at test) as compared with different exemplars. No age differences were found in the magnitude of this hypothesized perceptual similarity effect, with both age groups demonstrating higher recognition rates on the meaning,based recognition test, and larger priming effects on the priming task for same exemplars as compared with different exemplars. Older adults, however, had higher rates of false recognition than did the young adults for different exemplars on the identical recognition test. Overall, these results indicate that older adults were able to encode perceptual details of studied items, but only used those details when they did not need to be deliberately retrieved. This suggests that older adults may have difficulty employing the more effort, ful, controlled type of processing that helps differentiate between old and conceptually similar novel events, and without a strong manipulation to enhance perceptual qualities of studied items, older adults will continue to show high rates of false recognition and recall of critical novel words and objects. This conclusion suggests that older adults seem to be reliant on the semantic or conceptual information present in an event to drive recollection of that event, rather than use the perceptual qualities of an event to assist with its recollection. This argument is the central point of the next theoretical account, ing for age,related differences in false memories that we discuss here.

OVER-RELIANCE ON GIST PROCESSING

The argument that older adults rely on conceptual processing at the expense of percep, tual processing when encoding and retrieving information extends from an application of fuzzy trace theory (Brainerd & Reyna, 1990) to false memory. The main ideas of this theory are that two memory representations of an event are formed when it is initially processed-a trace that maintains the surface form of the object, word, etc. (i.e., the verbatim trace) and a trace that maintains the general semantic,conceptual features (i.e., the gist trace). Research on a variety of topics (e.g., memory for segmented speech; Wingfield, Tun, & Rosen, 1995) indicates that older adults tend to rely on gist,based processing rather than verbatim processing during memorial processing. As applied to false memories, it has been argued (e.g., Tun, Wingfield, Rosen, & Blanchard, 1998) that this reliance on gist,based processing makes older adults particularly vulnerable to false memories, especially in those situations where such processing is not an efficient way in which to retrieve information. For example, Tun et al. have demonstrated that age dif, ferences in false recognition are greatest when individuals must base their recognition decisions on item,specific information rather than general thematic information about studied word lists. In Experiment 2 of their study, the recognition test included both the standard critical lure used with the DRM lists (a strong associate of those items) and weak associates of list items. They hypothesized that the inclusion of these two types of items, encouraging participants to employ a more stringent verbatim,based processing

21.

FALSE MEMORY IN OLDER ADULTS

613

strategy to make recognition judgments as responding to items purely on the basis of relatedness, would produce less accurate recognition. Congruent with their hypotheses, Tun et al. found that older adults, relative to young adults, had higher rates of false recognition for both the standard (strong associate) critical lures and the weakly associated lures, reflecting a tendency to respond on the basis of semantic relatedness when making recognition judgments. Additional evidence for the role of gist-processing is provided by Koutstaal and Schacter (1997), who used a picture recognition paradigm in which the critical lure items were not likely to have been generated or directly activated during study. The critical lures used in this study consisted of detailed colored pictures of nonstudied exemplars (e.g., a German shepherd) of an object category (e.g., dogs) that had been presented with other studied exemplars (e.g., dachshund) during a study phase. This particular manipulation assesses false recognition of lures that share conceptual properties or general semantic similarity, rather than false recognition of lures that may have received direct activation through converging association, and so provides a cleaner test of the gist processing account than do methodologies that use the DRM (or similar) materials. False recognition of novel lures occurs with this methodology not because of a failure to use perceptual information to make recognition judgments, but rather because of a failure to make these judgments on the basis of conceptual similarity. Across three experiments, both age groups showed false recognition of these novel within-category lures, but older adults demonstrated substantially higher rates of false recognition (64%) than did the young adults (29%). This higher rate of false recognition was accompanied by lower hit rates for distinctive studied items that were not categorically related to other studied items, and for which gist processing would not aid in recognition. Finally, signal detection analyses revealed that older adults showed substantially higher sensitivity to (or willingness to act upon) gist information than did young adults. These results clearly implicate a reliance on gist processing in older adults that made it difficult for them to reject new items from previously studied categories. Although the evidence provided above points to gist versus verbatim processing differences between young and older adults as an important factor in false memory susceptibility, a different type of paradigm in which age differences are observed in false recognition provides data that implicate additional age-related processing differences that may also affect false memory susceptibility. Specifically, Jacoby and colleagues (Jacoby, 1999; Jennings &Jacoby, 1997), using a repetition-lag procedure with unrelated words, note that older adults are more likely than young adults to falsely recognize a repeated item. In this procedure, subjects are asked to study a list of words and later are given a recognition test for this list. In this test, new items are repeated once after a varying number of intervening items, and subjects are asked to identify the old items. All participants are warned about the repeated new items and are explicitly told that if an item is repeated, it is new and should be given a no response. A repetition error occurs if the new item is misidentified as old. Jacoby and colleagues have reported more repetition errors in their older adults than in the young adults. Similar effects of repetition have been found in false-fame paradigms, where repeated viewings of nonfamous faces or names have led to increased probabilities in older adults of incorrectly labeling the face or name

614

LAVOlE, MERTZ, RICHMOND

as famous, with opposite effects in young adults (e.g., Bartlett, Strater, & Fulton, 1991). These effects, while not identical to the false memory effects described earlier, are nonetheless memory errors that reflect the acceptance as old of something that was not part of an original experienced event. More importantly, however, they point to an additional set of processing factors and situations in which the older adult is more likely to misremember events. Jacoby identifies two specific processes, recollection and familiarity, as underlying memory judgments in this task. These processes have been hypothesized (e.g., Mandler, 1980) to be independent retrieval processes in dual-process models of memory. Of the two processes, familiarity is the faster, and generally reflects the strength of an item, such that previously encountered items will be more familiar (i.e., have a stronger memory trace) than new items. Recollection, on the other hand, is the more slowly operating consciously controlled retrieval process that allows for the retrieval of qualitative information about a study event. Jacoby has argued that the ability to respond correctly to repeated items in the recognition-lag procedure requires recollection of an item's prior occurrence and, in particular, recollection that a new repeated item had been presented earlier in the recognition test (and therefore should be rejected as an old studied item). More importantly, this recollection must be sufficient to overcome any automatic influence of memory that produces a feeling of familiarity for the repeated new item. A failure to do so will lead to a new item being incorrectly labeled as old, resulting in a repetition error. A growing body of evidence indicates that older adults show reductions in recollection, whereas familiarity processes remain unchanged across the life span (for an overview of these issues, see Light, Prull, LaVoie, & Healy, 2000), so to the extent that older adults have failures in recollection of an item's prior occurrence, their performance will be more heavily influenced by familiarity than is a young adult's, resulting in more errors. This tendency to respond on the basis of familiarity suggests that older adults will not just be more susceptible to false recognition of semantically related lures, but will also be more likely to falsely recognize any information that is repeated post-study, simply because that information has increased trace strength. The implications of this repetition effect for the elderly eyewitness is discussed later.

SUMMARY Our review of false memory susceptibility in older adults has provided details of a few select studies, primarily because these few studies illustrate key situations in which older adults are particularly vulnerable to committing memory errors. These include situations where careful source discrimination must occur (a) between highly semantically similar events (where a reliance on gist processing may limit an older adult's ability to accurately utilize the perceptual qualities of an event to distinguish it from related thoughts) or (b) when information becomes increasingly familiar, regardless of its conceptual relationship to the original event. The likely causes of the memory errors in these situations are an age-related difficulty in accurately monitoring sources of activation, coupled with a related failure to utilize verbatim/perceptual, or other controlled processing in an efficient manner. We do not believe that either one of these factors is the sole underlying

21. FALSE MEMORY IN OLDER ADULTS

615

cause of false memory susceptibility in older adults, as there is sufficient supportive evidence for both. It is important to note that although older adults are able to engage the processes necessary to suppress false memory, they do not appear to do so spontaneously, and so they are particularly vulnerable in everyday situations to the types of memory errors described in our brief review. The next section of this chapter examines the implications of these findings regarding eyewitness issues in the elderly.

IMPLICATIONS OF FALSE MEMORY SUSCEPTIBILITY FOR THE ELDERLY EYEWITNESS

Classic eyewitness research (e.g., Loftus & Palmer, 1974; Loftus, 1975, 1979) indicates that questioning procedures and post-event information can influence eyewitness memory. One well-established finding is that an individual's memory for an event or for objects/actions within an event can be altered by misinformation about actual observed events/objects or novel information about false events/objects. This alteration can produce striking memory errors, with individuals sometimes changing the details of a witnessed object (e.g., recollecting a stop sign for a yield sign) or even reporting the presence of objects (e.g., broken glass) that were not part of a witnessed event (but could have been, given the context of the original event). In these examples, the recollected information is conceptually congruent with the witnessed event, and so in this sense, these errors are similar to the types of errors seen in false memory studies like those described earlier. A reasonable (and rather obvious) conclusion to be made from the false memory literature, as extended to elderly eyewitness memory, is that older adults should be more susceptible to misinformation effects, as a type of memory error, than are young adults. Indeed, Loftus, Levidow, and Duensing (1992) report that one important factor related to the accuracy of eyewitness memory is age; older adults in their study were less accurate overall than were young adults in their recollection of events presented in a film clip of an assault. More importantly, however, Loftus et al. noted that older subjects were less able, relative to young adults, to detect post-event misinformation, and young adults were better at detecting discrepancies in queries about details of witnessed events. For example, one misleading question in their study asked participants if the male aggressor wearing the "black jacket" hit the victim over the head with a stick; the discrepant detail was the jacket color (it was actually blue). Older adults were less likely to detect this discrepant detail than were young adults. This particular outcome exemplifies what we think is one important implication of false memory findings for eyewitness situations, that is, that older adults are more susceptible to misinformation about event details (and not just general event information) than young adults are. These kinds of errors have been widely reported in the literature (for a review, see Bornstein, 1995) and likely arise out of a failure in older adults to effectively utilize perceptual details to accurately recollect events, especially if there is overall congruency between the witnessed event and the query about that event, as was the case in the study by Loftus et al. The changing of an event detail during questioning about a witnessed event produces a situation analogous to the typical false memory laboratory study, where critical lures can be

616

LAVOlE, MERTZ, RICHMOND

perceptually and/or conceptually similar to studied items. In these types of situations, the evidence strongly suggests that older adults are unable to use qualitative distinctive details about studied items or witnessed events to correctly reject related lures or misleading queries, and so the older adult is more susceptible to memory errors. The eyewitness situation, however, unlike the laboratory investigations of false memory, brings with it an additional factor that may further limit older adults' ability to effectively utilize perceptual details to support accurate recollection, that is, the witness's familiarity with the event and its details. Haber and Haber (2000) have argued that this is one important factor that may lead to errors in the recollection of a witnessed event, as a lack of familiarity with an event may lead to the observation of fewer details and the encoding of fewer distinctive details. Brimacombe, Quinton, Nance, and Garrioch (1997) have termed this factor an "encoding handicap" and suggest that a lack of familiarity with a specific environment may make that environment less memorable. Although their study does not provide a critical test of this factor, they argue that to the extent that a lack of familiarity with an environment is more present in one age group, age differences in memory accuracy for that event may appear. In support of their argument, Brimacombe et al. note that their older adult subjects were less accurate, and in particular less accurate regarding event details, in response to both direct examination questioning (where there were no leading or suggestive questions) and cross-examination questioning (which included leading questions) of a videotaped crime sequence occurring on a college campus. An extension of laboratory investigations of false memory suggests the simple conclusion here that there was an impairment in the older adults that led them to utilize event details that not only made them more susceptible to misleading questions, but also left them with an overall poorer memory for the witnessed event. However, in the more complex eyewitness situation, this impairment may be exacerbated by a lack of familiarity with the specific environment in which the witnessed event took place (as Brimacombe et al. suggest) that produces an additional failure to encode relevant details. This remains an open question, however, as research has not specifically examined older adults' memory for events that occur in environments more familiar to them as compared with less familiar environments. Such comparisons are needed to determine the types of crime circumstances in which an older adult is likely to have an overall less accurate memory and to determine whether such complex encoding situations produce an encoding handicap not typically seen in the more controlled laboratory studies investigating false memory susceptibility where event details are encoded but not utilized. The fact that older adults fail to utilize, or in some circumstances fail to fully encode, event details might lead one to conclude that they are consistently more susceptible to misleading questions and misinformation effects. This does not appear to be the case, however, inasmuch as some studies report less accurate overall memory in older adults relative to young adults, but no increased susceptibility to misinformation. For example, Coxon and Valentine (1997) measured age differences in accuracy for nonmisleading situations and suggestibility in misleading situations in adolescents/young adults aged 16-19 years and older adults aged 60-85 years. Subjects watched a simulated kidnapping and later were given a recall test with two types of questions: (1) questions contain-

21.

FALSE MEMORY IN OLDER ADULTS

617

ing noncritical information to assess for accuracy of memory for details and (2) critical questions, which for subjects in the misinformed condition contained misleading information about events in the video (for control subjects no such misleading information was presented). Young adults provided more accurate responses to the noncritical questions (63.1 o/o correct responses) than older adults (40.3% correct). For the critical questions, there were no reliable differences between the young (76.9% correct) and older adults (65.9% correct), although misled subjects were less accurate overall than the control subjects (50.95% correct versus 71.4% correct, respectively, for these two age groups). These findings indicate that although overall older adults were less able to give accurate accounts of the videotaped event, they were no less able than young adults to resist the influence of misleading post-event information in their recall of events. Similar findings have been reported by Bornstein, Witt, Cherry, and Greene (2000). Participants viewed a videotaped car accident and then completed an event recall questionnaire with either consistent (control) or inconsistent questions about events in the video, followed by a forced-choice recognition test. In general, participants in the consistent condition accurately recalled more information than misled participants did. As was the case in the Coxon and Valentine (1997) study, older adults showed poorer recall of general event details than did young adults, but they were not more susceptible to the misleading information, as evidenced by forced-choice test scores (older control subjects 80o/o correct, older misled subjects 75o/o correct, as contrasted to young adult scores of lOOo/o correct in the control condition and 65o/o correct in the misled condition). The lack of age differences in susceptibility to misleading information can be interpreted as being the result of an overall weaker memory for the witnessed event. This is a likely possibility, but other work refutes this argument. For example, Cohen and Faulkner (1989) found that unless older adults were misled, they performed as well as the young adults. Older misled subjects were more influenced by misleading post-event details than were the misled young adults. This finding is consistent with our argument that misinformation about event details will produce more memory errors in older adults than in young adults, an argument that appears to be in conflict with the findings of Coxon and Valentine and Bornstein et al. An important factor to consider here, which may provide some resolution to this conflict, is that the manner in which the misinformation was presented in the Cohen and Faulkner study was a much stronger manipulation than were the manipulations in the studies by Coxon and Valentine and Bornstein et al. Specifically, Cohen and Faulkner introduced the misinformation through a written narrative about the event presented after the videotape of the original event was viewed (control subjects read an accurate narrative). Such a manipulation creates a situation where careful recollection of event details is necessary to overcome the influence of the misinformation and any feeling of familiarity that may be arising from its presentation in the post-event narrative. In the studies by Coxon and Valentine and Bornstein et al. the misinformation was placed only within critical questions, a much weaker manipulation (although one that still misleads subjects, as evidenced by the main effect of misinformation seen in the misled relative to control subjects in both studies). To provide an example, misled subjects in the Coxon and Valentine study were asked, ''After running into the corridor to use the telephone, what did the nurse tell security to do?" (in the video

618

LAVOlE, MERTZ, RICHMOND

the nurse ran into a bedroom). Control subjects were asked, "What did the nurse ask security to do?" In both of these questions, subjects are asked about the overall actions of a key player (i.e., a nurse notifying security), and, in that sense, the question may be encouraging subjects to recollect the general set of circumstances of the witnessed event, rather than the specific details of that event (i.e., the location of the nurse when she notified security). As such, these questions encourage the use of gist-processing in order to recollect the event. As was discussed earlier in our review of the false memory literature, Tun et al. (1998) noted that when subjects were encouraged to use gist-processing toretrieve studied list items, there were no age differences in susceptibility to false recall or false recognition (but there was a nonsignificant trend for a main effect of age in correct recall, i.e., young adults recalled more than the older adults). Age differences in false memory susceptibility were apparent only when verbatim processing was made more salient to the retrieval task demands. A similar situation may be operating here, where the type of questioning employed in the misleading conditions of the studies by Coxon and Valentine and Bornstein et al. does not create a demand for verbatim processing, and so young and older adults show no differences in susceptibility to the misleading information, as the two groups are equally susceptible. This is another area open for research within the domain of elderly eyewitness memory research, as the false memory research suggests that retrieval tasks that encourage verbatim processing are more likely to yield age differences in memory errors than are tasks that are dependent upon gist processing. Employing questioning that requires retrieval of only a general set of circumstances may produce overall less accurate reports of events by elderly eyewitnesses, but it may not be sufficient to make them more susceptible to misinformation than young adults. This susceptibility may only occur when accurate event retrieval is reliant on verbatim processing. The presentation of misleading post-event information has been shown to influence memory for an event, and additional research suggests that the number of times a witness is exposed to post-event information influences his or her susceptibility to that information. It is not uncommon for eyewitnesses outside of the laboratory to be questioned multiple times about the event they experienced. Not only may they may have to recount the event for police/legal authorities-media sources or friends and family may question them as well. Each one of these recollections potentially exposes the witness to additional details (consistent or inconsistent) provided by post-event questions and increases the probability that the recollection may be tainted. The existing research on young adults supports this claim, indicating the existence of a negative relationship between frequency of exposure to misleading information and memory for an event. Specifically, as the number of exposures to misleading information increases, so does the susceptibility to false memory for the suggested information (Mitchell & Zaragoza, 1996; Warren & Lane, 1995; Zaragoza & Mitchell, 1996). Little research, however, has been conducted examining the effects of repeated exposure of post-event information on older adults. We think this is an important area of investigation within elderly eyewitness memory research, as false memory studies indicate that young and older adults respond differently to repeated information. As a reminder, in the DRM paradigm the repetition of

21.

FALSE MEMORY IN OLDER ADULTS

619

actual list items has been shown to reduce false memory susceptibility in young adults, but this is not always the case in older adults. With a different paradigm, Jacoby (1999) notes that repeated new items are generally rejected by young adults, but in older adults they are more likely to be accepted as old items because of a failure to use recollection to overcome these items' increased familiarity. Collectively, these findings have two implications for the elderly eyewitness. First, these findings suggest that repeated actual event information (the eyewitness analogy to the word lists used in false memory research) may not become more memorable to the older adult, and so regardless of how many times actual event information is encountered, the elderly eyewitness may be unable to use this information to overcome memory suggestibility. Karpel, Hoyer, and Toglia (2001) conducted a study that provides some insight into this issue by examining the effects of repeated exposures to an event on memory suggestibility. Participants watched a simulated theft presented in a series of slides one or two times. Two versions of the theft were created, each with two critical items (i.e., a can of Coke and a Vogue magazine in one version, and a coffee jar and a wrench in the other version); subjects saw one of these two versions during a study episode. Half of the subjects were exposed once to the series of slides, and the other half viewed the series twice (with a 5-minute delay between series). Two different questionnaires were used to assess memory accuracy and memory suggestibility: Questionnaire 1 assessed memory accuracy for presented items, and Questionnaire 2 assessed both memory accuracy for real items and memory suggestibility for critical items. We highlight only those results from the second questionnaire to simplify our discussion. On Questionnaire 2, memory for presented items was higher in young adults than it was in older adults, but only young adults showed an effect of repeated exposure on memory for real items. That is, young adults performed more accurately after two exposures to the events than did the older adults. For critical items, a main effect of age was noted, in which older adults were more susceptible to suggestion than were the young adults, regardless of the number of event exposures. This finding is congruent with our argument that older adults are unable to use repeated information to override the familiarity of suggested information and are more susceptible to memory errors than young adults. This is likely to arise from age differences in processing, where the tendency in older adults to respond on the basis of familiarity and to employ gistbased processing at the expense of verbatim processing creates a situation in older adults where the mere repetition of an event, without highlighting attention to detail, strengthens memory for its thematic content. As such, we would not expect memory for details to increase with repetition, and consequently older adults are more susceptible to misinformation. Future research in this area should be focused on ways to make accurate event details more memorable when that information is repeated (e.g., highlight attention to details or provide more repetition) to see whether older adults are consistently resistant to event exposure effects on memory suggestibility. A second implication to be drawn about the effects of repetition from the false memory research is that new post-event information becomes more memorable because of an age-related failure to use recollection (e.g., recollection of the context in which that misinformation appeared, or a failure to recollect its source) to overcome this information's

620

LAVOlE, MERTZ, RICHMOND

increased familiarity (relative to control conditions). These effects have important implications for the elderly eyewitness, as they suggest that the older adult may be more susceptible to the repeated effects of misinformation than are young adults, in the sense that age differences in susceptibility become greater with repetition. One of the few studies that has examined the effects of repeated exposure to misinformation in older adults was conducted by Mitchell, Johnson, and Mather (2003), who examined suggestibility effects as a consequence of the number of times participants were exposed to misinformation (zero, one, or three exposures) within post-event questions. Results indicated that although older adults were more suggestible overall than young adults, age did not interact with number of repeated exposures (i.e., one versus three). These findings indicate that young and older adults show equivalent effects of repeated suggestion. As very little research has examined the effects of repeated exposure to misinformation in older adults, however, any general conclusions regarding these effects cannot be made until more research is conducted.

CONCLUSIONS

This chapter has highlighted a few selected studies of false memory susceptibility in older adults to draw connections from that research to research on the elderly eyewitness, for which there is a scant literature. The false memory literature points overwhelmingly to an increased susceptibility in older adults to false recognition and false recall of novel information. This is especially the case if that information is related (either semantically or perceptually) to studied information, but the mere repetition of novel information also serves to increase the probability of its being falsely recognized later. These are troubling findings for anyone interested in the accuracy of elderly eyewitnesses, as they suggest that an older adult's testimony about a witnessed event is more likely to be tainted by post-event information than is a young adult's. We have highlighted a few circumstances in which we think this is particularly true, with a discussion of some relevant research from the eyewitness memory literature. To summarize our main points, the false memory research implicates at least two underlying factors in older adults that produce increased susceptibility to memory errors: source-monitoring deficits and encoding/retrieval processing deficits. We believe that these two factors are not mutually exclusive, but rather operate together to make it more difficult for older adults to accurately recollect events. This is not to say, however, that older adults are consistently less accurate in their recollection of events, as they are able, under specific circumstances, to perform as well (or as poorly, as the case may be) as young adults. These circumstances include situations where recollection of detailed event information is not necessary for retrieval, but rather the general gist or theme of that event can drive retrieval. It would be the rare case in legal settings, however, where this mode of retrieval is optimal. Rather, witnesses are more likely to be asked to recount their memory for details of an event. The false memory literature clearly implicates this mode of retrieval as problematic for older adults, not because they fail to encode these details,

21.

FALSE MEMORY IN OLDER ADULTS

621

but rather because their reliance on familiarity and gist-based processing makes them less likely to utilize these details in their recollection of that event. They are then more likely than young adults to be misled about event details and may even be less resistant to post-event misinformation. There remains, however, a gap between the false memory research and the eyewitness memory research on older adults, despite the fact that the false memory research has much to say about the set of circumstances under which the elderly eyewitness's memory may be less accurate. We have offered a few suggestions about some potential research that might serve to bridge this gap, with the hope that this research will be conducted.

LEGAL IMPLICATIONS

The summary we have provided about memory accuracy in the older adult raises legitimate concerns about the accuracy of the elderly eyewitness. However, this review also points out that the elderly eyewitness does not necessarily possess less accurate memories for a witnessed event than does a young adult. There are a number of circumstances and conditions in which young and older adults will perform similarly. These situations are those in which the witness, regardless of age, is asked to testify about general event circumstances. Nonetheless, the older adult will be placed at a distinct disadvantage in the typical courtroom or investigative setting where he or she is asked to describe details about a witnessed event. Because older adults rely more on gist-based processing, especially at retrieval, than do young adults, reported memory for details is often weaker in this age group. As our review demonstrates, the attempt to report details that either (a) may not have been sufficiently encoded or (b) may be difficult to discriminate from other similar event details is likely to produce less reliable and more error-filled recollection of ortestimony about this type of information. This does not make the entirety of an older adult's event recollection or testimony suspect, however; indeed, an older adult's memory for general event information is quite accurate. It is important to note that the type of memory errors we have focused on in our discussion are memory errors for highly plausible, yet not actually encoded, stimulus information. Older adults do not tend to show higher rates of false acceptance of implausible information. For example, Malmstrom and LaVoie (2002), using a picture recognition paradigm, found that the rejection of unstudied probe items consistent with the theme of a previously presented picture array was more difficult in older adults than in young adults. For inconsistent probe items, however, there were no age differences. These findings highlight an important point related to at least one aspect of elderly eyewitness memory. That is, although the older adult may be more susceptible to memory errors, these errors do not represent fabrications or confabulations (i.e., the false recollection of completely implausible event details). As such, the general veracity of an older adult's event recollection or testimony can be trusted. However, to the extent that the older adult is required to testify to event details, particularly those that might be considered peripheral, his or her testimony will be much less reliable and error-prone. It is important

622

LAVOlE, MERTZ, RICHMOND

to weigh these research conclusions in light of the individual testifying, however, in the sense that although older adults as a group are more susceptible to memory errors, there is a tremendous amount of individual variability, such that any single older adult may possess the cognitive competencies of a younger adult and is therefore no less reliable than younger witnesses. The larger issue here is whether a reduced memory for details of a witnessed event is legally relevant, that is, requires a judge to alert a jury to the potential unreliability of an older eyewitness's testimony. The answer to this question is dependent, of course, upon the importance of these details to the overall case, as is probably true of all eyewitness testimony. Caution must be exercised in the acceptance of testimony about fine details, regardless of the witness's confidence in these recollections, because the probability that these recollections contain errors is higher (but not guaranteed) than is the case for recollection of general event actions. The research literature indicates that this type of information is most likely to be error-filled in the older adult. To rule all testimony provided by older adults as unreliable would be unjustified, however, as much of what they do remember about events is accurate. Again, the research literature suggests that investigative questioning at the scene or by attorneys in the courtroom that encourages gist-based or general event recollection will lead to accurate responding by an older adult. The research literature does indicate, however, that questioning that requires verbatim processing by an older adult is most likely to produce error-filled recollections, but further research is needed in field settings to clarify the types of situations in which these recollections are significantly less reliable than younger adults' recollections. Until that research is conducted, older eyewitnesses should not be treated differently from younger witnesses regarding testimony about general crime/event circumstances.

REFERENCES Balota, D. A., Dolan, P. 0., &Duchek, J. M. (2000). Memory changes in healthy older adults. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 395-409). Oxford: Oxford University Press. Bartlett, J. C., Strater, L., & Fulton, A. (1991). False recency and false fame of faces in young adulthood and old age. Memory & Cognition, I9, 177-188. Bornstein, B. H. (1995). Memory processes in elderly eyewitnesses: What we know and what we don't know. Behavioral Sciences and the Law, 13, 349-363. Bornstein, B. H., Witt, C. J., Cherry, K. E., & Greene, E. (2000). The suggestibility of older eyewitnesses. In M. B. Rothman, B. D. Dunlop, & P. Entzel (Eds), Elders, Crime, and the Criminal]ustice System (pp. 149-162). New York: Springer. Bower, G. H., Black, J. B., & Turner, T. J. (1979). Scripts in memory for text. Cognitive Psychology, II, 177-220. Brainerd, C. J., & Reyna, V. F. (1990). Gist is the grist: Fuzzy-trace theory and the new intuitionism. Developmental Review, IO, 3-47. Bransford, J.D., & Franks, J. J. (1971). The abstraction oflinguistic ideas. Cognitive Psychology, 2, 331-350. Brimacombe, C. A. E., Quinton, N., Nance, N., & Garrioch, L. (1997). Is age irrelevant? Perceptions of young and old adult eyewitnesses. Law and Human Behavior, 2I, 619-633.

21.

FALSE MEMORY IN OLDER ADULTS

623

Brown, A. S., Jones, E. M., & Davis, T. L. (1995). Age differences in conversational source monitoring. Psychology and Aging, 10, 111-122. Budson, A. E., Daffner, K. R., Desikan, R., & Schacter, D. L. (2000). When false recognition is unopposed by true recognition: Gist-based memory distortion in Alzheimer's disease. Neuropsychology, 14, 277-287. Cohen, G., & Faulkner, D. (1989). Age differences in source forgetting: Effects on reality monitoring and on eyewitness testimony. Psychology and Aging, 4, 10-17. Coxon, P., & Valentine, T. (1997). The effects of the age of eyewitnesses on the accuracy and suggestibility of their testimony. Applied Cognitive Psychology, 11, 415-430. Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58, 17-22. Ghetti, S., Qin, J., & Goodman, G. S. (2002). False memories in children and adults: Age, distinctiveness, and subjective experience. Developmental Psychology, 38, 705-718. Haber, R.N., & Haber, L. (2000). Experiencing, remembering, and reporting events. Psychology, Public Policy, and Law, 6, 1057-1097. Hedden, T., & Park, D. C. (2003). Contributions of source and inhibitory mechanisms to agerelated retroactive interference in verbal working memory. Journal of Experimental Psychology: General, 132, 93-112. Henkel, L.A., Johnson, M. K., & De Leonardis, D. M. (1998). Aging and source monitoring: Cognitive processes and neuropsychological correlates. Journal of Experimental Psychology: General, 127, 251-268. Israel, L., & Schacter, D. L. (1997). Pictorial encoding reduces false recognition of semantic associates. Psychonomic Bulletin & Review, 4, 577-581. Jacoby, L. L. (1999). Ironic effects of repetition: Measuring age-related differences in memory. Journal of Experimental Psychology: Learning, Memory, & Cognition, 25, 3-22. Jennings, J. M., &Jacoby, L. L. (1997). An opposition procedure for detecting age-related deficits in recollection: Telling effects of repetition. Psychology and Aging, 12, 352-361. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin,

114,3-28.

Karpel, M. E., Hoyer, W. J., & Toglia, M.P. (2001). Accuracy and qualities of real and suggested memories: Nonspecific age differences. Journal of Gerontology, 56B, P103-P110. Kensinger, E. A., & Schacter, D. L. (1999). When true memories suppress false memories: Effects of aging. Cognitive Neuropsychology, 16, 399--415. Koutstaal, W. (2003). Older adults encode-but do not always use-perceptual details: Intentional versus unintentional effects of detail on memory judgments. Psychological Science, 14, 189-193. Koutstaal, W., Reddy, C., Jackson, E. M., Prince, S., Cendan, D. L., & Schacter, D. L. (2003). False recognition of abstract versus common objects in older and younger adults: Testing the semantic categorization account. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29,499-510. Koutstaal, W., & Schacter, D. L. (1997). Gist-based false recognition of pictures in older and younger adults. Journal of Memory and Language, 37, 555-583. LaVoie, D. J., & Faulkner, K. (2000). Age differences in false recognition using a forced choice paradigm. Experimental Aging Research, 26, 367-381. LaVoie, D. J., & Willoughby, L. (1998, April). Priming illusory memories in young and older adults. Paper presented at the Cognitive Aging Conference, Atlanta, GA. Light, L. L., Prull, M. W., LaVoie, D. J., & Healy, M. R. (2000). Dual-process theories of memory in old age. InT. J. Perfect & E. A. Maylor (Eds.), Models of cognitive aging (pp. 238-300). New York: Oxford University Press. Loftus, E. F. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7, 560-572.

624

LAVOlE, MERTZ, RICHMOND

Loftus, E. F. (1979). Eyewitness testimony. Cambridge, MA: Harvard University Press. Loftus, E. F., Levidow, B., & Duensing, S. (1992). Who remembers best? Individual differences in memory for events that occurred in a science museum. Applied Cognitive Psychology, 6, 93-107. Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589. Malmstrom, T., & LaVoie, D. J. (2002). Age differences in inhibition of schema-activated distractors. Experimental Aging Research, 28, 281-298. Mandler, G. (1980). Recognizing: The judgment of previous occurrence. Psychological Review, 87, 252-271. McCabe, D. P., & Smith, A. D. (2002). The effect of warnings on false memories in young and older adults. Memory & Cognition, 30, 1065-1077. Mitchell, K. J., Johnson, M. K., & Mather, M. (2003). Source monitoring and suggestibility to misinformation: Adult age-related differences. Applied Cognitive Psychology, 17, 107-119. Mitchell, K. J., & Zaragoza, M. S. (1996). Repeated exposure to suggestion and false memory: The role of contextual variability. Journal of Memory and Language, 35, 246-260. Norman, K. A., & Schacter, D. L. (1997). False recognition in younger and older adults: Exploring the characteristics of illusory memories. Memory & Cognition, 25, 838-848. Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. Roediger, H. L., McDermott, K. B., & Robinson, K. J. (1998). The role of associative processes in creating false memories. In M. A. Conway, S. E. Gathercole, & C. Cornoldi (Eds.), Theories of memory (Vol. II, pp. 187-245). East Sussex, England: Psychology Press. Schacter, D. L., Curran, T., Galluccio, L., Milberg, W. P., & Bates, A. F. (1996). False recognition and the right frontal lobe: A case study. Neuropsychologia, 34, 793-808. Schacter, D. L., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger and older adults: The distinctiveness heuristic. Journal of Memory and Language, 40, 1-24. Schacter, D. L., Koutstaal, W., Johnson, M. K., Gross, M. S., & Angell, K. E. (1997). False recollection induced by photographs: A comparison of older and younger adults. Psychology and Aging, 12, 203-215. Schacter, D. L., Norman, K. A., & Koutstaal, W. (1998). The cognitive neuroscience of constructive memory. Annual Review of Psychology, 49, 289-318. Schacter, D. L., Verfaellie, M., Anes, M.D., & Racine, C. (1998). When true recognition suppresses false recognition: Evidence from amnesic patients. Journal of Cognitive Neuroscience, 10, 668-679. Smith, A. D. (1975). Partial learning and recognition memory in the aged. The International Journal of Aging & Human Development, 6, 359-365. Smith, S.M., Ward, T. B., Tindell, D. R., Sifonis, C. M., & Wilkenfeld, M. J. (2000). Category structure and created memories. Memory & Cognition, 28, 386-395. Spencer, W. D., & Raz, N. (1995). Differential effects of aging on memory for content and context: A meta-analysis. Psychology and Aging, 10, 527-539. Tun, P. A., Wingfield, A., Rosen, M. J., & Blanchard, L. (1998). Response latencies for false memories: Gist-based process in normal aging. Psychology and Aging, 13, 230-241. Underwood, B. ]. (1965). False recognition produced by implicit verbal responses. Journal of Experimental Psychology, 70, 122-129. U.S. Census Bureau. (2004). U. S. interim projections by age, sex, race, and Hispanic origin. Retrieved March 18, 2004 from U.S. Census Bureau, Population Division Web site, http://www.census .gov/ipc/www/usinterimproj/

21.

FALSE MEMORY IN OLDER ADULTS

625

Warren, A. R., & Lane, P. (1995). Effects of timing and type of questioning on eyewitness accuracy and suggestibility. In M.S. Zaragoza, J. R., Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds), Memory and testimony in the child witness (pp. 44-60). Thousand Oaks, CA: Sage. Wingfield, A., Tun, P. A., & Rosen, M. J. (1995). Age differences in veridical and reconstructive recall of syntactically and randomly segmented speech. journal of Gerontology: Psychological Sciences, 50, P257-P266. Zaragoza, M. S., & Mitchell, K. J. (1996). Repeated exposure to suggestion and the creation of false memories. Psychological Science, 7, 294-300.

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

22 Eyewitness Memory in Older Adults Chris J. A. Moulin University of Leeds, United Kingdom

Rebecca G. Thompson Research Institute for the Care of the Elderly, St. Martin's Hospital, United Kingdom

Dan B. Wright University of Sussex, United Kingdom

Martin A. Conway University of Leeds, United Kingdom

Stereotypes of aging are generally negative (for example, that older people are slow and forgetful). This contrasts with stereotypes of the personality traits of older people, which are typically positive (for example, honest, trustworthy, kind, and reliable). Both of these stereotypes extend directly to the judicial system and have a detrimental effect on the perceived credibility of the older adult as an eyewitness. For instance, Ross, Dunning, Toglia, and Ceci (1990) reported that older adults' testimonies were rated more negatively than those of young eyewitnesses on accuracy and suggestibility. However, the older adults were rated equal to the young adults in terms of the honesty and integrity of their testimonies. Similarly, Brimacombe, Quinton, Nance, and Garrioch (1997) reported that older adult eyewitnesses are perceived as less accurate, confident, and competent but more honest than young adults. More recently, Kwong See, Hoffman, and Wood (2001) reported that young adults rated an 83-year-old female witness to be less competent but more honest than a 23-year-old female witness. Thus there is a disparity between the perceptions of integrity of the older adult eyewitness and the accuracy of the very same eyewitness. The fact that older adults are perceived to be more honest could possibly mean that they are better eyewitnesses, although this stereotype is likely to be misleading (and is actually somewhat neglected in the research literature). The fact that they are perceived 627

628

MOULIN ET AL.

to be less accurate suggests that they are poorer eyewitnesses. One of the goals of this chapter is to evaluate this stereotype of older adults as less accurate eyewitnesses. Such stereotypical views of aging extend to expert opinion on older adults as eyewitnesses, as reported by Kassin, Tubb, Hosch and Memon (2001), whose research concentrated on the Frye criterion for admissibility of expert testimony. In a survey of 64 experts in eyewitness testimony, 51% tended to agree (at the least) with the statement, "Elderly witnesses are less accurate than are younger adults." Only 23% of the full sample of experts indicated that this statement had no support or that the literature was inconclusive; 50% of experts said that this observation was reliable enough to present in courtroom testimony, but only 77% reported that their opinion on the reliability of this statement was based on their reading of published, peer-reviewed, scientific research. From this survey of experts, two main facts can be gleaned. One, older adults, for the most part, are seen as less proficient eyewitnesses. Two, this fact is not as reliably demonstrated as other eyewitness phenomena. In comparison with the above, 70% of experts said that the statement "Young children are less accurate as witnesses than are adults" was reliable, and 91% said that this was based on their reading of the scientific literature. This chapter considers the evidence for older adults being unreliable eyewitnesses, based on the notion of cognitive decline. A theoretical explanation of poor memory in older adults is presented, with the aim of indicating why it is reasonable to assume, in general, that older adults are less accurate as eyewitnesses. The main thesis of this chapter is that healthy aging is associated with particular memory deficits and that these memory difficulties should have an impact on eyewitness testimony. Not all memory systems decline with age, and not all memory processes are worse in older adults (for a review of memory in healthy older adults, see Balota, Dolan, and Duchek, 2000). For an example of this line of argument, on the basis of laboratory-based studies of learning of verbal materials, it has often been demonstrated that older adults forget information more quickly than younger adults (e.g., Wheeler, 2000). Eyewitness research has also considered this increased forgetting in older adults. Memon, Bartlett, Rose, and Gray (2003) showed that in a task where participants viewed a videotape of a crime and then had to detect the perpetrator in a lineup, there were no age differences after a short test delay interval (35 minutes), but that after a considerably longer interval (1 week), reliable age differences appeared, such that the older adults were less accurate. This example shows a correspondence between the mainstream literature and findings from studies that directly evaluate eyewitness testimony in older adults: in this case, the theory that older adults forget information more quickly when initially matched to younger adults is demonstrated with eyewitness materials and in a typical eyewitness task. Where direct evidence from studies of eyewitness ability in older adults is lacking, however, one should be able to make inferences from the mainstream cognitive aging and memory literature. We might have inferred how older adults perform over long delays with eyewitness materials, based on our reading of the literature on laboratory tasks on word list materials. It is beyond the scope of this chapter to give a complete overview of cognitive aging theory, and so we focus on the idea that older adults' memory function is characterized

22.

EYEWITNESS MEMORY IN OLDER ADULTS

629

by a series of deficits common to frontal lobe damage and that this affects memory in a particular manner. It is argued that this idea is a useful one for describing and categorizing the effect of age on eyewitness testimony, rather than a conceptual or theoretical tool for understanding cognitive aging. There are a number of other issues that are beyond the discussion here. Most notably, peripheral physical deficits in older adults can lead to perceptual and sensory difficulties, with a clear impact on eyewitness ability. However, as in the research literature on cognitive function on laboratory tasks, it is assumed in this review that the differences between young and older adults reported here are due to cognitive differences, after such sensory and perceptual deficits have been taken into account. For a study of the interaction between sensory and cognitive functions, see Lindenberger and Baltes (1994).

OLDER ADULTS' MEMORY AS FRONTAL LOBE DYSFUNCTION

Contemporary aging research tends to take a neuropsychological perspective, trying to identify the underlying neural mechanisms behind memory decline in older adults. The driving force behind this approach has largely been the similarity between older adults' cognitive difficulties and the pattern of impairment that occurs after acquired head injury to the frontal part of the brain. As demonstrated below, the frontal lobes are thought to be responsible for several functions crucial to eyewitness testimony: ascribing confidence judgments, judging temporal order, reporting the source of information, and so on. Therefore, memory dysfunction is not merely a stereotype of the aging process. Countless studies (for reviews see Balota, Dolan, & Duchek, 2000; Kester, Benjamin, Castel, & Craik, 2002) have demonstrated that healthy aging (i.e., "normal" aging in the absence of dementia, stroke, or other neurological impairment or degeneration) is associated with a decline in memory function and that this gives rise to different patterns of brain activation in young and old people (e.g., Cabeza, 2002). Most convincingly, cognitive decline in older adults is related to known changes in the brain, in which older adults show a somewhat inevitable loss of brain cells. That is, memory loss in older adults is probably not due merely to societal norms (e.g., the cliched view that memory fails with age, or more technically through diminished self-efficacy), psychological processes secondary to aging (e.g., depression, grandparenting), or physical difficulties that increase with the aging process (e.g., chronic pain, general poor health, poor hearing), but is based on underlying neuropsychological changes. Such a model of cognitive decline based on age-related changes in the brain is a useful context in which to discuss memory dysfunction, since armed with the knowledge of which brain locations are most affected by age, it is possible to make predictions about which particular aspects of memory function are most affected by the aging process. The most often cited model of memory change with healthy age is that of memory aging as frontal lobe dysfunction. Although a useful description of memory function in older adults, this viewpoint receives equivocal evidence in the literature (for critical reviews see Greenwood, 2000, and Perfect, 1997).

630

MOULIN ET AL.

This model of memory dysfunction rests on the observations that the most marked atrophy of cells in the aging brain is in the frontal cortex (e.g., Ivy, MacLeod, Petit, & Markus, 1992), that older adults show deficits on tasks related to frontal function in batteries of cognitive tasks, that these correlate with memory performance (e.g., Parkin & Walter, 1991), and that there is a general similarity between frontal patients and older adults on the patterns of performance across experimental tasks (e.g., Perfect, 1997). Where the model is less convincing is that researchers seldom test alternatives of this hypothesis-in that batteries of cognitive tests rarely include nonfrontal tasks (Greenwood, 2000; Parkin & Java, 2000; Perfect, 1994). Furthermore, studies of even large groups of frontal patients suggest that there is considerable heterogeneity in frontal damage populations, with measures of frontal function only having weak intercorrelations (Duncan, Johnson, Swales, & Freer, 1997). Most strikingly, even on different versions of one popular frontal test, the Stroop task, there is very little relation between one version and another for older adults (Shilling, Chetwynd, & Rabbitt, 2002). Thus defining what frontal function is, and assuming it to be uniform across older adults, seems to be a more difficult task. However, recent advances in neuroimaging have offered more support for the frontal model. For instance, in a review of the event-related potential literature, Friedman (2000) has shown that memory tasks allied to electrical activity in the prefrontal cortex involve the areas where there are large age-related differences-with a lack of brain activity in the old and a corresponding difficulty with searching memory and providing rich contextual information. Other work indicates that older adults need to recruit more of the frontal lobes than younger people in order to carry out the same memory task. A common finding is that whereas young people recruit a small, unilateral region of the frontal lobes, older adults show more extensive activation across both lobes (e.g., Langenecker & Nielson, 2003). Finally, Braver and Barch (2002) present a model of cognitive decline in aging that implicates the frontal lobes, drawing on the theory that levels of the neurotransmitter dopamine in the frontal lobes are particularly low in older adults. Dopamine is a brain chemical that appears to be particularly important in arousal and attention, although it is most commonly thought of as crucial to the control of motor function. Braver and Barch suggest that neuroimaging indicates that the frontal lobes are responsible for inhibition, attention, and memory function, and that in particular this dysfunction in older adults is related to a low level of dopamine. However, most of the research focus on the frontal lobes has arisen because of striking similarities between the types of memory tasks that frontal patients have difficulty with and the types of task that yield the largest differences between groups of young and old participants (see Perfect, 1997). Below several prominent themes in the aging and memory literature are discussed, all focusing on the notion that memory performance is similar in older adults and people with frontal lobe damage. For reviews of the effects of frontal lobe damage on cognition see Stuss and Levine (2002).

The Frontal Lobes: Impacts on Memory Here several memory and aging phenomena are addressed: that recall is more impaired than recognition, deficits in source memory, deficits in metamemory, and deficits with

22.

EYEWITNESS MEMORY IN OLDER ADULTS

631

inhibitory control of memory. Interestingly, these frontal~ type problems of aging also re~ fleet most of the theoretical focus of studies of age and eyewitness ability. A number of other frontal deficits common to aging are not discussed here (i.e., difficulty with mem~ ory for spatial information [Foisy, 1994]; difficulty with memory for temporal informa~ tion [Kausler, Salthouse, & Saults, 1988]), but it is highly probable that these difficulties would also affect eyewitness ability.

Recall More Impaired Than Recognition Numerous empirical studies have indicated that whereas older adults' recall of informa~ tion may be impaired, recognition performance, even for the same materials, tends to be preserved (e.g., Craik &Jennings, 1992). This has clear applications to eyewitness tes~ timony: it is quite possible that older adults will be unable to recall the events of a scene, but will be more likely to be able to correctly recognize aspects of the scene. Of course, with recognition being an "easier" task than recall, this pattern would be true for younger adults also-but the effect is magnified in older adults. Theoretical research in this area has focused on this difficulty issue. Critically, Craik and McDowd (1987) demonstrated that by making recognition extremely difficult by dividing attention during test (asking participants to perform a visual recognition test while being administered a recognition test on headphones), young and old adults still showed no differences in recognition performance. Differences in recall were retained, even though the recall test was much easier (and yielded higher levels of performance) than recognition. Thus, on quantity of information retrieved, older adults and younger controls seem equal on recognition tasks. However, there is substantial evidence to suggest that the quality of recognition performance in older adults is impaired. Tulving (1983) pioneered a memory test procedure in which participants are required to report their subjective ex~ perience of remembering. In this task, when responding yes or no to whether they have encountered an item before, participants are also asked to give an evaluation of the basis of this answer. Typically, they are required to evaluate whether they "remember" an event or find it "familiar" (sometimes referred to as "just knowing"). Remembering an event includes a sense of "pastness" and an ability to retrieve contextual information about encountering the stimulus before, such as remembering what you were thinking and feeling when you bumped into a man in a local shop a few days ago. Finding an item familiar is a judgment based on a lack of such rich evocative information, such as the type of memory we have for facts. These subjective appraisals of memory quality are dif~ ferent from other metacognitive judgments discussed below, such as confidence. People can judge an item as merely being familiar, but still be 100% confident that they have en~ countered the word before (Gardiner & Richardson~Klavehn, 2000). These two dif er~ ent judgments, of remembering and familiarity, are proposed to map onto two different memory processes-recollection (a conscious retrieval process) and familiarity (a more automatic, unconscious component). There is considerable evidence that these two forms of memory are anatomically and functionally distinct (see Yonelinas, 2002, for an over~ view). As developed below, aging is thought to disrupt recollection but not familiarity (e.g., Davidson & Glisky, 2002; Yonelinas, 2002), and this has several effects on memory that would be of relevance in eyewitness situations.

632

MOULIN ET AL.

Parkin and Walter (1992) carried out the recollective experience procedure on older adults with standard word~ task experiments. They demonstrated that older adults report significantly fewer remember responses, indicating that their memory performance is characterized by a lack of contextual information and an over~reliance on the mere familiarity of a previously encountered stimulus. This finding was replicated by Perfect, Brown (1995). In a follow~up, Perfect and Dasgupta (1995) ex~ Williams, and Anderton~ amined justifications of recollective experience in older adults. They asked participants to "think aloud" during study, examining what introspections and observations partici~ pants made at study. At test, participants made the usual recollective judgments, as above, and were asked to justify their response. The younger adults showed clear consis~ tency in their justifications of their remember responses and their reports of thoughts during study, reflecting an ability to consider the contextual information and subjective state of an earlier study episode. In comparison, when older adults made remember judg~ ments, Perfect et al. found that there was an inconsistency between the justification given at retrieval and what they reported at study. They suggested that this indicated a tendency to produce post~ hoc rationalizations of their memory experience in the lack of a rich memory of the study episode. This suggests that older adults' memory may be un~ reliable in this regard: they may reconstruct possibly incorrect contextual information to with justify the sensation of remembering. Such an idea would be worthy of follow~up word experiments. the use of eyewitness tasks, rather than laboratory~based

SOURCE MEMORY

A qualitative assessment of recognition memory seems to suggest a deficit in the mem~ ory for contextual information. Similarly, older adults, like frontal head injury patients, have shown deficits in source memory. Source memory can be thought of as the ability to retrieve the context or provider of a certain piece of information, usually assessed only for information that a person has correctly retrieved. For example, given that you re~ member that a man has been seen loitering around the loading bay, can you remember who told you this information? The evidence that older adults demonstrate difficulties with this kind of memory mostly comes from tests of general knowledge. For instance, in a study by Schacter, Kasniak, Kihlstrom, and Valdiserri (1991), young and old partici~ pants were presented with novel and trivial information by either of two people. At test, participants were asked who had told them the fact. Even though there were no group differences in memory for the facts, older adults were significantly worse at judging who had presented them the information. This sort of difficulty is a robust finding of cogni~ tive studies of aging. Such a problem of source amnesia is thought to be behind the false fame effect found in older adults (e.g., Dywan &Jacoby, 1990)-this is where older adults but highly familiar stimulus due to show a tendency to attribute fame to a non~famous presentation in an experimental setting. Such a deficit in source is likely to be critical for performance in eyewitness situa~ tions, where detailed information about the contextual details in which a person en~ countered information or encoded an event is likely to be important. Moreover, the fact

22.

EYEWITNESS MEMORY IN OLDER ADULTS

633

that older adults may be able to correctly provide some correct information about an event, such as the color of the getaway car, does not necessarily point to the fact that they will be able to correctly remember where they encountered it. Indeed, the literature on eyewitness testimony in older adults supports this view. Finally, source memory deficits can be implicated in unconscious transference, where people falsely identify a familiar but innocent bystander when asked to pick out a perpetrator from a lineup (e.g., Perfect & Harris, 2003). The implication here is that people misremember the source or identity of a face: they remember a bystander as being the perpetrator. The sections on recall and recognition and source memory point to memory reports in older adults that lack detail. Such findings have led many researchers to suggest that older adults' performance is characterized by an over-reliance on gist, that is, general information without details (e.g., Tun, Wingfield, Rosen, & Blanchard, 1998). This theme has been picked up by eyewitness researchers. Several studies conclude that older adults are less accurate in their recall of events and provide fewer details in their recall of witnessed events (e.g., Brimacombe et al., 1997; List, 1986), with the interpretation that older adults encode more of the general gist information rather than distinctive information. Eyewitness memory is known to be extremely vulnerable to disruption when witnesses are exposed to misleading post-event information. Cohen and Faulkner (1989) suggest that an over-reliance on gist contributes to the finding that older adults are more misled by post-event false information than young subjects. A further demonstration of this vulnerability is through the exposure to mugshots of innocent people-the mugshot exposure effect. Eyewitnesses to both real and experimental situations are more likely to choose an innocent face at a subsequent lineup if it has been presented previously in a mugshot display (Lindsay & Wells, 1985; Wells & Seelau, 1995). Memon, Hope, Bartlett, and Bull (2002) examined older adult participants who had witnessed an event with a specific target face. This was followed by a mugshot identification (without the target but with one critical foil face) and then by a target absent lineup (containing the foil). Older adults were more likely to select the critical foil, based on the familiarity of the face. Because familiarity judgments are thought to be based on less rich memory signals, this creates an increased margin for error. This empirical finding has a direct correspondence with the theory that older adults lack recollection, are over-reliant on gist memory (as above), and fail to establish the "source" of a facecriminal or innocent bystander.

METAMEMORY

Metamemory considers people's beliefs about their memory function and their ability to control and monitor their memory processes (e.g., Nelson & Narens, 1994). There is equivocal data as to whether older adults lack the ability to control and monitor their memory. Typically, studies of metacognition use people's predictions of performance, compared with their actual performance as a gauge of how well they can monitor their memory system. Studies of metacognition focus on the magnitude of people's judgments compared with their performance (i.e., calibration; someone who predicts they

634

MOULIN ET AL.

will remember 10 items should actually remember 10) or on the relative accuracy of their judgments. In relative accuracy, the magnitude of the prediction is not of interest; the focus is on merely whether an item with a higher prediction is more likely to be remembered than an item with a lower prediction. A robust finding is that older adults tend to overestimate their memory performance, and they are not as well calibrated as younger people (e.g. Connor, Dunlosky, & Hertzog, 1997). Thus, the literature suggests that older adults have a deficit in metamemory, inasmuch as they cannot accurately rate how good their memory is (e.g., Light, 1991). In general, the aging process appears to produce predictions of performance that are overconfident-suggesting that older adults are not aware of their failing memory. Of course, such overconfidence, where supported by the literature, would be of importance to an assessment of eyewitness ability. However, if one focuses on the relative accuracy of metamemory, it appears that older adults evaluate their memory function as well as young controls (Connor et al., 1997). Older adults can judge as accurately as young people (and at above-chance levels) which of a group of items are more likely to be remembered, presumably because part of having a well-calibrated memory performance is that memory performance is in keeping with expectations. For instance, if you take into account memory performance, age differences in predictive accuracy disappear (Perfect & Stollery, 1993). This suggests that a deficit in metamemory may actually only be another manifestation of the problem with forgetting. We suggest that, on the whole, as long as relative accuracy is the focus of study, and we do not expect people to be able to have perfectly calibrated evaluations of their own ability, then older adults are as metacognitively competent as younger adults. Even older adults with Alzheimer's disease make appropriate evaluations of their performance if an approach that emphasises sensitivity to memory impairment is used (Moulin, Perfect, & Jones, 2000). The key point for eyewitness testimony is that this approach suggests that during questioning, older adults will be able to reliably intuit which of two facts about an event is the more difficult for them to remember, or which one they are less sure of. But this may not actually relate to their actual accuracy. They may know that one thing is easier to remember than another (e.g., the color of a vehicle is better remembered than the license plate), but in fact the memories of both may be unreliable or the facts forgotten. It is clear that older adults do have some specific difficulties with metamemory, such as a lack of a relationship between predictions of future performance and time spent studying materials (Dunlosky & Connor, 1997), but it is unclear how such specific deficits might relate to eyewitness testimony, especially when so many studies indicate that metacognitive evaluations in the courtroom are largely unreliable in comparison with evaluations of general knowledge retrieval (e.g., Perfect & Hollins, 1999). In addition, research has shown that although older adults are very good at making evaluations of performance on tests of well-learned general knowledge materials, they are much less accurate at making introspections about performance on recently learned information (Souchay, Isingrini, & Espagnet, 2000) more akin to the types of knowledge you might be expected to probe in an eyewitness situation.

22.

EYEWITNESS MEMORY IN OLDER ADULTS

635

One difficulty with the metacognition literature is that whereas most laboratory memory studies indicate overconfidence in older adults, in the context of eyewitness testimony, it has been shown that older adults possess lower confidence in their abilities to recall information accurately. This lower confidence may directly relate to the negative perception held about the older adult eyewitness. For example, Brimacombe et al. (1997) reported that older adults tend to use less confident language in their recall of events (e.g., I think or I am not sure). In contrast, Searcy et al. (2001) reported that older adults who have high (selfrated) memory efficacy and who were able to recall high levels of information about the initial event were also significantly more likely to produce false choices including identification of an innocent foil (when the face of the perpetrator was present) and were less likely to reject foils in a target-absent situation. Similarly, in Cohen and Faulkner's (1989) study of post-event suggestibility in older adults, they found that when older adults made errors, they were more confident in the accuracy of their erroneous responses than were younger adults. The overall pattern of findings in this area is represented well in a study by Scogin, Calhoon, and Derrico (1994). Three groups of participants of approximately 21, 67, and 80 years were showed a videotape of a crime and were subsequently asked to complete a multiple-choice questionnaire about the event. Performance changed with age in the predictable manner, with impairments in accuracy for the older groups. However, in no group were there significant correlations between their performance and their metamemory evaluation of it.

INHIBITION OF MEMORY

One of the more recent theories of age-related cognitive decline considers that cognitive difficulties are due to problems with adequately inhibiting irrelevant thoughts or controlling unnecessary processes (e.g., Hasher & Zacks, 1988). In its most extreme form (e.g., Hasher & Zacks, 1988), this theory suggests that all age-related cognitive difficulties (e.g., attention, working memory, speed of processing) are secondary to inhibitory deficits. In this guise, the theory competes with the frontal hypothesis of aging, as a global account of cognitive difficulties in aging, although the two theories are not incompatible, inasmuch as frontal patients are known to have particular difficulties with inhibition (e.g., Stuss & Levine, 2002). Here we discuss it in a more limited capacity, with the idea that inhibition of memory is frontal, and frontal patients have difficulty inhibiting items in memory, sometimes called intentional forgetting (e.g., Conway, Harries, Noyes, Racsma'ny, & Frankish, 2000). Inhibition has recently become of interest to eyewitness testimony research (e.g., Macleod, 2002; Wright, Loftus, & Hall, 2001). In particular, memory theorists have identified that the process of repeatedly retrieving one piece of information leads to forgetting of related material (retrieval-induced forgetting; e.g., Perfect, Moulin, Conway, & Perry, 2002). Similarly, people have been shown to be able to wilfully apply similar

636

MOULIN ET AL.

mechanisms to prevent unwanted, irrelevant material from entering memory (directed forgetting, e.g., Bjork & Bjork, 1996; and think/no think, Anderson & Green, 2001). Several programs of research are indicative of inhibitory dysfunction in older adults (e.g., Hasher, Quig, & May, 1997; Zacks, Radvansky, & Hasher, 1996). That is, older adults have difficulties inhibiting information from memory-they remember something that they are told to forget. One often cited example used by researchers to explain the applicability of an inhibitory deficit in older adults is itself forensic: "it is possible that older adults' difficulty in inhibiting the continued processing of information designated as irrelevant or wrong could have an impact on their ability as jurors to comply with a judge's instructions to ignore testimony that has been stricken from the record" (Zacks et al., 1996, p. 155). Most of the evidence for inhibitory difficulties comes from tasks in which participants are asked to complete the last word of high-doze sentences (e.g., Hasher et al.). For example, participants are given the sentence: Before you go to bed, tum off the , with a very high likelihood that they will finish the sentence with light. The experimenter then provides a solution that has to be remembered, which either confirms (instructed to remember the word light) or disconfirms (instructed to remember another word, e.g., radio) the participant's response. The critical condition is the one in which the experimenter's answer disconfirms the high-doze solution that the participant has given. On disconfirm trials, the task is to remember the experimenter's solution and inhibit or forget the first answer. At test, participants are given a different set of sentences to complete. Proficient inhibition is shown by completion of sentences with words other than those seen in the previous stage. Deficient inhibition is shown by participants producing responses that were originally disconfirmed. The standard finding is that older adults show equal recall for confirmed and disconfirmed items, whereas the young recall more of the confirm items than the disconfirm items. Another method that has been used to assess the inhibitory function of older adults is the well-established directed forgetting paradigm (e.g., Bjork & Bjork, 1996). In this paradigm participants study items for a subsequent memory test, but between studying the items and the final test they are told to forget a subset of those items. With one version of the directed forgetting task, Zacks, Radvansky, and Hasher (1996) showed that older adults were less likely than young controls to inhibit to-be-forgotten material. As a result, they were more likely to fail to remember the to-be-remembered material, and more likely to make intrusion errors from the to-be-forgotten list. Such inhibitory errors have clear implications for eyewitness testimony (although no direct test of inhibition in older adults for eyewitness materials has been carried out). The main point here is that memory errors due to age have largely been characterized as errors of omission, but clearly, the failure to inhibit items in memory could lead to errors of commission-a type of memory error that, until recently, has been overlooked in the aging literature. The failure to inhibit distracter items on a recall test will lead to false positives or intrusion errors, but in a lineup, this type of failure may lead to the wrong person being convicted. Indeed, there is evidence that older adults are more susceptible to this kind of error. Searcy, Bartlett, Memon, and Swanson (2001) report that older

22. EYEWITNESS MEMORY IN OLDER ADULTS

637

adults make more false identifications (of innocent individuals) than young adults when the perpetrator is absent from the lineup. The higher rate of false positives in lineup identification parades has been reported elsewhere (Memon, Hope, Bartlett, & Bull, 2002; Searcy, Bartlett, & Memon, 1999, 2000). The conclusions from these studies are that older adults make lineup choices based on familiarity judgments rather than explicit recognition judgments, but equally this deficit could be conceptualized as a failure to inhibit highly familiar but irrelevant material at retrieval. This idea ties in with the distinction between recollection and familiarity (above): older adults remember events on the basis of automatic processes, and the mere familiarity with an item, and lack the conscious retrieval or recollection necessary to filter out erroneous memories, or irrelevant thoughts. Inhibitory errors will lead to other types of eyewitness difficulties. Older adults have been shown to be unable to wilfully forget irrelevant information, suggesting that when trying to remember the events of a scene, their memory for important details may be hampered by the retrieval of irrelevant information. Moreover, if older adults later learn that a fact that they learned was incorrect (e.g., that the car was a Volkswagen and not a Ford), they will have difficulty in overwriting this previously believed fact. Such inhibitory errors could be behind the observation that older adults are more susceptible to misleading information than younger adults (Bomstein, 1995). It is not that older adults are more gullible than their young counterparts, but that they cannot help but act on their memory and previous experience.

COGNITIVE IMPAIRMENT AND EYEWITNESS TESTIMONY

From the above, it is clear that older adults have detectable cognitive memory deficits, and it is reasonable to assume that these laboratory-based findings will extend to memory performance in the context of eyewitness testimony. However, it is also worth noting that the memory performance of older adults is susceptible to large inter- and intraindividual variability, and performance in this group is considerably more heterogeneous than that of younger adults (MacDonald, Hulsch, & Dixon, 2003). Moreover, there is some emergent evidence that older adults' memory problems are worse on formal laboratory tests of memory, with memory performance in real life not so impaired (e.g., Henry, MacLeod, Phillips, & Crawford, 2004). Additionally, a number of studies from different laboratories have indicated that the notion of cognitive difficulties in aging causing eyewitness errors is simplistic. In particular, studies indicate that older adults are influenced by the type of material that they are required to memorize. It has long been known that there are clear effects of race in face recognition, with people being better at recognizing faces from their own race (e.g., Wright, Boyd, & Tredoux, 2001). Similarly, it has been established that the same effects occur with age. Most notably, Wright and Stroud (2002) demonstrated that in target-present lineup participants perform significantly better at identifying the culprit when the targets and

638

MOULIN ET AL.

foils are from their own age group. In target-absent lineups, there are no age differences in performance, regardless of the ages of the foils. This is evidence that older adults may only appear to be less accurate eyewitnesses because of the targets and foils with which they are working. Interestingly, Wright and Stroud's study's "older" adults ranged from 35 to 55 years old-demonstrating clear age differences at an age when most researchers working on aging would probably not expect too great a difference between young and old subjects on memory tasks. Similarly, Perfect and Harris (2003) examined the innocent bystander effect in aging, using either young (19-21 years old) or old (68-74 years old) faces as stimuli. Groups of old (mean age approximately 67) and young (mean age approximately 22) participants saw photographs of target faces (perpetrators), then a series of photographs of nontarget innocent bystanders. A week later, participants were given lineups composed of bystanders and were asked if any of the target faces were present. When younger faces were used, older adults showed higher levels of unconscious transference errors, whereas when old faces were used as stimuli, there were no group differences in misidentification of a bystander. Despite clear theoretical motivations to believe that there should be deficits in eyewitness memory in older adults, these studies call into question the reliability of other findings that compare young and old, but only present younger stimuli. At the very least, these studies suggest that to some extent age differences in eyewitness testimony may be due to social bias, rather than cognitive dysfunction alone. Given that the laboratory-based studies are reliable and extend to day-to-day memory function, what effects might they have on eyewitness testimony? To address this question and by way of summary, a number of different eyewitness phenomena are presented in Table 22.1. Kassin et al. (2001) offer a framework by which to consider the hypothetical effects of age-related memory impairment on eyewitness memory. They suggest that 16 eyewitness phenomena are reliable enough to present in courtroom testimony. The effects of aging on these phenomena are reviewed, with the exception of hypnotic suggestibility, alcoholic intoxication, and child suggestibility, which are not within the scope of this chapter. Table 22.1 summarizes the eyewitness phenomena in question, and whether on the basis of empirical studies of eyewitness testimony in aging or theoretical underpinnings of memory and aging, it is reasonable to suggest that older eyewitnesses may differ from their younger counterparts.

HOW TO MAXIMIZE THE ACCURACY OF OLDER ADULT EYEWITNESS TESTIMONIES

The studies reviewed above do not paint a very positive portrait of the older adult as a competent eyewitness. The evidence suggests that older adults are less accurate in their recall of events, recall fewer details of events, have a greater tendency to include misleading information into their recall of events, and make choices and identifications based on familiarity rather than recollection judgments. Bearing in mind the high value that is placed upon eyewitness reports in court cases, it is important to establish situations where the accuracy and credibility of the older adult eyewitness can be maximized. A logical initial step toward achieving this is to gain an appreciation of the conditions under

22.

EYEWITNESS MEMORY IN OLDER ADULTS

639

TABLE 22.1. Eyewitness Testimony Phenomena and Putative Age Effects

Eyewitness Evidence

Theoretical Rationale

Topic

Statement

Weapon focus

The presence of a weapon impairs an eyewitness's ability to accurately identify a perpetrator's face.

Unknown

Unknown

It is possible that older adults will show increased focus-due to perceptual decline affecting peripheral vision. However, it is also conceivable that because of inhibitory deficits, they will not show the normal focus-they will fail to inhibit less relevant (i.e., nonweapon) information.

Lineup instructions

Police instructions can affect an eyewitness's willingness to make an identification.

Unknown

Unknown

There is no reason to suggest that age will interact with this phenomenon.

Exposure time

The less time an eyewitness has to observe an event, the less well he or she will remember it.

Unknown

Age effect

Laboratory-based research on word lists suggests that the larger the cognitive demands on a participant at encoding (e.g., by reducing study time), the larger the difference between young and old groups. Thus, this exposure time effect should be even larger for older adults.

Forgetting curve

The rate of memory loss for an event is greatest right after the event and then levels off over time.

Age effect

Age effect

The fact that older adults forget faster than younger adults has been demonstrated in eyewitness studies and more mainstream tasks.

Accuracy confidence

An eyewitness's confidence is not a good predictor of his or her identification accuracy.

No age effect

Equivocal

Studies of eyewitness testimony in older adults suggests that confidence is unaffected by age, inasmuch as both young and old group's evaluations are unrelated to performance. There is some evidence that older adults should be even less accurate than controls.

Post-event information

Eyewitness testimony about an event often reflects not only what witnesses actually saw but information that they obtained later on.

Age effect

Age effect

Because older adults are prone to source confusions, it is likely that they will be more likely to confuse actual events with information acquired later on. They may also be more prone to remembering information that was later found to be untrue, whereas young people will be able to forget this information.

Wording of questions

Eyewitness testimony about an event can be affected by how the questions put to that witness are worded.

Unknown

Unknown

The effect of age on this phenomenon is unknown, but it is likely that environmental support given to older adults at study will aid retrieval, but it may also increase false remembering.

Evidence

(continued)

640

MOULIN ET AL.

TABLE 22.1. (Continued)

Topic

Statement

Eyewitness Evidence

Theoretical Rationale

Evidence

Unconscious transference

Eyewitnesses sometimes identify as a culprit someone they have seen in another situation or context.

Equivocal

Age effect

Attitudes and expectations

An eyewitness's perception and memory for an event may be affected by his or her attitudes and expectations.

Unknown

Age effect

Cross-race bias

Eyewitnesses are more accurate when identifying members of their own race than members of other races.

Unknown

Unknown

Outside of cohort effects due to societal differences between young and older groups, there is no reason to expect different race effects across groups. On the other hand, there is evidence that there is a cross-age bias.

Confidence malleability

An eyewitness's confidence can be influenced by factors that are unrelated to identification accuracy.

Unknown

No age effect

One factor known to affect both confidence and accuracy is age. There is no direct test of how older adults' confidence changes in comparison with younger adults. There is no reason to suggest that older adults will be any more or less susceptible to malleability.

Mugshotinduced bias

Exposure to mug shots of a suspect increases the likelihood that the witness will later choose that suspect in a line-up.

Age effect

Age effect

Older adults tend to use familiarity as the basis for reporting that a stimulus has been encountered before. Coupled with this is a lack of source or contextual information, indicating that older adults will be more susceptible to this bias than young.

Presentation format

Witnesses are more likely to misidentify someone by making a relative judgment when presented with a simultaneous (as opposed to sequential) lineup.

Unknown

Unknown

There is no reason to suspect that age will interact with this phenomenon.

Source confusion and failures of inhibition at encoding are likely to lead to unconscious transference. However, direct tests of this notion suggest that the age effect in eyewitness testimony is somewhat dependent on the materials used. Older adults have been shown to be overreliant on semantic-memory-based gist processing. This suggests that they may overuse their expectations and preexisting attitudes in eyewitness situations.

Each statement is assessed. "Eyewitness evidence" suggests that there is empirical data from studies of young and old on eyewitness materials. "Theoretical rationale" suggests that aside from direct evidence, the older adult memory literature indicates a possible age difference.

22.

EYEWITNESS MEMORY IN OLDER ADULTS

641

which older adults perform poorly at eyewitness-type tasks, and subsequently to try to minimize the older adult's exposure to these. The usual framework for discussing these sorts of issues is between system variables and estimator variables (Wells, 1978). System variables are those under the control of the judicial system, and thus we have some control over them, for example, how to ask questions that minimize the effects of suggestibility. Estimator variables are things such as exposure duration, which we do not have control of; indeed, in the context of this chapter, a major estimator variable might be the age of the witness. Estimator variables are likely to be important to decisions about eyewitness credibility and may well be informed by laboratory studies comparing the memory, perception, and general cognitive performance of younger adults and older adults. The focus here, however, as with most practical suggestions for eyewitness testimony, is system variables (for example, see Wells et al., 2000), because with the nature of eyewitness situations, it is clearly impractical to support the older adult at the encoding stage. It would be impossible to increase exposure times and reduce distractions when a crime is witnessed. These situations typically occur rapidly and are very stressful. Therefore a more logical approach is to maximize support at the retrieval stage. Craik (1986) proposed that older adults can be helped at retrieval by environmental support, with the greatest benefits being observed when the older adult has to engage in a self-initiated task (e.g., recalling to mind the events of a crime). The discussion now turns to a review of techniques that may act to improve the quality of the older adult's eyewitness testimony. 1. Question type: The use of a direct questioning technique (e.g., What color was the thief's hair?) has been shown to produce more accurate testimonies in young and older adults than a cross-examination technique (e.g., Did the burglar kick or hit the window to break it?). Improvements in accuracy gained with the use of the direct questioning technique are greatest in older adults. For example, Brimacombe (1997) reported an accuracy level of just 48%, using the cross-examination technique in older adults, which increased to 78% with the direct questioning technique. Of course, asking direct questions goes against the principles of the Cognitive Interview (CI; Fisher & Geiselman, 1992), which recommends open questions that elicit free recall. Free recall should be more accurate than response to questions, which are open to bias. However, older adults are likely to have much poorer free recall. Our interpretation of these two contradictory pieces of information is that direct questions lead to more correct information and to more errors. We propose that this pattern is likely to be accentuated with age. That is, older adults may show a greater benefit of direct questioning, but at the cost of more erroneous information being produced. Further research should explore these possibilities, but a suggestion based on the principles of the cognitive interview and taking into account the limitations in memory as discussed above is to start with open questions and move to direct questions, followed by closed (recognition memory-type) questions. An innovative approach may consider subjective evaluations of source, recollective experience, and confidence alongside the retrieval of content, as a way of trying to reduce false memories based on gist processing (although many studies suggest that false memories are retrieved with misleading sensations of recollection and high levels of confidence; Schacter & Dodson, 2001). It is likely that the best way of reducing the impact of false

642

MOULIN ET AL.

memories on eyewitness testimony is for the professionals involved in the questioning process to be aware of the cognitive abilities of the witness and the impact of different kinds of questioning on different memory types. 2. Context reinstatement: Recreating the context of an event (e.g., witnessing a crime) that was initially encountered via mental imagery increases recall of the details of the event. The CI technique allows an eyewitness to revisit the source of the original source of the information. Participants are encouraged to think about how they were thinking and feeling at the time of the incident. Mello and Fisher (1996) reported that the CI technique stimulated the production of more information (without reducing accuracy) than a standard interview technique with older adult participants. However, Searcy et al. (200 1) reported no increase in the accuracy of recall in CI compared with a standard interview technique; thus the evidence for the effectiveness of the CI technique in older adults remains equivocal. 3. Crime scene variables: The research reviewed above suggests that older adults are less accurate at recalling the details of a crime scene and are more susceptible to false identifications in a lineup parade. This is, of course, potentially damaging; however, it should not go without note that the majority of the crime scenes used in the experiments reviewed above have used young perpetrators. It appears that intergenerational identification is less accurate than intragenerational identification, suggesting that one way to improve older adult's eyewitness ability is to consider the material on which he or she is testifying. For minority race lineups, one solution has been to take more care in selecting the people used (making sure that the people are all drawn from the same racial group). We propose that the same sort of care should be exercised in creating lineups where age is likely to be critical (e.g., young person witnessing an old perpetrator). 4. Identification procedures: Two identification procedures are typically used by police: a simultaneous lineup and a sequential lineup. 1 In the simultaneous lineup situation, participants are exposed to all faces at the same time and asked to select a perpetrator. In contrast, during a sequential lineup participants are exposed to only a single face at any one time and are asked to make a judgment on the face before moving on to the next face. If participants identify a face as the perpetrator, the procedure is stopped and no further faces are shown. The sequential exposure technique has been shown to reduce false identifications in target absent situations (Lindsay & Wells, 1985). Memon and Gabbert (2003) reported that false identification rates dropped from .90 (in a simultaneous lineup) to .40 (in a sequential lineup situation) when the target perpetrator was absent. No difference in accuracy rates was reported with the target present, although there was a trend for increased false identification by the older adults. Therefore, when a lineup parade does not contain the perpetrator, the accuracy of older adults' responses can be improved via the use of a sequential lineup.

Lindsay and Wells (1985) argue that the sequential and simultaneous lineups require different response strategies. In the simultaneous lineups people can choose the person who looks the most like their memory of the culprit. However, they argue that in the sequential lineups people need to adopt a threshold strategy. Although the Wells et a!. (2000) report recommends either type of lineup, ongoing research is considering the merits of each approach. 1

22.

EYEWITNESS MEMORY IN OLDER ADULTS

643

SUMMARY Older adults have been shown to produce a number of errors of omission and commis~ sion in memory that can be seen to be of direct relevance to eyewitness situations. The types of memory errors described in older eyewitnesses are captured in the notion that the older adult's memory can be described as similar to the pattern resulting from ac~ quired damage to the frontal lobes. Both the eyewitness and cognitive neuropsychology of aging literatures have identified source~monitoring errors and overreliance on famil~ iarity or gist as particularly important. A number of theoretical viewpoints, such as re~ duced recollective experience, the relative preservation of familiarity, the impairment of conscious retrieval (recollection), and inhibitory dysfunction, are well represented in the mainstream aging literature, and more studies of eyewitness ability and aging ought to consider these issues. Finally, more recent work indicates that older adults are affected by the age of the face stimuli that they are tested on, suggesting that the question for those interested in eyewitness testimony is not only how age affects eyewitness abilities, but more how it interacts with the subject matter in question.

REFERENCES Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by executive control. Nature, 410(6826), 366-369. Balota, D. A., Dolan, P. 0., & Duchek, J. M. (2000). Memory changes in healthy older adults. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 395-409). New York: Oxford University Press Bjork, E. L., & Bjork, R. A. (1996). Continuing influences of to-be-forgotten information. Consciousness and Cognition, 5, 176-196. Bornstein, B. H. (1995). Memory processes in elderly eyewitnesses-What we know and what we don't know. Behavioral Sciences & The Law, 13,337-348 Brimacombe, C. A., Quinton, N., Nance, N., & Garrioch, L. (1997). Is age irrelevant? Perceptions of young and old adult eyewitnesses. Law and Human Behavior, 21 (6), 619-634. Cabeza, R. (2002). Hemispheric asymmetry reduction in old adults: The HAROLD Model. Psychology and Aging, 17, 85-100. Cohen, G., & Faulkner, D. (1989). Age differences in source forgetting: Effects on reality monitoring and on eyewitness testimony. Psychology & Aging, 4 (1), 10-17. Connor, L. T., Dunlosky, ]., & Hertzog, C. (1997). Age-related differences in absolute but not relative metamemory accuracy. Psychology and Aging, 12, 50-71. Conway, M. A., Harries, K., Noyes, J ., Racsma'ny, M., & Frankish, C. R. (2000). The disruption and dissolution of directed forgetting: Inhibitory control of memory. Journal of Memory and

Language,43(3),409-430.

Craik, F. I. M. (1986). A functional account of age differences in memory. In F. Klix & H. Hagendorf (Eds.), Human memory and cognitive capabilities, mechanisms and performances (pp. 409422). Amsterdam: Elsevier, North-Holland. Craik, F.l. M., &Jennings, J. M. (1992). Human memory. In F.l. M. Craik & T. A. Salthouse (Eds.), Handbook of aging and cognition (pp. 51-83). Hillsdale, NJ: Lawrence Erlbaum Associates. Craik, F. I. M., & McDowd, J. M. (1987) age-differences in recall and recognition. journal of Experimental Psychology-Learning Memory and Cognition, 13, 474-479. Davidson, P. S. R., & Glisky, E. L. (2002). Neuropsychological correlates of recollection and familiarity in normal aging. Cognitive, Affective & Behavioral Neuroscience, 2, 174-186.

644

MOULIN ET AL.

Duncan, J., Johnson, R., Swales, M., & Freer, C. (1997). Frontal lobe deficits after head injury: Unity and diversity offunction. Cognitive Neuropsychology, 14, 713-741. Dunlosky, J., & Connor, L. T. (1997). Age differences in the allocation of study time account for age differences in memory performance. Memory and Cognition, 25, 691-700. Dysart, J., Lindsay, R. C. L., Hammond, R., & Dupuis, P. (2001). Mugshot exposure prior to lineup identification: Interference, transference and commitment effects. Journal of Applied Psychology, 86, 1280-1284. Dywan, J., & Jacoby, L. (1990). Effects of aging on source monitoring-Differences in susceptibility to false fame. Psychology & Aging, 5, 379-387 Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques for investigative interviewing. Springfield, IL: Charles C. Thomas. Foisy, P. (1994). Age-related deficits in intentional memory for spatial location in small-scale space-A metaanalysis and methodological critique. Canadian Journal on Aging, 13, 353-367. Gardiner, J. M., & Richardson-Klavehn, A. (2000). Remembering and knowing. In E. Tulving & F.l. M. Craik (Eds.), Handbook of memory (pp. 229-244). New York: Oxford University Press. Greenwood, P. M. (2000). The frontal aging hypothesis evaluated. Journal of the International Neuropsychological Society, 6, 705-726 Hasher, L., Quig, M. B., & May, C. P. (1997). Inhibitory control over no-longer-relevant information: Adult age differences. Memory & Cognition, 25(3), 286-295. Hasher, L., & Zacks, R. T. (1988). Working memory, comprehension, and aging: A review and a new view. In G. H. Bower (Ed.), The psychology of learning and motivation (Vol. 22, pp. 193225). San Diego: Academic Press. Henry, J.D., MacLeod, M.S., Phillips, L. H., & Crawford, J. R. (2004). A meta-analytic review of prospective memory and aging. Psychology & Aging, 19, 27-39. Ivy, G. 0., MacLeod, C. M., Petit, T. L., & Markus, E. J. (1992). A physiological framework for perceptual and cognitive changes in aging. In F. I. M. Craik and T. A. Salthouse (Eds.), Handbook of aging and cognition (pp. 273-314). Hillsdale, NJ: Lawrence Erlbaum Associates. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28. Kassin, S.M., Tubb, V. A., Hosch, H. M., & Memon, A. (2001). On the "general acceptance" of eyewitness testimony research-A new survey of the experts. American Psychologist, 56, 405-416. Kausler, D. H., Salthouse, T. A., & Saults, J. S. (1988). Temporal memory over the adult lifespan. American Journal of Psychology, 101, 207-215. Kester, J.D., Benjamin, A. S., Castel, A. D., & Craik, F. I. M. (2002). Memory in elderly people. In A. D. Baddeley, M.D. Kopelman, & B. A. Wilson (Eds.), The handbook of memory disorders (2nd ed., pp. 543-567). Chichester, UK: Wiley. Kwong See, S. T., Hoffman, H. G., & Wood, T. L. (2001). Perceptions of an old female eyewitness: Is the older eyewitness believable? Psychology and Aging, 16 (2), 346-350. Light, L. L. (1991). Memory and aging: Four hypotheses in search of data. Annual Review of Psychology, 42, 333-376. Lindenberger, U., & Baltes, P. B. (1994). Sensory functioning and intelligence in old-ageA strong connection. Psychology & Aging, 9, 339-355. Lindsay, R. C. L., & Wells, G. L. (1985). Improving eyewitness identifications from lineups: Simultaneous versus sequential lineup presentation. Journal of Applied Psychology, 70, 556-564. List, J. A. (1986). Age and schematic differences in the reliability of eyewitness testimony. Developmental Psychology, 22, 50-57. MacDonald, S. W. S., Hultsch, D. F., & Dixon, R. A. (2003). Performance variability is related to change in cognition: Evidence from the Victoria longitudinal study. Psychology & Aging, 18,510-523. Macleod, M. (2002). Retrieval-induced forgetting in eyewitness memory: Forgetting as a consequence of remembering. Applied Cognitive Psychology, 16, 135-149.

22. EYEWITNESS MEMORY IN OLDER ADULTS

645

Mello, E. W., & Fisher, R. P. (1996). Enhancing older adult eyewitness memory with the cognitive interview. Applied Cognitive Psychology 10(5), 403-417. Memon, A., Bartlett, J., Rose, R., & Gray, C. (2003). The aging eyewitness: Effects of age on face, delay, and source-memory ability. Journals of Gerontology Series B-Psychological Sciences and Social Sciences, 58, 338-345. Memon, A., & Gabbet, F. (2003). Improving the identification accuracy of senior witnesses: Do prelineup questions and sequential testing help? Journal of Applied Psychology, 88(2), 341-347. Memon, A., Hope, L., Bartlett, J., & Bull, R. (2002). Eyewitness recognition errors: The effects of mugshot viewing and choosing in young and older adults. Memory and Cognition 30(8), 1219-1227. Moulin, C. J. A., Perfect, T. J., &Jones, R. W. (2000). Evidence for intact memory monitoring in Alzheimer's disease: Metamemory sensitivity at encoding. Neuropsychologia, 38, 1242-1250. Nelson, T. 0., & Narens, L. (1994). Why investigate metacognition? In J. Metcalfe & A. P. Shimamura (Eds.), Metacognition: Knowing about knowing. London: MIT. Nunez, N., McCoy, M. L., Clark, H. L., & Shaw, L.A. (1999). The testimony of elderly victim/ witnesses and their impact on juror decisions: The importance of examining multiple stereotypes. Law and Human Behavior, 23(4), 413-423. Parkin, A. J., & Java, R. I. (1999). Determinants of age-related memory loss. InT. J. Perfect & E. Maylor (Eds.), Models of cognitive aging. Oxford: Oxford University Press. Parkin, A. J., & Walter, B. M. (1992). Recollective experience, normal aging, and frontal dysfunction. Psychology and Aging, 7, 290-298. Perfect, T. J. (1997). Memory aging as frontal lobe dysfunction. In M.A. Conway (Ed.), Cognitive models of memory. Hove, UK: Psychology Press. Perfect, T. J., & Dasgupta, Z. R. R. (1997). What underlies the deficit in reported recollective experience in old age? Memory & Cognition, 25, 849-858. Perfect, T. J., & Harris, L. J. (2003). Adult age differences in unconscious transference: Source confusion or identity blending? Memory & Cognition, 3, 570-580. Perfect, T. J., & Hollins, T. S. (1999). Feeling of knowing judgements do not predict subsequent recognition performance for eyewitness memory. Journal of Experimental Psychology-Applied, 5, 250-264. Perfect, T. J., Moulin, C. J. A., Conway, M.A., & Perry, E. (2002). Assessing the inhibitory account of retrieval-induced forgetting with implicit-memory tests. Journal of Experimental Psychology: Learning Memory and Cognition, 28(6), 1111-1119. Perfect, T. J., & Stollery, B. (1993). Memory and metamemory performance in older adultsOne deficit or two? Quarterly Journal of Experimental Psychology, 46, 119-135. Perfect, T. J., Williams, R. B., & Anderton-Brown, C. (1995). Age-differences in reported recollective experience are due to encoding effects, not response bias. Memory, 3, 169-186. Ross, D. F., Dunning, D., Toglia, M.P., & Ceci, S. J. (1990). The child in the eye of the jury: Assessing mock jurors' reactions to child victim witnesses: Factors affecting trial outcome. Law and Human Behavior, 14, 5-23. Schacter, D. L., & Dodson, C. S. (2001). Misattribution, false recognition and the sins of memory. Philosophical Transactions of the Royal Society of London, 356, 1385-1393. Schacter, D. L., Kaszniak, A. W., Kihlstrom, J. F., & Valdiserri, M. (1991). The relation between source memory and aging. Psychology & Aging, 6, 559-568. Scogin, F., Calhoon, S. K., & Derrico, M. (1994). Eyewitness confidence and accuracy among three age cohorts. Journal of Applied Gerontology, 13, 172-184. Searcy, J. H., Bartlett, J. C., & Memon, A. (1999). Age differences in accuracy and choosing in eyewitness identification and face recognition. Memory & Cognition, 27, 583-552. Searcy, J. H., Bartlett, J. C., & Memon, A. (2000). Influence of post-event narratives, lineup conditions and individual differences on false identification by young and older eyewitnesses. Legal and Criminological Psychology, 5, 219-235.

646

MOULIN ET AL.

Searcy, J. H., Bartlett, J. C., Memon, A., & Swanson, K. (2001). Aging and lineup performance at long retention intervals: Effect of metamemory and context reinstatement. Journal of Applied Psychology, 86(2), 207-214. Shilling, V. M., Chetwynd, A., & Rabbitt, P.M. A. (2002). Individual inconsistency across measures of inhibition: An investigation of the construct validity of inhibition in older adults. Neuropsychologia, 40(6), 605-619. Souchay, C., lsingrini, M., & Espagnet, L. (2000) Relations between feeling-of-knowing and frontal lobe functioning in older adults. Neuropsychology, 14, 299-309. Stuss, D. T., & Levine, B. (2002). Adult clinical neuropsychology: Lessons from studies of the frontal lobes. Annual Review of Psychology, 53, 401--433. Tulving, E. (1983). Elements of episodic memory. New York: Oxford University Press. Tun, P. A., Wingfield, A., Rosen, M. J., & Blanchard, L. (1998). Response latencies for false memories: Gist-based processes in normal aging. Psychology & Aging, 13, 230-241. Wells, G. (1978). Applied eyewitness testimony research: System variables and estimator variables. Journal of Personality and Social Psychology, 36, 1546-1557. Wells, G. L., Malpass, R. S., Lindsay, R. C. L., Fisher, R. P., Turtle, J. W., & Fulero, S. (2000). From the lab to the police station: A successful application of eyewitness research. American Psychologist, 55, 581-598. Wells, G., & Seelau, L. (1995). Eyewitness identification: Psychological research and legal policy on lineups. Psychology, Public Policy, and Law, I, 765-791. Wheeler, M. A. (2000). A comparison of forgetting rates in older and younger adults. Aging, Neuropsychology & Cognition, 7, 179-193. Wright, D. B., Boyd, C. E., & Tredoux, C. G. (2001). A field study of own-race bias for face recognition in South Africa and England. Psychology, Public Policy, and Law, 7, 119-133. Wright, D. B., Loftus, E. F., & Hall, M. (2001). Now you see it; now you don't: Inhibiting recall and recognition of scenes. Applied Cognitive Psychology, 15,471-482 Wright, D. B., & Stroud, J. N. (2002). Age differences in lineup identification accuracy: People are better at their own age. Law and Human Behavior, 26(6), 641-654. Yonelinas, A. P. (2002). The nature of recollection and familiarity: A review of 30 years of research. Journal of Memory and Language, 46,441-517. Zacks, R. T., Radvansky, G., & Hasher, L. (1996). Studies of directed forgetting in older adults. Journal of Experimental Psychology-Learning Memory and Cognition, 22(1), 143-156.

v CONCLUSION

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

23 The Relevance of Eyewitness Research: A Trial Lawyer's Perspective Don Thomson Charles Sturt University, New South Wales

My goal in this chapter is to identify and evaluate the potential impact of the research findings reviewed in the 22 chapters of this volume on the justice system, particularly the criminal justice system. My comments are from the perspective of a trial lawyer practicing within a common-law justice system. The plan of this chapter is first to give a thumbnail sketch of the criminal justice system, and then to identify the implications of the findings reported in this volume for the justice system. It should be noted that, notwithstanding the fact that all common-law jurisdictions adopted the common-law system of England at the time of colonization, since that time the common-law system in the various jurisdictions has developed in response to the specific demands and values of those jurisdictions, for example, the manner in which judges are appointed and jurors are selected in the United States is very different from the way it is done in England, Australia, or New Zealand. THE COMMON-LAW CRIMINAL JUSTICE SYSTEM

Three stages can be distinguished in the common-law criminal justice system: the investigatory or pretrial stage, the evidentiary or trial stage, and the offender management stage. The chapters in this volume are relevant to the first two stages. In theory, and often in practice, these two stages are discrete. They differ in many ways. The goal of the investigatory stage is to gather evidence from witnesses and crime scene analyses to determine whether an offense may have occurred and, if so, who should be charged with that offence. The goal of the trial stage is to determine, based on the evidence before the court, whether the act in question constituted an offense and whether the accused person was the person who committed that offense. The standard of proof required in the 649

650

THOMSON

investigatory stage is, at best, a more probable-than-not, whereas in the trial stage, the standard of proof is beyond reasonable doubt. The pretrial stage is in the hands of the investigatory agencies such as police, customs, tax, or revenue departments. If, and when, the investigatory agency has completed its investigation, the matter is passed on to the prosecutor's office. At that time the prosecutor assumes control of the proceedings. The prosecutor decides whether there is sufficient evidence to proceed to trial with the matter, whether further investigations are required, and whether to proceed further with the matter. The two stages differ also in the way evidence against the accused person is educed. The manner in which a prosecutor presents evidence against the accused in court is rigidly controlled by the laws of evidence; for example, in examination-in-chief, leading questions are not permitted, and hearsay evidence is excluded except under special conditions, for example, an official public document such as a marriage certificate, confessions (provided that they are voluntary), or the declaration of a dying person. In contrast, the investigator has much greater freedom in pursuing the course of his or her inquiry. The investigator can interview a suspect or any other witness multiple times, and whereas the electronically recorded interview with witnesses and defendants is unlikely to include leading and suggestive questions, unmonitored interviews and interrogations undertaken prior to the formal electronically recorded interviews are less likely to be constrained by the laws of evidence. Although the investigatory stage in many senses is independent of the evidentiary stage, what happens in the investigatory stage will substantially shape what will and can occur in the evidentiary stage. The common-law system of criminal justice can be contrasted with the inquisitorial system of criminal justice found in many European countries. In the inquisitorial system, the judge is charged with the responsibility of finding the truth. The judge, or more often the counsel assisting the inquiry, is responsible for ensuring that evidence is received from all relevant witnesses. It is the judge who will decide the verdict. In contrast, the common law has adopted an adversarial system of trial. The contesting parties determine what issues will be contested and decide what evidence will be put before the tribunal of fact, be that judge or jury. The judge's role in a trial by jury is restricted to ensuring that the parties adhere to the rules of the game, namely the rules of evidence, and instructing the jury on matters of law. It is the jury's role to consider the facts presented during the trial and decide the verdict. Where there is no jury, the judge assumes the role that would otherwise be exercised by a jury, in addition to ensuring that the trial is conducted according to the rules of evidence. The ultimate decision of the judge or jury therefore depends on what evidence is presented and how that evidence is presented. The process of obtaining evidence from witnesses falls into three stages. The first stage, the stage of direct evidence or examination-in-chief, is where the party calling the witness takes the witness through his or her evidence. This is the stage in which the witness has the opportunity to tell his or her story. The second evidentiary stage is the crossexamination. This is the stage in which the cross-examiner seeks to exploit the evidence given by the witness in direct evidence to the advantage of the contesting party and to undermine the credibility of a witness and the evidence given by that witness. The third evidentiary stage is reexamination. Reexamination is conducted by the party who questioned the witness in the first stage. This stage provides an opportunity to respond to any

23.

RELEVANCE OF EYEWITNESS RESEARCH

651

issues raised in cross-examination and to address any misleading impressions created during cross-examination. The authors of the 22 chapters, either explicitly or implicitly, raise issues for and problems with the current criminal justice system. These problems can be addressed in one of three ways. One way is radical: scrap the current system and replace it with another system. None of the authors are advocating this course of action. Another way to address identified shortcomings of the current criminal justice system is to effect specific changes that eliminate identified shortcomings. This appears to be the course of action being advocated by a number of the authors. However, because it is a system of criminal justice, changes cannot be effected in one part or aspect of the criminal justice system without having an impact on other parts or aspects of the system, for example, an unintended consequence of the elimination of jury trials potentially means more, as judges but not juries are required to give reasons for their decisions, and these reasons may not be grounds for appeals. A third way of addressing identified problems in the criminal justice system is the modification of the behavior of the parties in the criminal justice system in light of research findings. Many of the recommendations and conclusions of the authors fall into this category of reform.

INVESTIGATION ISSUES

The investigator in the criminal justice system has two major objectives. The first objective is to establish whether an offense has occurred and, if so, who committed that offense. The second objective is to prepare the case against the alleged offender for the prosecutor. Although these two objectives are related, there will be times when the procedures and requirements to achieve these two objectives are incompatible. Thus the emphasis in establishing what might have happened and who was responsible will be on maximizing the amount of detail about the event from witnesses. The more details that are obtained, the more leads the investigator has, even though many of the leads may turn out to be false trails. To obtain as much information as possible, investigators, such as police and child protection officers, will interview all witnesses, regardless of age, and may employ any technique that increases what witnesses report about the event in question. Details obtained from a witness, including hearsay evidence, may assist the police in uncovering other evidence that ultimately provides the basis for the successful conclusion of the investigation. The importance of distinguishing between the investigatory phase and the trial stage is underscored by the research findings outlined in the chapter by Pipe, Thierry, and Lamb. Very young children may lack the verbal skills to convey what happened to them, but they may well be able to reenact those events. Very young children may be unable to provide a free narrative about their experiences but, with appropriate cues or props, may be able to provide details that will materially assist the investigator's inquiry. However, the limitations in a young child's verbal and recall skills pose very real difficulties at the evidentiary stage, where the impact of suggestibility and inaccurate recall has the potential to have far-reaching consequences for the accused person. One recall deficit

652

THOMSON

apparent with young children is that of source monitoring. Young children are generally not as accurate as older children and adults in recalling a specific incident of a recurring event. Accurate recall by a witness of a particular event is of crucial importance, as the identification of that incident is necessary for the defendant to be able to marshal a defense against the allegation. The High Court of Australia in S v. The Queen (1989) outlined an additional reason or need to have particularization of charges. The court said that if evidence of allegation of repeated abuse is admitted without particularization of charges, it may be the case that one juror concludes that the defendant is guilty of committing one of the alleged offenses whereas other jurors do not, and each of these other jurors believes the defendant is guilty on a different charge. Thus, without particularization, the jury could find the defendant guilty, but on no one charge would there be any substantial agreement by the jurors. Pipe et al. have noted that some courts have concluded that "confusions between episodes does not necessarily cast doubt on the accuracy or credibility of young witnesses." Such a conclusion is inconsistent with what is commonly understood as accurate memory and flies in the face of research findings reported by Pipe et al. that younger children are more likely to forget the source of their memories than are older children. Although the emphasis in the investigatory stage is on finding out exactly what happened and thus eliciting as much information as possible, the investigation will progress much more quickly if interviewing techniques not only elicit the maximum amount of information from witnesses but also elicit a minimum of false and irrelevant details. One of the issues identified in a number of chapters is the role suggestibility is likely to play in shaping and reshaping memories during the investigatory stage. Perhaps the most pervasive and subtle form of suggestibility in the investigatory stage, and yet least scrutinized by courts, is confirmatory biases of investigators, or, as Melnyk, Crossman, and Scullin refer to it in their chapter, interviewer bias. Seldom do investigators seek evidence from witnesses that might disconfirm the investigators' hunch or hypothesis, and seldom do interviewers seek evidence which might support explanations other than the one they favor. The very questions asked, the answers that are ignored and the answers that result in further questions, the questions repeated, and the level of interest expressed in the tone of the interviewer's voice are all likely to convey to the witness the hypotheses being entertained by the interviewer, whether that interviewer is a parent or a police officer (Bruck, Ceci, Melynk, & Finkelberg, 1999; Thompson, Clarke-Stewart, & Lepore, 1997). The four major conclusions I would draw from the research reviewed by Melynk et al. on the suggestibility of children's memory are (1) that formal investigatory systems should be fashioned to maximize the completeness and accuracy of a child witness's account of events; (2) that all interviews of children by police officers be electronically recorded, not just the carefully rehearsed video-recorded interviews; (3) that the sequence of interviews, both formal and informal, to which the child has been subjected be documented; and (4) that the parties in a trial, the judge, the prosecutor, and the defense counsel, and the jury all be apprised of children's vulnerability to suggestion. I think the administering of a suggestibility test to witnesses to measure their suggestibility will have limited value. No investigating officer will ignore what a witness says because that witness has obtained a high score on a suggestibility scale, nor will the evi-

23.

RELEVANCE OF EYEWITNESS RESEARCH

653

deuce of that witness be excluded from the court because that witness has a high score on a suggestibility scale. The main value of administering such a scale is that the investigator and the tribunal of fact may scrutinize more thoroughly the sequence of interviews of that witness and the changes in that witness's account over interviews.

CREATION AND EVOLUTION OF MEMORIES

The thrust of the Davis and Loftus chapter is best captured by quoting a passage from that chapter: "original perceptions and judgments as well as later memory and judgments include a mixture of actual features of the original information, or added features resulting from schematic/inferential processing and/or influential information external to the event." The same conception of memory is sketched by Neuschatz, Lampinen, Toglia, Payne, and Cisneros in their chapter. Davis and Loftus review studies that show that people selectively attend to certain features of an event and that such feature selection will be determined by what the observer expects. Davis and Loftus hypothesize that these expectations are based on schemas or patterns that are built up from past experiences. Where there are gaps in what a person can recall, those gaps may be filled by features consistent with the schema. Furthermore, subsequent new information has the capacity to retroactively alter what is remembered about the event in question, aptly called hindsight bias. Media coverage of events is identified as one source of memory contamination. The Davis and Loftus chapter highlights that event memory is not static; rather, it is constantly evolving. The chapter by Davis and Friedman continues and extends the theme of the Davis and Loftus chapter. Davis and Friedman argue that memory for conversations is even more likely than memory for events to be forgotten over time and even more vulnerable to being distorted by subsequent events. The authors explore the question of how well people remember what exactly what was said, who said it, to whom the remarks were addressed, and when the conversation took place. I am unaware of any studies that have compared the impact on a juror of a witness recalling the gist of a conversation with the impact on a juror of a witness purporting to recall the conversations verbatim. My impression, as a trial lawyer, is that witnesses purporting to recall the exact words of a conversation have a greater impact on jurors. The findings reviewed by Davis and Friedman suggest that if jurors do give greater credence to the evidence of a witness who purports to recall conversations verbatim, that greater credence is probably misplaced. How do the investigator and the lawyer establish whether what the witness reports is an accurate or false memory? In controlled studies it has been found that false memories compared with true memories lack sensory details, contextual details, and details about the person's own feelings at the time of the event. However, as Neuschatz et al. point out, the investigator lacks a reference point against which to assess whether what is reported is a true or false memory. Two cases in which I have been involved during the preparation of this chapter illustrate the points made by Davis and Loftus and Davis and Friedman. In the first case, a witness observed a body falling from an overpass onto a roadway. In the first statement

654

THOMSON

taken from the witness soon after the event, she said she was sitting in the front passenger seat of a car that was stopped at a traffic light and was dimly aware of a man or men walking up the stairs of a nearby overpass. As she gazed into the night, waiting for the lights to change, a body fell from the overpass, crashing onto the road just in front of her car. It was clear to the witness that the body splattered on the road in front of her was lifeless. At this time the witness appeared to interpret the tragic event as either accidental or suicide. However, when interviewed again later, this time by a member of the police homicide squad, her story underwent subtle changes. She now stated that she saw a group of four young men climbing. In her evidence given at the coroner's inquest 5 years after the event, the witness told the coroner that because it was dark she could not see whether the people on the overpass were facing her or had their backs to her. However, in the preliminary trial, 1 year after the coroner's inquest, the witness testified that the overpass was well lit. Prior to the preliminary trial the witness had maintained that she could not say whether a car that drove away soon after she saw the body fall from the overpass was an old or new one, but in her evidence in court she described the car as being an old model. In a video-recorded reenactment of the event, asked by the police whether she had been able to hear whether the group of men were arguing as they moved across the overpass, the witness requested the police officers to go up on the overpass and argue, so that she could work out whether or not she would have heard an argument on the night when the event occurred. Also in the videotaped reenactment, when asked by one of the detectives where the people on the overpass were, the woman said, "I came here with some other police the other night, and we decided that the people must have been in front of the light pole because I can recollect seeing there was another light pole behind it." In an interview where she was hypnotized, the witness was asked whether she was able to get a picture of the deceased person's face. The witness replied, "The only picture that seems formed is that picture that has been spread out in the media." In this case, it was possible to identify the changes in the witness's account of events leading up to the death of the young man and to identify the subsequent events that may have been the cause of those changes. In addition, the defense had the handwritten statement of the witness made 3 days after the event. It had a typed version of the original statement. This typed statement, provided 6 days after the event, contained a number of additions and modifications to the handwritten statement. The third statement given to police by the witness was made 7 days after the previous statement, that is, 13 days after the event. The defense also had a transcript and audiotape of the witness recalling the event when the witness had been hypnotized. This hypnotic recall was made sometime within a 5-month period following the third statement. There was also a transcript and video recording of the witness being interviewed during the reenactment of the event 5 months after the witness had made her last statement. Finally, there was a transcript of the witness's evidence at the coroner's inquest 5 years later and a transcript of the witness's evidence at the preliminary triall year after the coronial inquest. In the second case, a defendant was accused of digitally penetrating a 5-year-old girl as she slept in her bed. The defendant was one of many who had been invited by .the

23.

RELEVANCE OF EYEWITNESS RESEARCH

655

mother of the girl to a New Year's Eve party. Some hours after the party had started, the mother found the defendant in the child's bedroom. The mother stated that the child told her that the intruder had touched her "fanny." The mother questioned the child and then told those present at the party that her child had been sexually interfered with. It is not clear whether any of the people attending the party then questioned the child. The mother phoned the father of the child, the mother's former boyfriend, telling him the child had been sexually abused. The father arrived soon after with his present girlfriend and another couple. After a heated exchange between the mother of the child and the father in front of the child, the father bundled the child up and drove off. Dur~ ing the journey to the father's house, the child was questioned by the four adults in the car. In the morning, when the child woke up, she was questioned by the father's girlfriend. That afternoon the child was subjected to a medical examination and questioned by the doctor. The medical examination was followed by a "disclosure" interview with two policewomen in preparation for the videotaped interview, which occurred straight after. In the trial, which took place 14 months after the alleged offense, the prosecution evidence comprised the testimony of the mother, the father, the father's girlfriend, the couple, the medical practitioner, the two policewomen, and, finally, the videotaped interview with the child. The only persons to have made any record of what the child said were the two policewomen, but this record did not contain the questions put to the child that elicited the child's responses. Aside from the police witnesses, the evidence of all the other witnesses was the gist of what they said the child told them. The absence of any record both of what questions had been put to the child and what the child said denied the court the opportunity to exclude the possibility that the child's videotaped statements could be attributed to suggestible questions put to the child by any of the persons who had questioned her prior to the videotaped interview. The findings of the studies described by Davis and Loftus, Davis and Friedman, and Neuschatz et al. have significant implications for investigators and lawyers, particularly defense lawyers. One implication the research findings have for investigators is that they should be aware that the information they provide to a witness as to the purpose of the investigation and the questions they ask have the potential to alter the witness's memory of the event that is the subject of the investigation. A second implication is that in~ vestigators should identify potential sources that may have already biased witnesses' accounts, for example, media coverage of the event or discussion between witnesses. For the lawyers the clear and unmistakable lesson to be learned from the findings reviewed by Davis and Loftus, Davis and Friedman, and Neuschatz et al. is that the content of testimony offered in court by witnesses should be scrutinized very carefully to identify discrepancies with witnesses' earlier accounts, particularly their first account. Lawyers would be well-advised to chart events to which an eyewitness has been exposed in the interlude between the event in question and the eyewitness's account of that event in order to identify what subsequent events may have influenced the eyewitness's memory of the event in question. The findings in these chapters also underscore the great importance of accurate recording of all interviews with witnesses as a means of maximizing the possibility of a fair and just outcome.

656

THOMSON

DELAY AND RELIABILITY OF EVIDENCE

The time that elapses between the commission of an offense and the investigation of that offense can range from hours to years. Likewise, the time that elapses between the conclusion of the investigation of an offense and the trial of the person(s) alleged to have committed that offense is likely to range from months to one or more years. The reasons for delays are many and varied, for example, disclosures of alleged sexual abuse in childhood are not made at the time the alleged offense occurs because the then child was too frightened or ashamed to do so, the investigation of the allegation is protracted, there is a backlog of other cases awaiting trial, and adjournments are sought and ob~ tained by prosecutors and defense lawyers. Thus, at the one extreme, war crime trials and childhood sexual abuse trials may occur 30, 40, or even 50 years after the alleged offense occurred. At the other end of the continuum, at the very least, it is likely that several months will have passed since an alleged offense occurred and the trial of the alleged offender begins. In assessing the impact of delay on what an eyewitness can recall or recognize, it is useful to distinguish between errors of omission and errors of commission. Overall, the pattern that emerges from the studies reviewed by Read and Connolly is that with in~ creasing delay there is an increasing number of errors of omission and commission, with errors of omission increasing more rapidly than errors of commission. However, as Read and Connolly point out, the rate of decline in memory performance is not just a function of delay per se, but also factors such as the extent to which attention is directed to an event and the type and quantity of subsequent activities and events, as also noted by Davis and Loftus, and Davis and Friedman. Read and Connolly point out that one of the difficulties confronting investigators and courts is that, in the absence of corroborating TV, there is no way of assessing the com~ evidence, such as footage from a closed~circuit pleteness or the accuracy of what a witness reports. A question posed by Read and Connolly, Reisberg and Heuer, and Smith and Gleaves is whether memories of emotional events, including those that may be personally traumatic, are different in kind or are more enduring than memories of mundane events. This issue has importance for investigators and courts because of the increasing number of allegations made by adults of being physically or sexually abused in childhood. The findings of the research in these three chapters appear to show that memory for emotional events follows the same patterns as memory for mundane events, but the decline in accuracy is slower. However, research has also shown that memories for emotional events have a vividness and are recalled with a certainty and an intensity that are not found with more mundane events. The vividness of these memories and the certainty of the complainants about what they remember clearly make an impression on investigators and jurors alike. Both investigators and courts need to be aware that vividness and certainty do not necessarily mean the memory is accurate. However, in the justice system it is not just memory that is a casualty of time. Where there are long intervals between an alleged offense and the investigation and trial, key witnesses die or become otherwise unavailable, and relevant documents are destroyed or misplaced. Witnesses and documents that may, on the one hand, corroborate com-

23.

RELEVANCE OF EYEWITNESS RESEARCH

657

plainants' allegations or, on the other hand, corroborate defendants' denial of wrongdoing are unavailable. The unavailability of witnesses and documents because of a long interval paradoxically means the investigation and trial will be more dependent on people's memories. These issues are illustrated in the following case. The 36-year-old niece of a defendant alleged that the defendant had sexual intercourse with her when she was 13 years of age. The niece said that she had told her mother that her uncle had raped her. The mother had since died. The defendant denied the allegation. At the time of the alleged rape he was 41 years old. His defense was that at the time of the alleged rape he had an erectile dysfunction and it was not physically possible for him to have done what his niece had alleged. The year after the alleged rape had occurred, the defendant had sought medical advice for his condition and subsequently had surgical intervention, which was unsuccessful. When the defendant attempted to document his dysfunction he discovered that the surgeon was dead, the hospital had closed down, and all records relating to his admission as a patient and his surgery had been destroyed in conformity with provisions of a statutory act, the Privacy Act. The prosecution case rested on the evidence of the niece and psychiatric reports, the defense case on the denial by the defendant that he had touched his niece and the defendant's wife who testified to her husband's impotence. Enactment of statutes of limitation in most jurisdictions is recognition that long delays in initiating legal action and bringing matters to trial compromises the ability of defendants to mount their defense. Under the provisions of a statute of limitation, legal action against the defendant is barred unless that legal action is initiated within a specified time. Statutes of limitation apply to civil litigation matters in most jurisdictions, in some jurisdictions statutes of limitation apply to less serious criminal offenses, but in most jurisdictions (outside the United States) there are no statutes oflimitation for serious criminal offenses. However, in some jurisdictions where the court is satisfied that the delay in prosecuting the defendant has severely disadvantaged the defendant, the charges may be dismissed on the grounds of abuse of process.

EYEWITNESS TESTIMONY, AGE, AND PSYCHOLOGICAL IMPAIRMENT

One of the identifiable issues examined in this volume is the accuracy of eyewitness testimony as a function of age and psychological impairment. In this section the implication of findings reported in the various chapters for children, the elderly, and persons with psychological impairment are considered.

Children Four chapters in the present volume examine issues concerning children's evidence. In their very comprehensive chapter, Pipe et al. trace developmental changes in testimonial skills from as early as 18 months of age. The findings from the research reviewed in this chapter suggest that children as young as 2 years of age can recall experiences, but

658

THOMSON

the quantity of details recalled is limited, and although prompts increase the quantity of details reported, the accuracy of details provided in response to prompts declines. This latter point is highlighted in the chapters by Melnyk, Crossman, and Scullin and Lamb, Orbach, Warren, Esplin, and Hershkowitz, which identify circumstances such as repeated questioning, prompts, drawings, and use of anatomically detailed dolls, which are shown to undermine the reliability of the child's account of events. In their chapter, Reyna, Mills, Estrada, and Brainerd distinguish between verbatim and gist memory. They propose that verbatim memory provides a means for checking the accuracy of gist memories and that verbatim memories are less developed in young children and these memories are lost more quickly over time. Given that children do not retain verbatim memories for long, the opportunity to check the accuracy of children's gist memory by reference toverbatim memory is very limited, particularly if a long time has elapsed since the event in question. The thesis of Malloy, Mitchell, Block, Quas, and Goodman is that the more information courts have on which to base their decision, the sounder that decision will be, a proposition with which few would argue. Although formal impediments to children testifying in common-law jurisdictions have been removed within the last 30 years, Malloy et al. argue that many of the practices and procedures in some jurisdictions create remain impediments through the stress they create for children. This stress in turn impairs the quantity and quality of the children's evidence and may also have short-term and long-term consequences for their well-being. Malloy et al. outline a number of techniques to alleviate the distress of children required to give evidence in court, as well as techniques likely to improve the accuracy and completeness of children's recall. These techniques include the videotaped interview of the child substituted for the child's live testimony or at least substituted for the child's direct evidence. In the former case, the videotaped evidence would be hearsay evidence, as the truth of the child's statements cannot be tested in court; in the latter case, the videotaped interview would not be hearsay, as the child would still appear in court and be subjected to cross-examination. A second technique noted by Malloy et al. is the use of closed-circuit television. Closedcircuit television makes it possible for children to testify in a room outside the courtroom in a more user-friendly environment. In some jurisdictions, for example, the United Kingdom and Australia, legislation has been enacted to require closed-circuit television is available for child witnesses. In other jurisdictions, such as the United States, children are allowed to give their evidence via closed-circuit television only in quite restricted circumstances. There appears to be opposition to its use by both prosecutors and defense lawyers. Malloy et al. also propose that a support person be available while the child is testifying. Support persons are already permitted and available in many jurisdictions, and this policy has received general acceptance. One proposal that has engendered considerable controversy is to have a parent, teacher, or person trained in interviewing children substitute for the child and to testify about the disclosures the child has made to them. This approach has found greater acceptance in inquisitorial systems than in commonlaw systems. There are two objections to this procedure. The first objection is that the evidence provided to the court by the substitute is the interpretation by the substitute of what the child has said. The second objection is that having a substitute testify as to

23.

RELEVANCE OF EYEWITNESS RESEARCH

659

what the child has said offends the hearsay rule in that the truth of what the child is re, ported to have said cannot be tested in court. One other proposal of Malloy et al. is the establishment of more Child Advocacy Centres, whose mission would be to coordinate child abuse investigations. The implementation of this proposal would, on the positive side, ensure that highly skilled persons interviewed the child. On the negative side, the risk of interviewer bias becomes greater if there is an assumption that children brought to an Advocacy Centre have been abused. It is important that the two issues of the trial of the defendant accused of abusing the child and the protection of the child are not confused. In a criminal trial, because it is the liberty of the defendant that is at stake, the defendant's rights must have priority over the needs of witnesses. In child abuse cases the needs of the child may become an issue if the defendant is acquitted through lack of evidence and the defendant returns to the home of the child, placing the child at risk. If there are concerns about the safety of the child, then action can be taken to protect the child under the provisions of relevant child protection acts. In child protection matters, it is the child's interests that are para, mount. The standard of proof required by the court to satisfy itself that the child is at risk is lower than the standard of proof in a criminal trial, and the rules of evidence against hearsay evidence are more relaxed. It is important to recognize that in a matter where child abuse is alleged the competing interests may be between the interests of the child and the interests of society and individuals who want to gain a conviction. If it is the case, as Malloy et al. assert, that requiring a child to give testimony in court is distressing and can lead to shorHerm and long,term consequences, then society at large, and pros, ecutors in particular, must seriously consider whether the child should be called to give evidence and whether the criminal trial should proceed. The prosecutor should carefully weigh the prospects of a conviction against the probable distress for the child witness.

Elderly Witnesses Testimony of elderly eyewitness is the focus of three chapters in this volume. The thrust of the argument in two of these chapters, the chapters by Moulin, Thompson, Wright, and Conway and by Mueller, Johnson and Ceci, is that the basic processes underpinning complete and accurate memory have been shown to decline with age, with a consequen, tial impairment of the reliability of testimony and identification by elderly witnesses. Thus, Moulin et al. detail (1) the crucial role the frontal lobes play in remembering and (2) the diminution of frontal lobe activity in elderly persons. Mueller, Johnson and Ceci examine memory of the elderly as a function of encoding, storage and retrieval. The third chapter by LaVoie, Mertz, and Richmond argues that older adults rely heavily on gist information and are more susceptible to false memories and less able than young adults to detect post,event misinformation. In all three chapters it is noted that the el, derly are more likely to have source,monitoring problems. The pattern of recall of the elderly as outlined in all three chapters bears a marked similarity to the pattern of recall of children as described by Pipe et al. in their chapter. It raises the question as to whether the techniques proposed by Malloy et al. for the interviewing of children should be extended to the older witness.

660

THOMSON

Witnesses with Psychological Impairment The chapter by Soraci et al. examines the reliability of eyewitness testimony of people with psychological impairments. The psychological impairments include mental retardation, autism, schizophrenia, Alzheimer's disease, and dissociation. In addition, the authors of this chapter explored the effects on memory of trauma, alcohol, and marijuana. The general finding to emerge from the studies, with some exceptions, was that people with psychological impairment and people who had experienced trauma were likely to recall less, report more false information, and have greater difficulty in source monitoring. What are the implications for the criminal justice system of the findings reported in the above chapters? The findings in all of the chapters alert investigators and lawyers to sorts of memory errors made by children, elderly persons, and persons with psychological impairments. Three of the chapters offer techniques to facilitate recall. However, one very significant omission in the chapters on child witnesses is any discussion as to who is a child and relating the research to relevant legislation. While it is the case that age is operationally defined in the research reviewed by the authors, the discussion in the chapters is simply about children or sometimes about young children and older children. Findings that pertain to children aged 3 and 4 may not pertain to children aged 11 and 12 or children aged 16 and 17. The category of persons who are children is defined in various pieces of legislation. In some jurisdictions, for some purposes, a child is any person under the age of 18, but, for other purposes, a child is any person under the age of 16, and yet, for other purposes, a child is any person under the age of 15. Findings obtained from children aged 15 may have no legal application if the legal definition of a child is under the age of 15. Of course, it may be the case that the legally defined categories of children have no psychological basis, itself a matter worthy of research. Although researchers are now differentiating elderly by labeling age ranges (e.g., youngold, middle-old, and old-old), the category of persons who are elderly or older adults also is not specified in a legal sense by the authors of those chapters. In contrast with the category of persons who are children, there are no legal reference points as to who constitutes an older adult or an elderly person. A major limitation in the application of the findings reviewed by the authors showing that children, older persons, persons with psychological impairments, persons who have experienced trauma, and persons who have consumed alcohol or drugs are generally less reliable witnesses is that this research provides no clues as to whether a particular person will or will not be able to provide reliable evidence. Given the wide individual differences in memory accuracy found for children, older persons, and persons with psychological impairments, one cannot assume that the evidence of every child witness, every elderly witness, or every person with a psychological impairment is unreliable. No prosecutor or defense counsel would automatically exclude a key witness on the basis that the key witness was a child, an elderly person, or a person with a psychological impairment. However, based on the research findings, prosecutors and defense lawyers should cast around for an expert to testify about these findings and thus sensitize jurors to the problems.

23.

RELEVANCE OF EYEWITNESS RESEARCH

661

CONFESSIONS AND ALIBIS

In a criminal trial evidence that the accused person has confessed to the offenses for which he or she has been charged is perhaps the most powerful evidence the prosecutor can assemble. Confessional evidence on its own is enough to lead to the accused person being convicted. It is a presumption in common law that no person would voluntarily confess to something he or she did not do. Where an accused person has confessed to committing the offenses with which he or she is charged, the avenues of appeal are very limited for the person who has confessed to the offense. Because confessional evidence is so damning, confessions obtained under coercion or confessions obtained by inducements are inadmissible in most common-law jurisdictions. Alibis are an important line of defense for the accused person. If the accused person can establish that he or she was somewhere else at the time the offense was committed, then, in theory at least, the accused person would not be convicted for that offense. Thus, in some ways, alibis are to defendants as confessions are to prosecutors.

Confessions Kassin's chapter on false confessions is a very important one. Although there may be broad agreement that physical torture constitutes coercion, there is less consensus about such matters as sleep, food, and water deprivation, and almost no recognition of other psychological coercion. Similarly, although there may be some consensus about blatant inducement, such as the offer of money, sex, or reduction of charges, there is little agreement (nor understanding) about the role of psychological inducement. Kassin distinguishes three types of false confession: voluntary, coerced-compliant, and coerced-internalized. Voluntary false confessions are those offered to police without external pressure, and coerced-external false confessions are those provided by the suspect to avoid harm to him/herself or to members of his/her family or induced through reward. It must be noted at this point that the same classification can be applied to valid confessions. The focus of the Kassin chapter is coerced-internalized false confessions. Five components of coerced-internalized false confessions are identified by Kassin: a vulnerable suspect, the suspect is confronted with false incriminating evidence, the suspect is encouraged to believe his/her memory is faulty, the suspect infers his/her culpability, and the suspect creates a plausible account of how the offense was committed. Even in cases where the coercion or inducement has been blatant, the accused person has an evidentiary problem in attempting to establish that the confession was coerced and it is false. However, the problem is even more acute when the confession is a coercedinternalized one, a state of affairs noted by Davis and Friedman in their chapter. Often all the accused person can do is deny the claim made by the investigating police officer that the confession was made voluntarily. Kassin notes that U.S. courts have refused to exclude evidence of coerced-internalized confessions. The same position with respect to the exclusion of coerced-internalized confessions is held by courts in many other commonlaw jurisdictions.

662

THOMSON

Based on his review of the role that confessions have played in a number of cases in the United States and his analysis of the social and cognitive factors that lead an innocent person believing that he/she committed an offense, Kassin identifies two policy implications. The first policy implication is that confession evidence obtained by police by lying to the accused person that they have evidence inculpating the accused person should be inadmissible. The second policy implication is that the interrogation of the accused by the police should be videotaped. Kassin notes that videotaping would allow judges and jurors to evaluate the tenor of the formal interrogation, whether or not physical or psychological coercive tactics or physical or psychological inducement was used. It would also allow the judge or juror to evaluate whether the details contained in the confession match the evidence heard in the trial. The final point Kassin makes is that the videotape would allow the judge or jury to evaluate the confession on the basis of that confession containing details that could only be known by the offender. Kassin notes that only four states in the United States have videotaping requirements. Electronic recording of police interviews of the accused persons was introduced into Australian jurisdictions some years ago. Initially, videotaping of police interviews was opposed by the police. Today the police enthusiastically embrace the practice, as it has gone a long way to eliminating allegations of coercion and "verballing." One consequence of the videotaping is that the standard of interviewing presentation has become quite polished. Indeed, it could be argued that the interviewing police have in mind the judge and the jury when they are doing the interrogation. My concern as a defense lawyer is that the interrogations are done so professionally that few ask the question as to what went on before the videotape commenced rolling. If the lying to the accused person or if psychological coercion or inducement occurs, but does so before the tape starts rolling, videotaping the formal interview will make the tasks of either convincing the judge that the confessional evidence was coerced and therefore should not be admitted or convincing the jury that the confessional evidence should be disregarded much harder. As Kassin points out, video recording should not just occur when the suspect is providing his/her detailed confession; rather every time the suspect is interviewed, that interaction should be recorded. Failure to video-record all interviews should be grounds for excluding evidence of the confession.

Alibis Within the community at large and within the legal society in particular, the alibi defense evokes negative reactions; indeed, it would seem to have a status similar to that of malingering for persons claiming compensation for injuries sustained at work: both are seen as self-serving. In their chapter Burke, Turtle, and Olson provide a useful analysis of alibis. The first requirement of a credible alibi is for suspects to be able to recall where they were at that time, what they were doing and, who else was there. The second requirement is for corroboration of that alibi. If the corroboration takes the form of other persons attesting to the fact that the suspect was with them, those other persons will also have to remember where they were at that time, whether they remember the suspect

23.

RELEVANCE OF EYEWITNESS RESEARCH

663

also being there. Burke et al. point out that the longer the delay, the greater the difficulty both for the suspect and the persons corroborating the alibi. The third requirement for a credible alibi, if other persons are providing corroboration of that alibi, is that these other persons be seen as credible corroborators. Burke et al. note that frequently it is the case that either there was no other person to corroborate the claim by the suspect that he/she was somewhere else when the offense was committed or that person has some relationship to the suspect. The important contribution of the chapter by Burke et al. is that it identifies an area of research that has been largely overlooked. Basic research establishing the strategies people employ when asked to recall where they were at a designated time, what they were doing, and who else was there may provide investigators, lawyers, judges, and jurors a basis for evaluating the content of an accused person's alibi. Ultimately, it may be important to ensure interviews with suspects who claim an alibi are videotaped. Both Kassin and Burke et al. note that findings from extensive eyewitness identification research have led to systemic changes in identification procedures conducted by police in the United States. The impetus for significant changes in the way courts receive confession evidence and alibis awaits the findings from research and these findings to be disseminated widely to lawmakers, lawyers, and judges.

MEMORY AIDS IN THE INVESTIGATORY STAGE

Two techniques that have been employed by investigators in an attempt to enhance witnesses' memories are hypnosis and the cognitive interview. The chapter by Mazzoni, Lynn, and Kirsch examines hypnosis; the other chapter by Fisher and Schreiber is primarily concerned with the cognitive interview. Any evaluation of the usefulness of a technique employed to facilitate memory ultimately depends on the objective of the investigator, a point implicit in Fisher and Schreiber's discussion on measuring accuracy of memory. A technique that increases the number of false details reported by witnesses will increase the chance that the investigation will stray. However, that cost may be one that should be borne if the technique not only increases the reporting of false details, but it also elicits additional accurate information. The greater the number of leads the investigator has, the more likely it is that the investigation will be successfully concluded.

Hypnosis After reviewing the literature, Mazzoni et al. identify three issues in the use of hypnosis as an investigatory technique. The first issue is that the increase in details reported by hypnotized persons does not represent a "genuine enhancement" of recall, as many of the new details reported are inaccurate. The second issue is that hypnotized persons are vulnerable to suggestive utterances of the hypnotist, resulting in the creation of false memories. The third issue is that hypnotized witnesses are more confident about what

664

THOMSON

they report, whether or not the details are accurate or inaccurate. Similar issues have been commented upon in the chapter on false memories by Neuschatz, Lampinen, Toglia, Payne, and Cisneros. The concern of Mazzoni et al. about hypnosis as an investigatory technique is that the content of what a witness has reported while under hypnosis is likely to form the basis of his or her court testimony. Because of the great conviction witnesses who have been hypnotized have about the accuracy of their evidence, it is likely to be accorded greater weight than it should by jurors. Establishing during cross~examination that the content of the testimony was based on what witnesses reported under hypnosis with the objective of making jurors sceptical of that evidence is unlikely to succeed; in~ deed, it may result in that evidence being given greater weight. Nor is it likely that in~ structions by the judge warning the jurors of the dangers of evidence originally obtained under hypnosis will result in jurors assigning lesser weight to that evidence. Because the prejudicial impact of testimony elicited through hypnosis during the investigation is likely to far outweigh its probative value and because the methods courts have of ad~ dressing this impact are ineffective, in many jurisdictions evidence that has been ob~ tained under hypnosis is held to be inadmissible. There is one other significant issue relevant to the use of hypnosis that is not dis~ cussed by Mazzoni et al. Witnesses who deliberately give false statements in hypnotic interviews, if confronted with the falsity of their account, can disown that account by asserting that what was disclosed under hypnosis was in fact a product of the hypnosis, that the false statement must have been planted in their mind by the hypnotist, that they had no intention of misleading the court, and therefore that they could not be held responsible for the false information.

Cognitive Interview The cognitive interview, an investigatory technique developed by Fisher and his associ~ ates, has been adopted by many investigation agencies. The current version of the cog~ nitive interview has been designed to take into account cognitive, social, and communi~ cation processes. For each of these processes, Fisher and his associates have formulated procedures. Thus, the procedures relevant to the cognitive processes include context re~ instatement, allowing witnesses to recall everything about the target event that is salient to the witnesses before attempting to direct witnesses' recall to conform to the investi~ gator's conception of what is relevant, varied retrieval, admonishment against guessing, and minimizing the opportunity for witnesses' recall to be contaminated by other exter~ nal sources. The social processes include establishing rapport and providing the oppor~ tunity for active witness participation. The communication process includes procedures that allow the investigators to convey their investigative needs to witnesses and which allow the witnesses to tell their stories. Close examination of these procedures suggests that a potential conflict between the various procedures and how an investigator con~ ducts an interview will depend on the weight given to any particular procedure. Thus on the one hand, investigators are to attempt to minimize witnesses' reconstruction of

23.

RELEVANCE OF EYEWITNESS RESEARCH

665

their memories; on the other hand, witnesses are encouraged to think about the event "in many different ways." Furthermore, what is required in court is to determine what the particular witness experienced at the time the target event occurred, not what that witness would have experienced had that witness been somebody else, as was the case in the Anderson and Pichert (1978) study. In the Anderson and Pichert study university students read a story from the perspective of a burglar or a home purchaser. They then recalled the content of that story from the perspective they had when they read the story. They then recalled again, switching their perspective from burglar to home pur~ chaser, or from home purchaser to burglar. Although it may be important for the investigator to convey to witnesses the inves~ tigator's objectives to minimize witnesses' withholding of valuable information and to minimize irrelevant information, there is the danger of witnesses editing their reports to fit the expectations of the investigators. Instructions to witnesses not to guess make sense if it is assumed that guesses are random or are based on irrelevant considerations. However, if guesses are based on partially relevant information, then they may provide the investigator with useful material. Rapport has been identified as an essential compo~ nent of the cognitive interview. It is worrying that methods employed in rapport building have also been implicated as establishing circumstances conducive to memory distortion through suggestive utterances (e.g., Ceci & Bruck, 1993) One important matter canvassed by Fisher and Schreiber is the limitation of the cognitive interview. Fisher and Schreiber identify two limitations, namely, the cognitive interview does little to improve eyewitness identification if the offender is present in the lineup, and the cognitive interview cannot reverse the corruption of witnesses' memo~ ries caused by earlier exposure to misleading information. This latter limitation has sub~ stantial significance both for investigators and for defense counsel. What it means is that any account given by witnesses in a police interview, even with cognitive interviews, should not be taken at its face value until and unless it can be established that the wit~ ness has not been exposed to prior suggestive utterances. The likelihood that witnesses have not been exposed to suggestive utterances is not high, particularly when the wit~ nesses are children. In many jurisdictions a "disclosure" interview precedes the video~ taped interview. One implication for defense attorneys is that they will need to establish the context of the videotaped interview, such as what questioning of the child occurred prior to the interview. Fisher and Schreiber also explore in their chapter the important question of what police investigators actually do when interviewing witnesses. Findings from a recent study they carried out suggest that police practice in interviewing has changed little in the past 20 years: police officers frequently interrupted witnesses when witnesses were ended questions were the exception rather than the giving their account of events, open~ rule, and many of the police questions were suggestive and misleading. Lamb (2000) found that the perceptions of police of their interviewing were quite inaccurate; the po~ lice officers claimed that most of their questions were open~ended when in fact they were closed questions. These findings can only reinforce concerns about the effect of "disclo~ sure" interviews by police on witnesses' reports in subsequent videotaped interviews.

666

THOMSON

FACTORS THAT MAY AFFECT JURORS' DECISIONS IN CHILD SEXUAL ABUSE CASES

In their chapter Bottoms, Golding, Stevenson, Wiley, and Yozwiak have comprehensively reviwed findings of studies that have examined factors that may affect jurors' decisions in child sexual abuse cases. Thus studies have explored whether characteristics of the alleged perpetrator, the child victim, the juror, and the circumstances surrounding the disclosure influence jurors' decisions. The characteristics examined include gender, race, and age. The circumstances include those surrounding the disclosure of the sexual abuse, type of abuse, and courtroom and trial factors. Given the widely divergent findings about gender and, to a lesser extent, race of jurors, victim, and defendant, an inescapable con, elusion to be drawn is that the particular facts of the vignettes/transcripts employed in the different studies and/or the manner in which these vignettes were presented inter, acted with the gender and race of the juror, victim, and defendant. For example, Bottoms et al. note that some studies have found that female jurors favor female perpetrators and female victims, whereas male jurors favor male perpetrators and male victims. Other studies have found that the decisions of male jurors were different, depending on the gender of the victim, but gender of the victim did not influence female jurors. In contrast to the characteristics of juror, victim, and defendant, there are clearer patterns of the effect of the other factors reviewed by Bottoms et al., for example, age of victim, prior criminal history of defendant, manner in which child witness evidence was presented in court, and disclosure characteristics. However, some of these factors had an impact on measures of credibility but not on the verdict. More importantly, it is not clear what implications the findings about the factors shown to influence decisions whether a child victim has been sexually abused have for the legal system or social policy. Whereas in eyewitness research, the researcher can assess the effect of the experimental manipu, lation against what the eyewitness actually saw, with studies examining juror decisions, there is no right or wrong decision. The findings reviewed in this chapter have clear im, plications for how to increase or decrease guilty decisions, but the findings are silent about whether the various factors produce a more just outcome.

CREDIBILITY

The course of an investigation and the outcome of a trial are determined in large mea, sure by the credibility of the witnesses. Credibility of witnesses may be assessed by the witnesses' truthfulness, their competence, and their confidence. Two of these factors, their truthfulness and their confidence, were the subject of chapters in this volume.

Truthfulness Griesel and Yuille have described in their chapter a range of techniques and procedures that have been developed to detect deception. These techniques include the polygraph, nonverbal behaviors, and analyses of statements.

23.

RELEVANCE OF EYEWITNESS RESEARCH

667

With the exception of Japan, findings or conclusions from polygraphs are not admissible. In common-law jurisdictions, evidence of an expert about the observations of a witness's nonverbal behaviors would also be inadmissible. This position is probably well founded, given the findings of DePaulo (1992, 1994) that there is no exclusive relationship between an internal state and a specific nonverbal behavior. Although the expert's opinion is excluded, it is very likely that judges and jurors place great reliance on nonverbal behavior, as illustrated by a recent decision in an Australian court: Whilst the material included in her affidavit gives the appearance of being quite unequivocal, this appearance did not survive her cross-examination. I do not accept significant parts of her evidence. It was internally inconsistent in a number of respects. She was argumentative at times, evasive at others. She did not present well in the witness box. On a number of occasions I formed the view that the evidence she gave was made up on the spot to get her out of what she perceived to be a problem. Frequently she paused for periods, some of which were extended, and appeared to be casting around for an answer which she regarded as suitable rather than addressing the question directly.

Expert opinion based on an analysis of the statements is admissible in many European and Scandinavian courts indeed, for cases of allegations of child sexual abuse such an analysis is required. In common-law jurisdictions, having an expert inform a jury of his/her conclusions about the credibility statements of a witness made outside the court are likely to be excluded on the basis of the common-law rule if that witness is an adult. With allegations of child sex abuse it is likely that any expert opinion would be determined on a case-by-case approach. However, techniques and procedures to detect deception are used extensively in criminal investigations by investigators who have received formal training and by investigators who have received no formal training based on their intuition. What is a matter of concern is training programs that disseminate inaccurate information. It is also important that these techniques not be used by investigators to coerce confessions from suspects by providing false information about the findings.

Confidence A truism around the courts is that the impact a witness has on a jury is not just what witnesses say but how they say it. This view is shared by the U.S. Supreme Court, at least with respect to eyewitness identification. In Neil v. Biggers (1972) the U.S. Supreme Court identified the level of certainty a witness had in his/her identification of the defendant as the offender. In England, the Devlin Inquiry (1976) had come to quite the opposite conclusion. Juries there were to be warned about putting too much weight on the confidence of eyewitnesses in their identification of the defendant. Based on the review of research findings of Shaw, McClure, and Dykstra, it would appear to be the case that the Devlin Inquiry got it right and the U.S. Supreme Court got it wrong. In essence Shaw et al. found that at best there is no relationship between confidence and accuracy; at worst, they are negatively correlated.

668

THOMSON

CONCLUDING REMARKS

What are the important lessons to be learned from the research findings reviewed in this volume? Perhaps the most important thing that emerges from these findings is the many hidden factors identified by researchers that have an impact on the outcome of a criminal trial. These factors would include psychological coercion that courts in common-law jurisdictions have refused to acknowledge; they also include the limitations and frailties of the major participants involved in the criminal justice system, the witnesses, the investigators, the lawyers, and the judges. For too long, insufficient attention has been paid by lawyers and judges to the difficulty in establishing alibis. What is interesting is that many of the shortcomings identified in the chapters can be addressed by technology. All interviews with victims, suspects, and other witnesses should be electronically recorded, and that recording should constitute evidence before the court. Unless there are good reasons for not doing so, the absence of recording should be grounds for excluding that interview as evidence. In that way courts would be better placed, or, said another way, they would have a better data base on which to make their decision. Finally, based on the research findings reviewed in this volume, the question of whether closed-circuit television should be available for elderly witnesses as well as for children should be pursued. REFERENCES Anderson, R. C., & Pichert, J. W. (1978). Recall of previously unrecallable information following a shift in perspective. Journal of Verbal Learning and Verbal Behavior, 17(1), 1-12. Bruck, M., Ceci, S. J., & Melnyk, L. (1999). The effect of interviewer bias on the accuracy of children's reports and interviewers' reports. Paper presented at the Biennial Meeting of the Society for Child Development, Albuquerque, NM. Ceci, S. J., & Bruck, M. (1993). The suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113 (3), 403-439. DePaulo, B. M. (1992). Nonverbal behaviour and self-presentation. Psychological Bulletin, Ill (2), 203-243. DePaulo, B. M. (1994). Spotting lies: Can humans learn to do better? Current Directions in Psychological Science, 3(3), 83-86. Devlin, Rt. Hon. Lord (1976). Report to the Secretary of State for the Home Department of the Departmental Committee on Evidence of Identification in Criminal Cases. London: Her Majesty's Stationery Office. Neil v. Biggers, 409 U.S. 188, 34L.Ed.2d401 (1972), cert. denied,444 U.S. 909, 100 S.Ct.221, 62 L.Ed.2d 144(1979). S v. The Queen, [1989] HCA 66, (1989) CLR 266 F.C. (21 December 1989), available at www.austli.edu.au/au/cases/cth/HCA/1989 /66.html/ State Line Rail Authority of NSW v. Earthline Constructions Pty Ltd and ORS, S8/1997 (13 March 1998), available from www.austlii.edu.au Thompson, W. C., Clarke-Stewart, K. A., & Lepore, S. (1997). What did the janitor do? Suggestive interviewing and the accuracy of children's accounts. Law and Human Behavior, 21 (4), 405-426.

Author Index

Numbers in italics indicate pages with complete bibliographic information.

A

Abbot, W. F., 208, 224 Abelson, R., 430, 444 Ahlin, D. S., 115, 425, 573 Abrams, S., 183, 192 Ackerman, B. P., 482, 501,503 Ackil, ]. K., 385,397, 406,421, 442,444, 501,507 Adam, S., 457,471 Adams, D., 291 Adams, M., 263, 294 Adams, M. ]., 86, 114 Adams-Price, C., 375, 393, 587,597 Adaval, R., 26, 52, 197, 221, 237 Aderman, D., 526 534 Adlaf, E., 282, 293 Adler, S. A., 214,234 Adolphs, R., 85, 98, 111, 112 Akehurst, L., 255,260, 351,357,361,362,365, 369, 373, 374,381,393 Akrami, N., 203,224 Alain, M., 198, 236 Albert, G., 472 571 Albin, R. S., 513 535 Aldridge, J., 444, 558, 569 Aleman, A., 270, 290 Alexander, K. W., 87, 112, 122, 149,460,466,472, 476, 498,505,550,571,573 Alexander, M. C., 520, 537 Allan, K., 184, 190, 209,228, 249,257, 589,599 Allen, J. ]. B., 103, 114 Allen, L.A., 511, 519,534 Alley, T. R., 376, 393

Allison, M., 587,598 Allport, G. W., 11, 14, 40, 42, 201 Allwood, C. M., 211,224, 383, 391, 392,394 Aloia, M., 274, 293 Alpert, J. L., 140, 146 Altmann, E. M., 133, 146 Amadar, Z., 106, 115 Amador, M., 66, 67 Amador, X., 270,291 Aman, C., 355, 366, 408, 423, 441, 446, 538 Amato, S. L., 345, 353, 367, 369 Ambadar, Z., 301,319 Ambady, N., 198,234 Anas, A., 219,225 Anderson, B., 580,597 Anderson, D. E., 363, 365 Anderson, G., 319, 602 Anderson, J. R., 583, 604 Anderson, M. C., 219, 222, 224, 302, 303,304, 316, 636,643 Anderson, N.D., 581,597 Anderson, R. C., 59, 76, 665, 668 Anderson, S., 112 Anderson, S. ]., 120, 146 Anderton-Brown, C., 632,645 Andreassen, 0., 554, 572 Andresen, B., 270, 294 Andrews, B., 125, 147, 316 Anes, M. D., 605, 624 Angell, K. E., 217,234,583,602,610,624 Anis, A., 184, 189 Ansfield, M. E., 363, 365 Arasteh, K., 288, 293

669

670

Araya, T., 224 Arbuthnott, D. W., 216, 224 Arbuthnott, K. D., 216,224 Archer, R. L., 526, 534 Arenberg, D., 582, 599 Arkes, H. R., 219,224, 229 Armilio, M. L., 591, 603 Arnold, M. C., 271,294 Aronson, L., 120, 124, 153, 318, 335 Arps, K., 43 Arrigo, J. M., 93, 111 Asch, S. E., 211, 224 Aschermann, E., 353, 367 Ashworth, A., 322, 336 Askari, N. E. A., 204, 233 Assefi, S. L., 214, 224, 288, 291 Astington, J. W., 556, 569 Atchley, R. C., 577, 598 Atkinson, G., 277, 296 Attias, R., 513, 534 Aubrey, M., 536 Auman, C., 591,599 Austin, J. L., 16, 43 Autry, M. W., 28, 34, 50 Avaltroni, )., 570 Axelrod, S., 16, 42 Ayers, M.S., 214, 224

B Bachman, R. B., 600 Back, S., 511,514, 515,519,534 Baddeley, A. D., 59, 76, 113, 292, 295, 644 Bahrick, H. P., 119, 123, 147 Bahrick, L. E., 93, 111, 122, 123, 147, 150, 152, 462, 469,

471,475 Bahrick, P. 0., 119, 147 Bailey, D., 149 Bailey, S. E., 127, 150, 204,229 Baker, S. J., 521, 541 Baker-Ward, L., 105, 114, 118, 126, 148, 152, 405,406, 407,408,421,423,425,433,446,454,455,456,461, 462,466,467,469,470,472,475 Balch, W. R., 104, 111 Baldwin, C. L., 580, 602 Ball, C., 569, 569 Ball, P.J., 511,525,536 Ballard, P. B., 305,316 Balota, D. A., 243,260, 274, 275,291,582,598, 606, 622,628,629,643 Baltes, P. B., 629, 644 Banaji, M., 35, 48, 208,231 Banaszynski, T. L., 202, 227 Banipal, R., 141, 147 Banse, R., 45, 228 Barabasz, A., 379, 394 Barabasz, M., 379, 394

AUTHOR INDEX

Baradaran, L. P., 432, 450 Baranoski, M., 114 Barchet, P., 319 Barclay, C. R., 133, 147 Bargh, J. A., 198, 225 Bar-Hillel, M., 342, 365 Barker, C. H., 178, 192 Barnett, D., 540 Barnett, G. 0., 15, 47 Barnett, M.A., 514,534 Barnhardt, T. M., 307, 318 Barnier, A.]., 219,225 Barnsley, R. H., 119,153 Baron, L., 517, 537 Barr, D. J., 14,42 Barr, R. J., 108, 111, 442, 444 Barrett, A.M., 291 Barrett, H. M., 95, 114,310,318 Barrett, L. F., 207,226 Barsalou, L.W., 59, 79 Barth, R. P., 550, 570 Barthel, J., 179, 189 Bartlett, F. C., 11, 14, 40, 42, 60, 76, 123, 147, 240, 241, 251,257,481,504 Bartlett, J. C., 45, 584, 585, 598, 601, 602, 614,622, 628, 633,636,637,645 Bartol, A. M., 53, 76 Bartol, C. R., 53, 76 Basden, B. H., 248, 249, 257 Basden, D. R., 248, 257 Bass, B. A., 517,541 Bass, E., 322, 335 Bates, A. F., 605, 624 Bates, E., 23, 26, 42, 47 Bates, M. E., 284, 285, 295, 296 Batt, J., 208, 224 Batterman-Faunce, J. M., 150,412,414,422,423, 449, 455,461,472,510,538,540,559,565,571,573 Bauer, A., 416,425,455,551,573 Bauer, P. J., 150,416,458,469 Bauserman, R., 516,541 Baxter, J. C., 515,540 Baxter, J. S., 437,446 Bayen, U. J., 34, 42, 50, 205, 225, 236, 294 Bean, P., 282,291 Beardsall, L., 515, 538 Beattie, A. E., 168, 173 Beavin, J. H., 17, 51 Beck, M., 385, 397, 501 Becker, D. V., 202, 232 Beckman, H. B., 37, 42, 48 Becquart, E., 202, 226 Beer, J., 466, 472 Begg, I. M., 184, 189, 219, 225 Behrman, B. W., 89, 111, 373, 393 Beidleman, W. B., 379, 397 Bekerian, D. A., 59, 63, 76, 441,444

AUTHOR INDEX

Beling, J., 513, 534 461, 462 Bell, M., 405, 407,534 475 425, 455, 461, 462, Bell, R. L., 305, 319 Belli, M., 526, 534 Belli, R. F., 120, 147, 178, 189, 583,598 Bellucci, S., 263, 296 Belmont, J. M., 262, 291 Belsky, J., 467,469 Bern, D. J., 182, 183, 189 Benda, B. B., 282, 292 Bendiksen, M., 301,319 Benjamin, A. S., 629, 644 Bennetto, L., 266, 269, 291 Benoit, G., 332, 337 Benoit, P. J., 15, 24, 30, 41,42 Benoit, W. L., 15, 24, 30, 41,42 Ben-Shakkar, G., 342,343,345, 347, 365, 369 Bentley, B. J., 513, 536 Berg, R. F., 550, 574 Bergman, E. T., 152, 209,234 249, 260 Berkowitz, R., 173, 319 Berliner, L., 250, 257, 540 Berman, G. L., 136, 147 Bernheim, H., 323, 327,335 Bernstein, B. E., 562, 569 Bernstein, D. M., 215,231, 243,258 Bernstein, E., 276, 291 Bernsten, D., 135, 147 Berresheim, A., 53, 54, 76 Bessenoff, G. R., 205,235 Betz,A. L., 121,155,162,173 Betz, S., 209, 225 Beuke, C. J., 216,235 Beversdorf, D. Q., 267,291 Bibi, U., 126, 154 Bidrose, S., 115, 118, 126, 147, 407,425, 573 Bifulco, A., 141, 147 Billings, J. F., 184, 190, 243,258, 277,293 Binet, A., 240,257,402,411,421 Bink, M. L., 35, 42 Birch, R. A., 432,451 Birnbaum, I. M., 306, 317 Birrell, P., 202, 234 Birt, A. R., 135, 146, 153, 184, 192, 214, 233, 339, 348, 362,368 Bishop, G. D., 531,540 Bjoerkman, M., 206, 237 Bjork, E. L., 224, 302, 303, 316, 636, 643 Bjork, R. A., 104, 115,222,224,235,302,303,306,316, 320,373,393,636,643 Bjorklund, B., 454, 469 Bjorklund, D. F., 406,421, 442,444, 454,455, 456,469, 470,487,489,504,548,569 Bjornsson, G., 186, 190 Black, J. B., 199, 225, 605, 622 Blackford, C., 413, 426 Blackwell, J. M., 243, 259, 502, 505

671

Blackwell, K. T., 513,536 Blades, M., 268,292, 425, 433,441,449,474,470, 557, 574 Blagrove, M., 186, 189, 373,374, 381, 393 Blair, I. V., 200, 201,230 Blanchard, L., 46, 585, 602, 603, 605, 606, 612, 624, 633,646 Blank, A., 26,47 Blank, H., 163,171,214,225 Block, S., 550, 569 Blois, W. 0., 216,236, 253, 260 Bluck, S., 96, 113, 139, 147, 463, 469 Blue, J., 462,470 Blume, E. S., 282, 322, 335 Bly, B. J., 284, 295 Boals, A., 213,230 Boat, B. W., 408,424,434, 437,438,447,464,471, 473, 558,571 Boatwright, L. K., 582, 603 Bock, J. K., 13, 31, 42 Bodenhausen, G. V., 35, 48, 591, 598 Boehm, L. E., 219, 224 Boller, K., 214,234 Bolls, P. D., 29, 42 Bolton, J. F. G., 514,515,534 Bomba, E., 233, 602 Bonacci, A. M., 28, 43 Bonanno, G. A., 29, 42 Bond, N., 129, 147, 167, 172 Boney-McCoy, S., 529, 542 Boon, J., 72, 76, 120, 150, 417, 422, 436,446 Borgida, E., 213,230, 434,446, 511,513,519,522,525, 528,529,535,539,542 Boriskin, J. A., 520, 537 Borkowski, J. G., 262,291 Bornstein, B. H., 97, 112, 120, 136, 138, 147, 149,463, 469,531,535,588,598,615,617,618,622,637,643 Bornstein, G. K., 65, 79 Borowiecki, J. J. III., 202, 233 Borza, M.A., 214,234 Bothwell, R.K., 324, 325, 327, 337, 338, 371, 372,373, 374,379,392,393,394,396 Bottoms, B. L., 408,416,421,422,423, 433,441,442, 444,446,510,511,512,513,514,515,517,518,519, 520,521,523,524,526,527,528,530,531,533,535, 536,53~538,541,542,556,571

Bottrill, K. V., 214,230 Boucher, J., 266, 268,291,292 Boudewyns, P. A., 92, 114 Bove, C., 602 Bowen, C. J., 560, 570 Bower, G. H., 35, 48, 67, 76, 172, 199, 218,225, 228, 303,306,316,592,605,622,644 Bowers, K., 379, 394 Bowler, L., 444, 569 Bowman, K., 104, 111 Boyce, W.T.,416,425,551,573

672

Boychuk, T. D., 356, 367 Boyd, C. E., 637, 646 Bradfield, A. L., 211, 237, 338, 372, 382, 385,386, 392, 393,397,587,598 Bradshaw, J. M., 214, 227, 442, 445 Bradshaw, T. L., 520, 527, 535 Brady, M.S., 256,257, 410,411,421 Brainerd, C. J., 118, 127, 147, 153, 214,225, 256,257, 277,295,314,316,437,442,444,457,458,470,479, 480,481,483,484,485,486,487,488,489,490,491, 492,493,495,496,497,498,499,500,501,502,504, 505,506,612,622 Brandstatter, H., 236 Bransford, J.D., 241, 242,257, 481,486,495,504, 605,622 Braun, B. G., 316 Braun, K. A., 214,225 Bray, N. W., 262,291,296 Brebion, G., 270, 274, 291 Bredahl, L. C., 112 Bredart, S., 202, 226 Bregman, N.J., 214,225 Brekke, N., 511, 513, 519,535 Bremner, J.D., 280,291 Brennan, K. H., 63, 77, 592, 599 Brennan, M., 434, 444 Brennan, R. E., 434,444 Breukelen, G. J.P., 119, 154 Brewen, C. R., 316 Brewer, D. D., 515,535 Brewer, J., 85, 112 Brewer, K. D., 515, 517, 520, 527,535 Brewer, M. B., 34, 42, 45, 198, 228 Brewer, N., 125, 127, 129, 136, 139, 147, 155, 167, 172, 235, 236 Brewer, W. F., 13, 26, 28, 31, 34, 42, 43, 204,211,225, 242,248,257,458,470 Brewin, C. R., 125, 135, 147, 148 Briggs, P., 442, 449 Brigham, J. C., 327, 337, 353, 354, 355, 356, 357, 360, 367,368,371,372,376,387,392,393,395,511,516, 519,535,539,541,586,598,601 Brigidi, B. D., 280, 297 Brimacombe, C. A. E., 174,587,588,596,598,616,622, 627,633,635,641,643 Britton, W. B., 82, 114 Brock, P., 67, 77 Brock, T. C., 196, 197, 199, 216, 228 Broder, A., 32, 49 Brokate, B., 282, 293 Broussard, S.D., 513, 516, 518, 523, 526,537 Brown, A., 306 Brown, A. L., 482, 504 Brown, A. S., 33, 34, 35, 40, 43, 607, 623 Brown, C. L., 65, 76 Brown, D., 129, 135, 142, 148,307,316,324,327, 328, 331,335,466,469

AUTHOR INDEX

Brown, D. A., 74, 76, Ill Brown, E., 373, 393 Brown, G. W., 141, 147 Brown, J. M., 28,43 Brown, L. S., 146 Brown, M. H., 22,51 Brown, N. R., 132, 151 Brown, P., 17, 43 Brown,R.,29,37,43,433,444,455,469 Brown, T., 462,471 Brownmiller, S., 519, 535 Brown-Whistler, S., 582,603 Bruant, A., 270, 292 Bruce, D., 214,233 Bruce, K. R., 283, 284, 286, 291 Bruce, V., 225 Bruck, M., 39, 43, 60, 73, 74, 76, 108, 111, 136, 148, 178, 184,189,214,225,241,257,355,360,365,401,402, 404,405,406,407,408,409,410,411,412,413,414, 417,418,421,422,424,429,441,442,444,458,461, 464,466,470,474,492,501,504,558,559,563,570, 652,665,668 Bruhn, A. R., 326, 335 Brustrom, J. E., 583,598 Bryant, P., 466, 476 Bryant-Brown, L., 406,423 Bryne, C. A., 135, 148 Buchanan, T. W., 85, 98, Ill, 112, 551,570 Buchner, A., 206, 227 Buck, J. A., 39, 51, 529, 535, 537, 541, 563,574 Buckhout, R., 91, 113 Buckley, J.P., 177, 180, 190, 223,230 Budesheim, T. L., 26, 27, 52 Budson, A. E., 274, 275,276,291, 611,623 Buehler, R., 36, 43 Buffalo, E. A., 84, 115, 131, 154 Bugental, D. B., 462, 470 Bulkley, J., 509, 538, 548,571 Bull, R., 64, 65, 69, 72, 78, 79, 84, 114, 118, 120, 131, 150,152,182,192,256,260,322,337,351,352,357, 361,362,363,365,368,369,378,395,406,422,425, 429,430,436,444,446,448,584,585,593,598,600, 601,633,637,645 Burch, M., 463 Burden, M. J., 107, 112 Burgess, C., 324, 332, 335 Burgess, C. A., 217,236, 260 Burgess, M. F., 217,236, 260 Burgess, S. L., 109, 113 Burggraf, C. S., 15, 28, 30, 50, 51 Burgwyn, E. 0., 467,469 Burgwyn-Bailes, E., 118, 148, 452,461 Burke, A., 87, 112 Burke, D. M., 306,316 Burke, T. M., 156, 161,164, 166, 167, 171, 172, 173 Burkett, A. D., 603 Burnett, C. B., 48

AUTHOR INDEX

Burnham, W. H., 323, 335 Burns, H. J., 68, 78, 178, 191,242,258,380,395,501 Burns, L., 515, 537 Burnstein, E., 26, 31, 47,49 Burt, C. D. B., 121, 148, 167, 172,215,230 Burt, M. R., 513, 535 Bushman, B., 28, 43 Bussey, K., 555, 570 Butler, E. A., 29, 43, 50 Butler, L. D., 306,317, 318 Butler, S., 407,421,464,472,557,558,570 Butterfield, E. C., 262, 291 Butterworth, J., 481, 482, 507 Buttow, P. N., 15, 48 Butz, D. A., 201, 233 Bylin, S., 222, 225, 226 Byrd, M., 581,598 Byrne, B., 85, 113

c Cabeza, R., 629, 643 Cahill, L., 85, 112 Caldwell, R. A., 524, 540 Calhoon, S. K., 375, 396, 587, 602 635, 645 Calhoun, K. S., 319 Calvini, G., 35, 48 Camp, C. J., 580,601 Campara, L. B., 74, 76, 435, 444, 549, 574 Campbell, H., 91, 113 Campbell,]., 320 Campbell, M.A., 146, 153, 339,348,368 Campbell, W. K., 21,43 Cancelli, A., 481, 482,506 Canli, T., 85, 112 Cardeiia, E., 317 Carey, P., 24, 49 Carini, B., 34, 42 Carlesimo, G. A., 263, 264, 296 Carli, L. L., 206, 225 Carlin, M. T., 262, 263, 291, 295 Carlson, E. B., 213,227, 276,278,291,292,302 Carpenter, K. M., 216, 229 Carr, C., 525, 539 Carroll, J. S., 22, 43, 202, 216, 225 Carter, A. Y., 481, 482, 483, 505 Carter, C. A., 416,421,433,442,444 Carter, J.D., 198,230 Carter-Sobell, L., 306,318 Cary, M., 469 Cashon, T. S., 286, 292 Cashmore, J., 565,570 Caspi, A., 125, 150 Cassel, W. S., 406,421, 442,444, 454,455,456,469, 470 Castel, A. D., 629, 644 Castelle, G., 208,231 Castelli, P., 511, 530,535, 571

673

Castro, A.M., 306,319 Cauffman, E., 571 Ceci, S. J., 39, 43, 60, 73, 74, 76, 78, 108, 111, 118, 136, 146, 148, 152, 153, 178, 184, 186, 189, 192, 213, 214, 225,226,232,241,257,355,360,366,368,401,402, 404,405,406,407,409,411,412,413,414,418,419, 421,422,423,424,426,429.437,440,441,442,444, 445,446,44~449,450,456,457,458,461,464,470,

474,475,476,492,501,504,536,538,539,541,558, 559,563,570,574,589,590,594,601,602,627,645, 652,659,665,668 Cederborg, C-A., 438, 445, 458, 463, 470, 474 Cegala, D. J., 24, 43 Cendan, D. L., 623 Cerrito, B. M., 491,505 Chaffin, M., 540 Chalfonte, B. L., 583, 598 Chambers, K. L., 214, 226 Chamblin, M. H., 517,537 Chan, F. M., 259 Chanowitz, B., 26,47 Chapleau, K. M., 200, 201,230 Chapman, A., 545, 570 Chapman, A. J., 75, 76 Chapman, M. V., 550, 570 Charlton, K., 44, 363, 365 Charman, S.D., 211, 237, 385, 397 Charney, D. S., 125, 154 Chase, W. G., 456, 470 Chechile, R. A., 262, 291 Chen, C.-F., 413,426 Chen, E., 567, 570 Chen, K., 549, 570 Cherry, K. E., 588, 598, 617, 622 Chetwynd, A., 630, 646 Chi, M. T. H., 450, 464 Chin, D., 67, 77 Chinman, G., 319 Chiu, C. Y., 22, 47, 220, 221,231 Choi, H., 306,308,317, 320 Christensen, T. C., 136, 153, 207,226 Christensen-Szalanski, J. J. J., 206, 226 Christianson, S.-A., 28, 43, 50, 87, 92, 112, 116, 120, 122, 129, 130, 132, 133, 134, 135, 148, 149, 150, 222, 225,226,234,551,570 Chrosniak, L. D., 583,599 Chu,J. A., 311,317 Cialdini, R. B., 28, 43, 216,229, 235 Cicchetti, D., 413,422, 491,505 Cima, M., 93, 94, 112 Ciranni, M. A., 222, 226 Clague, F., 274, 292 Claman, L., 562, 569 Clancy, S. A., 94, 114, 280, 281,292, 310,318 Clare, C. I. W., 65, 79 Clare, I. C. H., 265, 292, 293 Clark, H. H., 14, 22, 43, 220,226, 236

674

Clark, H. L., 512,514, 515,523,536,595,601, 645 Clark, M. C., 67, 76 Clark, S. L., 406, 423 Clarke-Stewart, K. A., 442, 446, 450, 652, 668 Clary, E. G., 528, 536 Clay, J., 112 Claycomb, L., 425, 433, 448 Clayton, L. D., 510,536 Clifford, B. R., 53, 54, 66, 67, 68, 77, 78, 79, 213, 214, 234,373,393,584,598 Clopton, P. L., 288, 296 Clubb, P. A., 118, 152, 405,421,440,448, 454, 458,461, 469, 4670 471, 475 Coan, J. A., 249, 258 Cockburn, A., 355,365 Cockman, D. W., 229 Coffman, K. J., 183, 190, 216, 217,229 Cohen, C. E., 205, 226 Cohen,G.,69,77, 120,146,588,598,609,617,623,633, 635,643 Cohen, J., 355, 365 Cohen, N.J., 132,151,164,173 Cohn, J. F., 349,365, 369 Cohn, M.A., 122, 148 Coifman, K., 29, 42 Colcombe, S. J., 26, 52, 197,237, 591,599 Colella, M. J., 363, 370 Colgan, B. A., 180, 189, 189 Collings, S. J., 519, 521,536 Collins, R., 61, 76 Collins, W. A., 437, 445 Compton, D. M., 263, 294 Conger, J. C., 29,47 Connelly, P. J., 159, 172 Connery, D. S., 179, 189 Connolly, D. A., 117, 118, 126, 129, 137, 140, 141, 146, 147, 148, 149, 153, 458,459,470 Connor, L. T., 634, 643, 644 Connors, E. T., 160, 172 Constantin, T., 151 Conte, R., 117,189 Conway, L. G., 209, 226 Conway, M.A., 84, 112, 121, 123, 132, 136, 148, 149, 162,163,167,172,219,225,635,643 Cook, A. E., 22, 43 Coon, V. E., 582, 583, 599 Cooney, L. M., 600 Cooper, B.S., 58, 77, 88, 112, 353, 370 Cooper, D. S., 373, 374, 394 Cooper, H., 44, 363, 365 Cooper, J., 224 Copeland, J. T., 169, 172 Copeland, J., 202, 226 Copeland, S. M., 204, 231, 248, 258 Corder, B. F., 513,536 Cordon, I., 408, 456, 462, 540, 573

AUTHOR INDEX

Cordon, I. M., 422, 471, 549, 557,562,570, 573 Corneille, 0., 202, 226 Cornish, W. R., 528, 542 Correll, J., 201, 226 Cortese, M. J., 291 Cortez, V., 462, 470 Corwyn, R. F., 282, 292 Costal!, A., 84, 114, 131, 152, 182, 192 Coupland, J., 22, 49, 590,598 Coupland, N., 590,598 Courage, M. L., 454, 455, 473 Courtney, A. S., 263, 294 Courtois, C. A., 146, 320 Cousins, C., 519, 539 Covey, J. A., 376, 391, 394 Cowan, N., 150, 151, 454,471 Cowley, E., 214,226 Cox, W. M., 286, 292 Coxon,P.,436,439, 590,600,612,613,614,619 Craig, J.D. R., 158, 172 Craig, R. A., 438,445 Craik, F. I. M., 581, 582, 583, 584,591,597, 598, 600, 601,603,629,631,641,643,644 Crasilneck, H. B., 335 Craske, M. G., 570 Craw, S., 111, 115, 425,573 Crawford, J. R., 637, 644 Crawford v. Washington, 9, 10, 524, 536, 564, 570 Crenshaw, M. C., 472, 571 Crespi, C., 288, 295 Crnic, K., 467, 469 Crocker, J., 196, 197, 199, 203, 236 Crombag, H. F. M., 84, 85, 112, 133, 149, 209,226 Cromer, C., 436, 448, 459, 474 Crosniak, L. D., 37, 46 Cross, T. P., 515, 517, 520, 536, 554 Crossman, A.M., 411,424,464, 551,654,660 Crowley, M.J., 511,512,514,515,519,528, 531,536 Crucian, G. P., 291 Cucciare, M., 203, 227 Culhane, S. E., 157, 158, 167,170, 172 Cull, C., 263, 294 Cully, D., 158, 172 Cummings, A. L., 29, 43 Curci, A., 151 Curran, H. V., 282, 283, 284, 285, 286, 292 Curran, T., 605, 624 Curtin, J. ]., 286, 292 Cutler, B. L., 67, 68, 77, 129, 136, 147, 149, 372, 375, 376,377,394,395,396,528,531,536,541,585, 587,598,601 Cutshall, J. L., 84, 116, 122, 126, 129, 130, 131, 135, 149, 155 Cuttler, C., 273, 294

AUTHOR INDEX

D

Daffner, K. R., 274,275,291,611,623 Dahl, L. C., 377, 394 Dale, P. S., 433, 445 Dalenberg, C. J., 307, 317 Daley, P. C., 513,536 Daly, J. A., 24,51 Daneman, M., 23, 50 Daniels, G. H., 33, 43 Danion, J. M., 270, 272, 292 Dark, V.J., 112 Darling, R., 22, 51 Darling, S., 88, 116 Dasgupta, N., 202, 227 Dasgupta, Z. R. R., 632, 645 Davey, S. L., 88, 89, Ill, 373, 393 Davidson, D., 107, 112, 437, 445, 503, 505 Davidson, G., 454,471, 631 Davies, G. M., 118, 155, 422, 436, 438, 525, 536, 550, 567,568,570,574 Davies, P. G., 201, 227 Davies, S. L., 422, 445, 450 Davis, D., 13,16, 21, 22, 23, 25, 27, 32, 33, 34, 35, 37, 38, 40, 41,43, 44, 48, 51, 124, 149, 197, 198, 199, 203, 205,206,207,208,210,213,215,216,220,223,226, 227, 231 Davis, J.D., 15, 40,47 Davis, L., 322, 335 Davis, M. H., 70, 75, 77,526, 534 Davis, M. R., 75, 77 Davis, S. L., 410,413,416,422,442,446,511,512,530, 532,535,536 Davis, T. L., 35, 40, 43, 7607, 623 Dawes, R. M., 207, 218,227 Dawson, V. L., 568, 572 Dawud-Noursi, S., 152 Day, J. D., 482, 504 Daylen, J., 82, 92, 93, 100, 116, 278,296 De Leonardis, D. M., 34, 48, 103, 113, 205, 232, 583, 601,607,623 De Poot, C. J., 212,235 de Rivera, J., 178, 182, 189 de Villiers, J., 433, 445 de Villiers, P., 433, 445 Deaux, K., 517, 536 DeCooke, P. A., 133, 147 Deese, J., 241, 242, 257, 263,292, 486,504, 623 Deffenbacher, K. A., 82, 92, 93, 96, 97, 98, 99, 100, 101, 102, 107, 111, 112, 120, 128, 149. 371,372, 373, 392, 393,394 de Haan, E. H. F., 290 Deitz, S. R., 513, 532, 536 Delamothe, K., 69, 78 DeLoache, J. S., 408, 422 DeMarie-Dreblow, D., 458,471

675

Dennett, J. L., 59, 63, 76 Dennis, N. A., 262, 291 Dent, H. R., 136, 137, 149, 265,292, 406,422, 429,433, 437,442,445,552,570 DePaulo, B. M., 22, 44. 339,340,348, 350,351, 352, 358, 361,362,363,364,365,667,668 Depret, E., 27, 45 Dermer, M., 27, 42 Derrico, M., 635, 645 Desikan, R., 274. 275,291, 611,623 Desmarais, S., 146, 283,295 Devine, D. J., 510, 526,536 Devine, P. G., 67, 78,516,517,536 Devitt, M. K., 353, 369 Devlin, Honorable Lord Patrick, 159, 171, 172, 667,668 DeVos, E., 515,536 Dewey, M. E., 514,536 deWinstanley, P. A., 24, 49 Dexter, H. R., 372, 377, 394, 531,536 Diamond, D. M., 82, 113 Diamond, S. S., 510, 530, 536, 542 Dickinson, J. J., 408,422 Didow, S., 457,472 Dietrich, D. E., 85, 112 Dill,D.l.,317 Dillehay, R. C., 531, 541 DiMatteo, M. R., 15, 29, 44 Dinges, D. F., 184, 189, 324, 327, 328, 337, 338 Dinnel, D. L., 131, 155 Dion, K. L., 200, 230 DiRubbo, M., 454,477 Diviak, K. R., 512,535 Dixon, A., 460, 474 Dixon, R. A., 637,644 Djdali, Y., 537 Dobson, M., 213, 227 Dodd, D. H., 214,227. 436,438,442,445, 448, 459,474 Dodge, M., 274, 293 Dodson, C., 178, 189 Dodson, C. S., 214,227. 272, 276,295,296,641,645 Dodson, J., 92, 116 Doherty, E., 536 Dolan, P. 0., 582, 598, 606, 622, 628, 629, 643 Dollar, K. M., 516,536 Donchin, E., 346, 347, 366 Donnelly, C., 112 Donnelly, D. A., 514,536 Dorado, J., 556, 572 Doran, A., 114 Dore, F., 350, 366 Doris, J., 429, 444. 445, 446, 447, 450, 470, 475, 573 Dostal, C., 425, 433, 448 Doueck, H. J., 515,537 Dovidio, J. F., 200,230 Drake, M., 326, 337 Drapalski, A. L., 207, 234

AUTHOR INDEX

676

Drew, D., 28, 49 Driskill, J., 574 Drivdahl, S. B., 213, 214,227, 385, 397, 501 Driver, E. D., 183, 189 Driver, R. E., 339, 370 Drizin, S. A., 176, 180, 187, 189, 189 Drucker, P.M., 557, 570 Drugge, J. E., 516,536 Drummey, A. B., 461, 471 Duchek, J. M., 291, 582,598, 606, 622, 628, 629,643 Dudas, R. B., 274, 292 Dudley, W., 582, 601 Dudukovic, N. M., 220, 221,227 Duensing, S., 588, 600, 615, 624 Duffy, C. M. M., 319 Duggan, L. M., 511, 512, 519, 527, 531,536 Dukes, R. L., 524, 536 Duncan, J., 242, 257, 630, 643 Dunford, B. B., 510, 536 Dunlop, B. D., 598, 602, 622 Dunlosky, J., 634, 643, 644 Dunning, D., 20, 44, 136, 137, 149, 184, 192, 227, 463, 471,594,602,627,645 Dunton, B. C., 200, 227 Dupuis, P., 644 Dupuis, S. E., 42, 205, 225 Dworkin, R. H., 24, 44 Dwyer, J., 170, 173, 176, 192 Dys, A., 188, 189 Dysart, J. E., 68, 80, 123, 124, 149, 286, 287,292, 585, 599,644 Dywan, J., 379, 394, 632, 644 E

Eakin, D. K., 213,214,227 Earles, J. L., 582, 583, 599 East, R., 77, 218, 228 Eaton, T., 525, 536 Ebbinghaus, H., 130, 149, 241, 257 Eberhardt, J. L., 201, 202, 227 Eberhardt, T. L., 29, 49 Eberle, P., 403, 422 Eberle, S., 403, 422 Ece, B., 133, 154 Eckerman, C., 457,472 Eddings, S. K., 21, 50 Edelson, M., 303, 317 Edelstein, R. S., 87, 96, 107, 112, 122, 149, 548,571, 573 Edery-Halpem, G., 117, 149 Edison, S.C., 454, 473 Edward, K., 352, 363, 369 Egan, M. F., 274, 293 Egerton, J. M., 410, 425, 482, 506 Eggert, L., 48 Egloff, B., 29, 43 Ehlich, P. J., 142, 154

Ehrenberg, K., 34, 45, 47, 228 Eich, E., 104, 112, 306, 317 Eich, J. E., 284, 292 Eigsti, I.-M., 413, 422 Eisen, M. L., 213, 227, 276, 278, 279, 292, 410, 413, 420, 422,442.446 Eisenberg, N., 514, 536 Ekehammar, B., 203, 224 Ekman,P.,253,257,339,340,341,348,349,350,351, 362,363,365,366,367,368 Elaad, E., 285,292, 341,343,345,346,366 EI-Ahmadi, A., 149 Eldridge, J. E., 158, 172 Elie C. J., 243, 259, 502, 505 Eling, P., 282, 293 Elischberger, H. B., 466, 471 Eliason, J. W., 319 Elliott, A. N., 511,540 Ellis, H. C., 104, 116 Ellis, N. R., 262, 292 Ellis, R., 214, 225 Ellsworth, P. C., 214,235 Elvevag, B., 292 Emrich, H. M., 112 Engelberg, E., 132, 135, 148, 149 England, P., 423, 571 Entzel, P., 598, 602, 603, 622 Epstein, D. M., 158, 172 Epstein, M.A., 511,512,517,535,537 Epstein, R. M., 37, 48 Er, N., 132, 133, 149, 154 Erdelyi, M. H., 324, 325, 335 Erdfelder, E., 206,227 Erickson, B., 596, 599 Erickson, D. J., 150 Erickson, E. A., 29, 43 Erickson, P. G., 282, 293 Ericsson, K. A., 30, 41,44 Erskine, A., 375, 394 Eslinger, T., 541 Espagnet, L., 634, 646 Esplin, P. W., 353,354, 356, 359, 360, 367, 416,417, 423,424,425,430,432,434,435,437,438,439,440, 443,444.44~448,449,450,456,468,473,475,476,

477,558,569 Estes, W. K., 219,227 Etcoff, N. L., 33, 51, 348, 366 Ethier, K. A., 517,536 Eugenio, P., 91, 113 Evans, S., 214, 235 Everson, M.D., 408,424, 434, 447, 464, 471, 473, 558,570 Ezaki, H., 49

F Fagiman, D., 330, 335 Falkner, K. L., 72, 77

AUTHOR INDEX

Fallot, R. D., 528, 538 Fanurik, D., 549, 574 Faries, J. M., 243, 258 Farinacci, S., 184, 189, 219,225 Farquhar, L., 574 Farrant, A., 268, 292 Farrar, M. ]., 141, 149, 458, 459,471 Farrell, R. A., 515, 517, 543 Farwell, L.A., 346, 347,366 Fasig, L., 459, 471 Fassler, 0., 322, 336 Fauchier, A., 437, 447 Faulkner, D., 588,598, 609,617,623 Faulkner, K., 608, 623, 633, 643 Faye, E., 580, 602 Fazey,J. A., 92, 112 Fazio, R. H., 200, 227 Feiring, C., 549, 551,570 Feldman, R. D., 580, 601 Feldman, R. S., 23, 50 Femina, D. D., 293 Ferguson, S. A., 34, 44, 583, 599 Ferguson, T. ]., 372,377,397 Ferrie, J., 112 Fiedler, K., 19, 44, 342, 344, 345, 347, 366 Fillmore, M. T., 285, 293 Finger, K., 67, 77, 215,233, 244,259 Finkel, N.J., 19,45 Finkelhor, D., 509, 514, 515, 516, 517, 519, 520,537, 539,554,572 Finkelman, D., 227,295 Finkenauer, C., 122, 149 Finlay, W. M. L., 263, 293 Finlayson, L. M., 513, 537 Fisher, J. E., 183, 189, 273,292 Fisher, P. A., 413, 425 Fisher, R. F., 129, 147, 167, 172 Fisher, R. P., 54, 55, 58, 59, 60, 61, 63, 64, 65, 66, 67, 68, 69, 70, 72, 75,77, 78, 79,129,136,149,173,353, 366,378,394,39~430,441,446,448,592,593,599,

601,641,642,644,646 Fishhoff, B., 206, 228 Fiske, A. P., 33, 45 Fiske, S. T., 28, 33, 45, 51, 169, 172, 197, 199,228 Fisler, R., 320 Fite, R., 324, 332, 336, 337, 338 Fitzgerald,]. M., 34, 45 Fivush, R., 93, 111, 118, 122, 123, 140, 141, 147, 149, 150,152,207,228,411,413,422,423,454,455,457, 458,459,461,462,463,469,471,472,473,474,475, 476, 548, 570 Flanagan, C., 425, 433, 448 Flannagan, D. A., 406,421 Fleck, K., 462,470 Fletcher, C. R., 13, 26, 42, 45 Fletcher, K. L., 291 Fligstein, D., 379,394

677

Flin, R., 119 125, 127, 130, 150, 410, 417, 422, 429, 436, 440,445,446,450 Flowers, K., 37,48 Foa, E. B., 280,297, 310,311, 317 Foisy, P., 631,644 Foley, L.A., 517,537 Foley, M.A., 35, 45, 50, 216, 230, 361, 367, 433, 436, 447, 459, 472 Folkard, S., 29, 49 Follette, W. C., 15, 21, 23, 25, 32, 34, 35, 38, 44, 197, 198,199,205,206,208,211,220,226 Follmer, A., 118, 152, 461,472, 475, 571 Fondacaro, R., 41,46 Fore!, A., 323, 335 Forgas, J.P., 218, 228 Forrest, K., D., 186, 190 Forrest, T. ]., 489, 492, 493, 504 ForsterLee, L., 512, 514, 522, 537 ForsterLee, R., 512, 514, 522, 537 Forsterling, F., 19, 50 Foss-Goodman, D., 511,515, 516, 518, 519, 542 Foster, H. H., 183, 190 Foster, R. A., 263, 294 Foushee, H., 526, 534 Fozard, ]. L., 582, 602 Fradin, S., 466,472 Fragale, A. R., 204, 228 Fraley, R. C., 29, 45 Francoeur, E., 40, 43, 108, 111,405,407,421,563,570 Frank, M.G., 61, 76, 200,228, 348, 363, 366 Frankel, R. M., 37,42 Frankish, C. R., 635, 643 Franks, J. ]., 60, 76, 241, 242,257,481,486,501,504, 605,622 Frazier, L. D., 235 Fredrickson, R., 322, 335 Freeman, C., 282, 293 Freeman, J. B., 515, 516, 537 Freer, C., 630, 643 Freud, S., 291,300,310,317, 323,335 Freund, J. S., 582, 603 Frey, L. M., 317 Freyd, ]. ]., 93, 112,247, 248, 250,257, 304,309,310, 317,319 Friedman, M. J., 142, 154 Friedman, R. D., 7, 9, 39, 43, 45, 226 Friedrich, W. N., 520, 537 Friesen, W. V., 348, 349, 350, 351, 366 Frischholz, E., 335 Frisina, D. R., 581, 601 Frisina, R. D., 581,601 Fromme, K., 288, 295 Fromuth, M. E., 516,536,542 Frost, P., 214,228 Fryman, H. M., 519, 537 Fukui, T., 49 Fulero, S.M., 19, 45, 68, 80, 173, 174, 397, 646

AUTHOR INDEX

678

Fulford, J., 600 Fulton, A., 45, 585,598, 614, 622 Fultz, J., 43 Furedy, J. J., 346, 365 Furtado, E., 475 Fussell, S. R., 14, 47, 49

G Gabarczyk, A., 233, 602 Gabbert, F., 184, 190, 209,228, 249, 257, 584, 585,586, 588,599,601,642,645 Gabora, N.J., 511,519,528,537 Gabrieli, G. H., 224 Gabrieli, J.D. E., 85, 112 Gabrielsen, R., 554,572 Gales, M.S., 178, 189, 583, 598 Gallai, D., 345, 347, 366 Gallant, C. L., 231 Gallo, D. A., 243, 245, 246, 247,257, 259,260, 267,295, 457,477,483,488,500,504,505,506 Galluccio, L., 184, 191,231,605,624 Gamst, G., 27,45 Gansler, D. A., 379, 396 Ganzel, B. L., 317 Garcia, L. A., 380, 392, 396 Garcia-Bajos, E., 204, 228 Gardiner, J. M., 631, 644 Gardner, W. L., 26, 28, 45, 198, 228 Garner, J.P., 29, 45 Garrett, S., 114 Garrioch, L., 587, 598, 616, 622, 627, 643 Garry, M., 108, 113, 134, 151, 184, 190, 191, 213, 214, 215,216,217,222,224,228,229,231,235,236,244, 250,257,258,260,288,291 Garven,S., 73, 74,77,209,210,228,381,396,404,405, 406,408,414,415,420,422,442,446,553,571 Garver, R. B., 323,324,331,335 Gavrilescu, D., 285, 293 Gawronski, B., 34, 45, 205, 228 Geddie, L., 466,472 Gee,S.,407,410,422,425,465,472,482,506,559,573 Geiselman, R. E., 54, 55, 58, 59, 64, 65, 66, 67, 69, 72, 74, 76, 77, 79, 136, 149, 353,366, 378,380, 387, 392, 394,396,430,446,592,593,599,641,644 Gencoz, F., 151 George, D. T., 284, 296 George, R. C., 53, 54, 56, 66, 77, 79 Gerhard, D., 381, 396 Gerkens, D. R., 308,317, 320 Gerrie, M., 213,214, 216,228 Gerson, L. W., 282, 293 Ghetti, S., 410, 422, 466, 472, 493,499,505, 511,535, 550,562,572,573,605,623 Giambra, L. M., 582, 599 Giannelli, P. C., 335 Gibbons, D., 183, 192

Gibbs, R. W., 183, 192 Gibling, F., 128, 155 Gilbert, D. T., 29, 39, 40, 46 Gilbert, J., 60, 78 Gilbert, M. S., 320 Giles, D. C., 389, 391, 392, 394 Giles, H., 590, 591, 598, 599 Giles,J. W., 410,422 Gilligan, C., 514, 537 Gilligan, P. A. T., 286, 292 Gilliland, T., Gilovich, T., 45, 200, 228 Gilstrap, L. L., 419, 422 Gimpert, N., 35, 50 Girgis, A., 15, 46 Gisle, L., 149 Giuliani, V., 26, 42 Glatt, R. L., 333, 338, 379, 396 Gleaves, D. H., 247, 248, 250,257, 300, 303,307, 308, 317, 320 Gleitzman, M., 464, 477 Glenberg, A.M., 104, 115, 306, 320 Glisky, E. L., 34, 46, 631 Glover, T. A., 213,237, 277,296 Gluck, D., 24, 25, 46 Gobbo,C.,410,412,422,426,456,457,458,468, 472,477 Godfrey, D. K., 28, 48 Goff, D. C., 272, 296 Goff, L. M., 123, 154, 244,258, 283,295 Goff, P. A., 201, 227 Goggin, W. C., 206, 228 Goh, C. L., 460,477 Gold, G., 24, 46 Gold, S. N., 311,317 Goldberg, A., 123, 150, 462,471 Goldberg, L. R., 310,317 Goldberg, T. E., 273, 274, 292, 293 Goldfinger, S. H., 24, 44 Golding]. M., 511,512,519,521,522,523,525,527, 53~538,542,563,571

Goldman, M.S., 295 Goldsmith, M., 60, 70, 78, 103, 114, 240,258 Goldstein, C. C., 169, 172, 180, 190 Goldstein, G., 266, 268, 294, 296 Gooderson, R.N., 157, 158, 159, 165, 171, 172 Goodman, G. S., 87, 108, 112, 115, 118, 122, 126, 141, 147, 149, 150, 186, 188, 192,355,366,408,409,410, 412,413,414,416,420,422,423,425,426,436,440, 441,442,446,449,455,456,457,458,459,460,461, 462,464,465,466,467,468,471,472,476,477,478, 493,505,509,510,511,512,514,515,518,519,520, 524,525,527,528,531,532,533,535,538,539,540, 541,548,549,550,551,554,556,559,560,561,562, 563,564,565,566,567,569,570,571,573,574,605, 623 Goodman, J., 129, 148

AUTHOR INDEX

Goodman-Brown, T. B., 548,571 Goodwin, D. W., 282, 293 Goodwin, J., 513, 534 Goodwin, K. A., 243, 260, 488, 507 Goodwin, S. A., 169, 172 Goossens, L., 582, 603 Gopnik, A., 410,422,466,472 Gordon, B. N., 214, 233, 264,294, 405, 413,421, 425, 455,457,461,466,467,469,470,472,474,475, 559, 571 Gordon, D. S., 548,571 Gordon, H. M., 141, 147 Gordon, L. L., 481, 504 Gordon, R. V., 186, 191 Gordon, S. E., 27, 46, 228 Gordon, T., 581, 600 Gordon-Salant, S., 581,599 Gorenflo, D. W., 16,51 Gorman, J. M., 270,291, 318 Goss, B., 28, 46 Gosslin, P., 350, 366 Gotlib, I. H., 125, 147 Gottlieb, M. C., 320 Gowdey, D. A., 567, 574 Graesser, A. C., 12, 13, 26, 27, 46, 196, 203, 204, 228, 242,259 Graf, P., 466, 472, 584, 603 Graham, K. S., 274, 282, 292, 293 Graham, S., 351,369,571 Granhag, P. A., 65, 78, 168, 173,210,211,224,230,363, 366,368,369,373,383,391,392,394 Grant, H. M., 104, 112 Gravitz, M. A., 335 Gray, C., 585, 601, 645 Gray, E., 513, 520,538, 539 Graziano, W., 27, 42, 306, 320 Greco-Vigorito, C., 570 Green,C.,219,224, 302,303,304,305,316,636,643 Green, J.P., 332, 335 Green, M. C., 196, 197, 199,228 Greenberg, D. L., 84, 112, 123, 133, 150, 154 Greenberg, M.S., 127, 150, 204,229 Greene,E.,209,213,231,242,258,513,540,588,598, 617,622 Greenhoot, A. F., 410, 423, 456, 466, 467, 472 Greenstock, J., 456,472 Greenwald, A. G., 34, 46, 62, 78, 201, 205, 207,229 Greenwood, K. M., 70, 77 Greenwood, P. M., 629, 630, 644 Gregory, W. L., 216,229 Gresham, A. W., 513, 522,539 Grice, H. P., 16, 17, 19, 27, 30,46 Grieve, R., 409,423, 441,442,446 Griffith, J.D., 529, 538 Grinker, R. R., 300, 317 Grlliso,T.,532,538,547,571 Groeneweg, J., 119, 155, 253, 260

679

Gross, J. J., 29, 43, 50, 82, 113,407,421,464,470, 472, 557,558,570,571 Gross, M.S., 217,234,583, 602,610, 624 Gross, S. R., 176, 190 Groves, J. E., 112 Gruendel, J., 436, 448, 458, 475 Gruneberg, M. M., 375, 396 Grupe, L. A., 291 Guarino, M., 288, 295 Gubin, A., 169, 172 Gudjonsson, G. H., 176,177, 178, 179,186,189, 190, 192,213,264,265,282,292,293,344,366,412,418, 419,423 Giilgoz, S., 133, 154 Gunderson, E. W., 289,293 Guth, M., 436, 448 Guzy, L., 62, 78 Gwyer, P., 67, 68, 78 Gwynn, M.l., 333, 338, 379,396

H Haber, L., 130, 150, 224, 229, 616, 623 Haber, R.N., 130, 150, 224, 229,616, 623 Hackett, C., 219,224 Hackney, A., 592, 600 Haddock, G., 271,295 Haden, C. A., 207,228, 455, 457,471, 472, 476 Haegerich, T. M., 511,512,513,514,515,526,531, 538,541 Hafstad, G. S., 211,229 Hagen, L., 108, 113, 134,151, 184, 191,215,217,231, 250,258 Haith, M. M., 90, 506, 519,538 Hakami, M. K., 24, 47 Hala, S., 266, 268, 293 Halberstadt, J., 202, 203, 229 Hall, C. L., 51 Hall, J. A., 198,230 Hall, K.,319 Hall, M., 213,237,635,646 Hall, S. R., 566,571 Hallberg, E., 29, 43 Halliday, H. E., 35, 43 Halverson, C. F., 436,448 Hamann, S., 85, 113 Hamilton, D. L., 229, 528, 538 Hamlin, S., 526, 538 Hamm, V. P., 585, 599 Hammersley, R. H., 136, 153 Hammond, D. C., 129, 148,307,316,323,324,331,335 Hammond, R., 644 Hamond, N. R., 410,423, 454,455,459,463,471, 472 Hancock, P. J. B., 225 Handa!, A., 222, 235 Hannesdottir, K., 186, 190 Hannigan, S. L., 204, 205, 229

680

Hansen, E., 541 Hansen, T., 37,48 Harbeck-Weber, C., 549,571 Hardt, J., 125, ISO Hardy, L., 92, II2 Harkins, S. W., 582, 602 Harner, L., 433, 446 Harper, D. N., 214,228, 229 Harries, K., 635, 643 Harris, L., 511,531,535 Harris, L. J., 633, 638, 645 Harris, R. J., 19,46 Harrison, L. D., 282, 293 Harsch, N., 84, 114, 132, 133, I5I Hart, C. L., 289, 293 Hart, R. P., 582, 602 Hartel, D. M., 581, 602 Hartman, M., 585, 599 Hartshorn, K., 454,477 Harvey, M. R., 307, 3I7 Harvey, P. D., 271,294 Hasel, L., 263, 295 Hasemann, D., 512,537 Hasher, L., 284,293, 297, 581, 585, 591,599, 601, 603, 635,636,644,646 Hashtroudi, S., 22, 34, 37, 44, 46, 47, 103, I 13, 183, I90, 216,222,230,247,258,271,277,293,410,424,460, 473,484,488,505,583,599,600,607,623,644 Haslam, N., 33, 45 Hasselmo, M. E., 94, I 13 Hastie, R., 25, 46, 197, 199, 206,229,233,513,538 Hastings, P. A., 511, 526,54 I Hatton, C., 264, 265, 294 Haugaard, J. J., 555, 57 I Haw, R. M., 68,78 Haward, L. R. C., 322, 335, 336 Hawkins, E. R., 204, 233, 248, 259 Hawkins, S. A., 25, 46, 206, 229, 584, 600 Hay, A., 13, 31,43 Hayes, B. D., 69, 78 Hayes, B. K., 488, 505 Hayne, H., 214,236,407,414,421,427,454,455,463, 464,470,472,477,552,557,558,563,570,57I,574 Hazan, C., 408,423, 441, 446 Hazen, K., 54 I Hazlett, G., II4 Hazzard, A., 462, 47I Healy, A. F., 119, I 50 Healy, M. R., 614,623 Heaps, C., 244, 258 Heath, C., 205, 228 Heather, N., 286, 292, 293, 294 Heber, S., 3I9 Heckers, S., 272, 296 Hector, J., 600 Hedden, T., 609, 623 Heeren, D., 270, 294

AUTHOR INDEX

Heim, C., 82, 93, I 13 Heishman, S. J., 288, 293 Hekkanen, S. T., 213,229 Helgeson, V. S., 5I9, 538 Hembrooke, H. H., 214,225,360,365, 42I, 458,476 Henaghan, M., 565, 573 Henderson, A. M. E., 266, 268, 293 Henkel, L.A., 183, I90, 216, 217,229, 245,258, 460, 472,495,505,607,623 Henry, B., 125, ISO Henry, J.D., 550, 562,57 I, 638, 644 Henry, L.A., 264, 265, 293 Henwood, K., 590, 598 Hepps, D., 462, 472 Herman, J. L., 322, 336, 550, 57 I Hernandez, E., 300, 3I7 Herndon, F., 384, 391, 395 Herscovici, B. B., 510, 538 Hershkowitz,!., 105, I 13, 353, 354, 356, 359, 360 367, 408,416,417,418,423,424,425,430,432,434,435, 437,438,439,440,441,443,444,445,446,447,448, 449,450,456,464,465,468,473,475,47 Hertel, P., 282, 294 Hertzog, C., 634,643 Herve, H. F. M., 58, 282, 297, 339, 368 Hess, T. M., 34, 46, 340, 350, 367, 406, 42I Hester, R. K., 282, 294 Hetherton, J., 515, 516, 538 Heuer, F., 84, 85, 87, 91, 109, I12, I 13, 115 Heyman, G. D., 410, 422 Hickam, D. H., 16, 49 Hicks, J. L., 33, 35, 42, 48, 229 Higgins, E. T., 41, 46, 219,229 Highhouse, S., 214, 230 Hijman, R., 270, 290 Hildebrandt, H., 282, 285, 293 Hildreth, K., 454, 473 Hill, K., 441,444 Hill, P. E., 552, 57 I Hill, S.M., 552,57 I Hilton, D. J., 19, 27,47 Hinton, I. D., 549, 574 Hira, K.,49 Hirschman, J. E., 355, 366, 441, 446, 462, 472, 538, 550 Hirshman, L., Hirst, W., 24, 25, 46 Hirt, E., 206, 230 Hitch, G. J., 211,232 Hitchcock, R., 441, 444 Ho, R., 512,537 Hock, H. H., 489, 504 Hodge, D., 215,233, 244, 250, 251,259 Hodges, J. R., 274, 292, 293 Hodgins, H. S., 230 Hoefer, E., 353, 367 Hofer, B., 232 Hoffman, D. D., 196, 230

AUTHOR INDEX

Hoffman, H., 522, 540 Hoffman, H. G., 184, 191, 201,210,216,229,230,595, 600,627,644 Hoffman, M., 214,235 Hoffrage, U., 230 Holland, H. L., 69, 72, 77, 378, 394 Holley, A., 65, 79 Holliday, R. E., 480, 484, 486, 488, 490, 492, 495, 496, 498,501,502,504,505,506,593,594,603 Hollin, C. R., 373, 393 Hollins, T. S., 634, 645 Hollnack, S., 93, 112 Holmes, D. L., 303,317,436,448 Holst, V. F., 127, 150, 204,230 Holt, A. R., 516,536 Holtgraves, T. M., 14, 15, 16, 17, 18, 19, 22, 23, 27, 31, 44,46,47,220,230 Honts, C. R., 341,342,343,345,353,354,355,358,367, 368,369 Hope, L., 585, 601, 633,637,645 Hopkins, S., 541 Horan, N., 432, 445 Horgan, T. G., 198,230 Horn. M., 309, 318 Horowitz, D., 104, 113,418,424,425,430,440,441,446, 447.448,465,473,475 Horowitz, 1., 322, 336, 537 Horowitz, M. J., 318 Horowitz, S. W., 341,354,355,356,357, 360, 367,368 Horselenberg, R., 185, 188, 190 Horsley, J. S., 322, 336 Horton, W. S., 15, 47 Hosch, H. M., 157, 158, 167, 170, 172, 283, 294, 373, 374,376,394,395,511,541,542,628,644 Hosier, S. G., 286, 292 Hotaling, G., 515,537 Houlihan, D., 43 House, A. T., 35, 45 Hovav, M., 367, 424, 446, 447, 473 Howard, J. D., 35, 42 Howe, M. L., 406, 410,423, 424,454,455,462,473, 491, 493,505 Howie, P.M., 375, 394, 560, 570 Hoyer, W. J., 588, 600, 619, 623 Hoyt, G., 114 Huang, M. T., 150 Huart, J., 202, 226 Hubbard, T. L., 197,230 Hubert, H. H., 355, 367 Hiibinette, B., 87, 112 Hudson, A. E., 457,475 Hudson, J. A., 141, 150, 410,423, 455,457,458,459, 461,471, 473 Hudson, J.l., 202,233 Hudson, S.M., 513,534 Huffman, M. L., 556, 572

681

Huffman, M. L. C., 73, 76, 148,411,413,421, 423, 442,445 Hughes, M., 409,423,441,442,446 Hulicka, l. M., 582, 599 Hull, R. G., 308, 320 Hulse, L. M., 128, 150 Hulse-Trotter, 557, 574 Hultsch, D. F., 582, 599, 645 Hummel, J. E., 26, 45 Hummert, M. L., 590, 595, 599, 600 Hung, L., 219, 225 Hungerford, A., 408, 423 Hunt, J. S., 213,230, 432,434,446, 450, 451 Hunt, M. J., 244, 258, 405, 416, 417, 426 Hunter, J. E., 524, 540 Hunter, M., 514,516, 524,538 Hunter, R., 58, 80, 289, 296 Husband, T. H., 184, 190, 243, 258 Hutcheson, G. D., 437, 446 Hyer, L.A., 93, 114 Hyman, l. E., 123, 131, 135, 148, 150,1 55,184, 186, 190, 220,230,243,244,249,250,253,258,277,278,293, 296,325,336,401 I

Iacono, W. G., 344, 345, 347, 367, 368 Ickes, W., 51 Igou, E. R., 30, 47 Imhoff, M. C., 408, 422, 433, 446 Inbau,F.E., 177,180,190,230 Infield, L. L., 24, 51 Inhelder, B., 482,483,506 Intons-Peterson, M. J., 592, 600 Iobst, A. D., 277, 295 Irion, A. L., 306,318 Isingrini, M., 634, 646 Isquith, P. K., 511, 512,514, 515,531,533,536, 538, 539 Israel, L., 495, 500, 505, 507, 605, 611, 623, 624 Ivy, G. 0., 630, 644

J Jackson, B., 120, 147 Jackson, D. D., 17, 51 Jackson, E. M., 623 Jackson, J. R., 220, 227 Jacob, J., 128, 155 Jacobs, W. J., 303, 305, 307, 309, 318 Jacoby, J.D., 82, 104, 114 Jacoby,K., 176,190 Jacoby,L.L.,201,233,234,585,600,613,614,619,623, 632,644 Jaenicke, C., 549, 574 Jakway, A., 263, 295 Jamieson, G. A., 184, 192 Janet, P., 300,318, 321, 323, 336

AUTHOR INDEX

682

Janowsky, D. S., 288, 296 Janus, E., 214, 226 Jarvis,}., 141,147 Java, R. 1., 630, 645 Javitt, J. C., 581,602 Jayne, B. C., 177, 190, 223, 230 Jenkins, L., 82, 114 Jennings, J. M., 613, 623, 631, 643 Jennings, L., 516, 540 Joab, A., 511,537 Joaquim, S. G., 319 Joffe, R., 58, 80, 289, 296 Johannes, S., 112 Johansson, M., 211,224 Johnson, B. C., 596, 599 Johnson, J. T., 384, 389, 395 Johnson, J. W., 481, 482,505 Johnson, M. K., 13, 32, 34, 42, 44, 46, 47, 48, 103, 113,183,184,190,191,199,205,214,216,217, 222,225,22~230,231,232,234,245,247,254,

258,271,277,293,361,367,410,424,433,436, 447,459,460,466,472,473,474,484,485,488, 495,502,505,583,588,598,599,600,601,602, 607,610,620,623,624,644 Johnson, R., 630, 643 Johnson, S.D., 391, 394 Johnson, T. C., 88, 115, 128, 154 Jones, C., 461, 473 Jones, D.P. H., 423, 437, 520,543, 548, 549, 559,571, 572, 573 Jones, E. B., 85, 113 Jones, E. E., 39, 40, 45 Jones, E. M., 34, 35, 40, 43, 607, 623 Jones, G. V., 306, 318 Jones, H. P. T., 371,397, 519,543 Jones, H. R., 410, 421 Jones, H. T., 587, 603 Jones, L. M., 509, 539, 554, 572 Jones, M., 404, 424 Jones, R. W., 634, 645 Jones, S., 412,424 Jonsson, A-C., 168, 173 Jordan, J. J., 589, 601 Jordan, R., 268, 294 Joseph, A. L., 259 Josephs, R. A., 286, 295 Josephs, S., 185, 190 Joye, N., 574 Judd, C. M., 200, 201,226, 230, 237, 377, 396 Julien, R. M., 282, 288, 293 Jung, S., 587,598 Jurkevich, L. M., 64, 77 Juslin, P., 206, 237, 373, 374, 394, 395 Jussim, L., 172 Justice, C., 209, 237, 249, 260, 385, 397

K

Kahn, R. S., 270, 290 Kahneman, D., 59, 78 Kaloupek, D. G., 311,318 Kalven, H., 528,531,539 Kamio, Y., 266, 296 Kanade, T., 349,365 Kannel, W. B., 581, 600 Kaplan, D. 1., 550, 572 Kaplan, M., 528, 539 Kaplan, S., 104, 113 Kardiner, A., 300,318 Karibian, D., 492, 493, 505 Karlin, R. A., 323, 330, 336, 337 Karlsson, 1., 134, 150 Karpel, M. E., 588, 600, 619,623 Kassin, S.M., 20, 47, 136, 150, 169, 170, 172, 176, 177, 178, 180, 181, 182, 184, 186, 187, 188, 189, 190, 191, 192,216,217,230,282,283,294,363, 368,628,638,644 Kastner, J., 521,541 Kastner, S., 521,541 Kaszniak, A. W., 645 Katz, A. N., 213,236, 567 Katz, E. R., 567, 570 Kausler, D. H., 24, 47, 631, 644 Kawakami, K., 200, 230 Kay, D., 330, 335 Kean, R. B., 524, 536 Keane, T. M., 310,311,318 Keating, L. M., 514,542 Kebbell, M. R., 53, 70, 72, 78, 264, 265, 294, 336, 376, 378,379,389,391,392,394 Keefe, R. S. E., 271,294 Keenan, J. M., 24, 26, 28, 29, 47, 48 Keeney, J. M., 39, 51, 426, 451, 525, 542, 574 Keiffer, J., 442, 447 Keillor, J. M., 291 Kelland,A.,406,423 Kelley, H. H., 169, 172,474, 528,539 Kelly, R. J., 513,539 Kemmelmeier, M., 15, 25, 32, 34, 35, 38, 44, 205, 220,226 Kemp, R. 1., 66, 79 Kemp, S., 121, 148, 167, 172, 215,230 Kemper, S., 26, 27, 47 Keniston, A., 437,445 Kennedy, M.A., 58, 88, 112 Kenney,R., 150,412,422,461,472,510,538,563,571 Kenny, D. T., 15, 48 Kenrick, D. T., 202, 232 Kensinger, E. A., 611, 623 Kent, G. G., 15, 47 Kent, J., 375,397, 584, 587, 603

AUTHOR INDEX

Kenyon, S., 514,536 Keogh,E.,213,234 Kerker, R. M., 168, 173 Kerr, N. L., 512,540 Kester, J. D., 629, 644 Ketcham, K., 198, 202, 207, 216, 217,231,548,572 Keuler, D.]., 207,230 Key, H. G., 522, 539 Keysar, B., 14, 42, 47 Khalsa, S., 326, 337 Kiechel, K. L., 178, 181, 184, 187, 190 Kiernan, B., 482,483,484,485, 496, 497,499, 501,506 Kihlstrom, J. F., 93, 99, 105, 113, 115, 135, 154, 280, 291,303,307,318,322,323,336,632,645 Kilpatrick, D. G., 568, 572 Kim, J. J., 82, 113 Kindt, M., 278, 279, 294 King, D. W., 142, 150 King, L.A., 150 King, M.A., 434,441,442,447 Kintsch, W., 23, 26, 42, 47 Kirby, P., 211, 232 Kircher, J. C., 341, 342, 344, 367, 438, 445 Kirchner, T. R., 285, 294 Kirkendol, S. E., 351, 364, 365 Kirouac, G., 350, 366 Kirsch, 1., 187, 192, 214, 216,232, 235, 251,258, 324, 332,335,336,33~338

Kitayama, S., 26,47 Klauer, K. C., 33, 34, 45, 47, 205, 228, 230 Klaus, P. A., 578, 579, 600 Klaver, J., 32, 186, 191 Kleck, R. E., 340, 350, 367 Klein, K., 213,230 Kleinhauz, M., 322, 336 Kleinknecht, E., 466, 473 Kleinsmith, L. J., 104, 113 Klettke, B., 26, 46, 196, 228 Kline, D. W., 580, 581, 600, 602 Klinger, L. G., 266, 295 Klinger, M. R., 266, 295 Kluft, R. P., 318 Knaack, D., 44, 149, 227 Knauer, E., 93, 112 Kneer, R., 498, 500, 504 Kneller, W., 585, 600 Knie, K., 12, 22, 49 Knight, D.P., 51 Knox, A., 120,150,417,422,436,446 Knox, J., 322, 336 Koehnken,G.,63,64,65,69,78,79,352,353,354,357, 360,362,363,365,36~369,406,424,593,600

Koeske, R., 456, 470 Koestner, R., 363, 370

683

Kong, A., 15, 47 Koocher, G. P., 513,537 Kopelman, M.D., 93, 113, 182, 191 Kopiec, K., 554, 572 Koriat, A., 60, 66, 70, 78, 103, 114, 240,258, 306,318 Kornbrot, D. E., 131, 151 Kossak, A., 600 Koutstaal, W., 184,191,217,231,235 Kovac, P., 436, 438 Kovera, M. B., 441,448, 511,513,522, 525, 528, 529, 539,540

Koyama, M., 25, 44, 149, 207,227 Kozak, M. J., 311,317 Krackow, E., 334, 336 Krafka, C., 67, 78, 378, 395 Kramer, T., 91, 113 Krausler, D. H., 584, 600 Krauss, R. M., 14, 22, 47, 220,221,231 Krausz, M., 270, 294 Kremnitzer, M., 342, 365 Krispin, 0., 367 Krosnick, J. A., 20, 30,47 Krug, K. S., 155 Krugman, R. D., 559, 572 Krull, D. S., 29, 45 Kuebli, J., 458,471, 473 Kuhl, B., 224 Kuhn, J., 414,423, 455, 464,472, 559,571 Kulas, J. F., 29, 47 Kulik, J., 29, 43 Kunda,Z., 197,199,202,203,231 Kurbat, M.A., 164, 173 Kusal, T., 44, 149, 227 Kutas, M., 104, 116 Kvavilashvili, L., 131, 151 Kwon, P., 459, 474 Kwong See, S. T., 595, 600, 627,644 L

La Rooy, D., 120, 136, 139, 151, 473 Lacey, S. C., 278, 295 Lachs, M. S., 579, 600 Laird, J., 557,573 Lamb, M. E., 74, 75, 79, 104, 113, 136, 140, 152, 153, 353,354,358,359,360,367,405,408,415,116,417, 418,423,424.425,429.,430,432,433,434,435, 437,438,439,440,441,442,443,444.445,446,447, 448,449,450,455,456,458,460,463,464,465,466, 468,470,473,475,476,477,478,523,535,541, 558,569 Lambert, A. J., 26, 27, 52, 221, 223 Lamphear, V., 435, 449 Lampinen, J. M., 204,231, 233, 243, 246. 248,258, 259,601

684

Lanceley, A., 590, 601 Landau, J. D., 33, 35, 48 Landis, T. Y., 456, 473 Landrine, H., 516,539 Landry, K. L., 357, 360, 367 Lane,P.,406,426,437,440,45I,618,625 Lane, S.M., 213,237,460,478 Laney, C., 91, 92, 113 Lang, A. R., 29, 42, 286, 292 Lang, P. J., 318 Langer, E. J., 26,47 Langford, S., 306,318 Langley, J., 125, 150 Lanz, M., 282, 293 Larsen, S. F., 37, 48, 112 Larsen, S. R., 121, 155 Larson, S.M., 451, 556,572 Larsson, A. S., 65, 78 Larus, D. M., 214,233, 405,421, 440,448, 454,461,469 Lassiter, G. D., 177, 191, 213, 231 Laulhere, T. M., 109, 113, 139, 147, 469 Laurel!, 0., 458, 470 Laurence, J. R., 323, 324, 336 LaVoie, D. J., 608, 609, 614, 621,623, 624 Law, S., 584, 600 Lawton, S. C., 482, 504 lc Doux, J. E., 219,233 Lea, J. A., 525, 540, 600 Leavitt, F., 307, 310,318,413,424 LeBaron, S., 551,576 Lee, F., 27, 48 Lee,J., 132,151,578,584,589,601 Lee, K., 213,231 Lee, Z., 186, 191 Lehman, D. R., 19, 30, 47, 109, 115, 134, 153, 244,259 Leibowitz, H. W., 62, 78 Leichtman, M.D., 73, 74, 78,405,407,412,413,422, 429,437,438,440,442,445,447,460,461,474 Leippe, M. R., 372, 373,382,395,410,424,510,519, 520,527,539,548,572,596,603 Lench, H., 551,573 Lenzenweger, M. F., 281, 292 Leo, R. A., 40, 41, 44, 49, 176, 177, 178, 187, 188, 189, 189, 191, 192 Leonard, A.M., 214,231 Lepore,S.J., 74,79,408,412,424,442,450,653,668 Lese, K. P., 320 Lesgold, A. M., 68, 76 Leskin, G. A., 307,311,318 LeVan, V., 601 Levidow, B., 588,600, 616,624 Levin, D. T., 203, 235 Levine, B., 631,635,646 Levine, L.J., 82, 96,109,113,132,139,147, 151,207, 218,234,464,469 Levine, M., 416,421, 434,445,511,515,522,536, 537, 538,539

AUTHOR INDEX

Levinson, S. C., 17, 22, 43, 48 Levitt, M., Ill, 122, 147, 152, 462,469,475 Levy, B. R., 591, 600 Levy, R. J., 511,539 Lewinsohn, J., 29,49 Lewis, D. 0., 293 Lewis, I. A., 515,537 Lewis, M., 551,570 Lexcin, F., 571 Ley, P., 15, 29, 48 Li, G., 19, 30,47 Liang, W., 37,48 Libby, L. K., 123, 151 Liben, L. S., 481, 482,505 Libkuman, T. M., 529, 538 Liebel, L. M., 136, 147,463,469 Lieberman, M. D., 23, 48 Liebmann, J. !., 231 Liegerg, G., 319 Lien, J., 349,365 Lifton, R.J., 178,191 Light, L. L., 306,318, 614,623, 634,644 Lilienfeld, S. 0., 322, 336 Lillie, A., 141, 147 Lim, P. L., 493, 505 Lim, R., 158, 172 Lincoln, R., 61, 76 Lind, E. A., 531,539, 596, 599 Lindberg, M.A., 412,424,442,447, 456,474 Lindenberger, U., 629, 644 Lindsay, D. S., 32, 47, 103, 105, 108, 113, 115,119, 126,128,136, 140, 141, 148, 151, 153, 178,183, 184, 189, 190, 191, 192, 213, 214, 215, 216, 222, 230,231,247,250,258,260,271,277,293,322, 337,339,367,372,392,395,401,410,411,412, 424,425,433,440,441,442,447,449,458,459, 460,470,473,474,476,484,485,488,502,505,522, 539,560,573,583,598,600,607,623,644 Lindsay, J. J., 44, 365 Lindsay, R. C. L., 80, 149, 158, 159, 167, 172, 173, 286, 292,372,377,395,397,525,526,539,541,563,573, 585,599,600,633,642,644,646 Linholm, K. J., 517,539 Linville, P.W., 169, 172 Lipovsky, J. A., 568, 572 Lips, H. M., 511,514,515,519,534 List, J. A., 582, 588, 601,633,644 Litz, B. T., 142, 154 Lloyd, F., 483, 484, 485, 487, 493, 502,506 Lobb,E.A., 16,48 Loewenstein, R. J., 318 Loftus, D. A., 104, 115 Loftus,E.F.,22,23,34,35,36,37,44,48,60,68, 73,76, 78, 86, 87, 91, 104, 108, 114, 115, 118, 122, 123, 129, 133, 140, 141,146, 148, 150, 152, 163,172, 173, 178, 184, 190, 191, 192, 198, 202, 205, 208, 209, 210, 212, 213,214,215,216,225,226,228,230,231,232,236,

AUTHOR INDEX

237,242,243,244,258,259,322,325,326,336,339, 368,371,380,381,394,396,413,421,442,445,501, 505,522,540,548,572,588,598,615,623,624, 635,646 Loftus, G., 87, 114 Logie, R., 211, 229 Loiselle, R. C., 262,291 Lombardo, P., 215,232, 253,258 London,K.,329,337,458,474,556,572 Lopez,P.,25,44, 149,207,227 Lord, C. G., 28, 48 Lovallo, W.R., 551, 570 Lucioli, R., 591,601 Luminet, 0., 122, 131, 132, 149, 151 Lundregan, T., 160, 172 Lundy, B., 122,152,462,475 Luo, Z., 107, 112, 256,260 Luszcz, M.A., 129, 148, 167, 172 Luus, C. A. E., 211,232, 372,384,387,395 Lyall, L. M., 51 Lykken, D. T., 347,348,367 Lynn, S. J., 108, 114, 213, 215,227, 232,322, 323, 324, 326,332,334,335,336,33~338

Lyon, T. D., 403,411,415,420,424,555,556,572 Lyons, E., 263, 293

M Maass, A., 91, 114, 387, 393 MacArthur, L. A., 540 MacCoun, R. J., 512,540 MacDonald, S. W. S., 637,644 MacDonald, T. K., 286, 292 MacEachron, A. E., 514,534 MacKeith,J.A.C., 179,190 MacKinnon, D. P., 72, 77, 378, 394 MacLaren, V. V., 343, 346, 368 MacLeod, C. M., 630, 644 Macleod, M., 635, 644 MacLeod, M. D., 222, 232 MacLeod, M.S., 637,644 Macmillan, N. A., 500, 506 MacMurray, B., 540 Macrae, C. N., 35, 48, 222, 232 MacWhinney, B., 24, 26, 28, 29, 47, 48 Magee, J. J., 348, 366 Magid, L., 185, 191 Mahoney, G. J., 481, 482, 505 Maldonado, J., 306, 318 Malinoski, P. T., 322, 326, 332, 334, 336, 338 Malloy, L. C., 556, 572 Malmstrom, T., 621, 624 Malone, B. E., 44, 366 Malpass, R. S., 67, 73, 74, 77, 78, 173, 174, 210,228, 397, 404,406,422,440,442,446,447,553,571,646 Malvagia, S., 215,232, 252,258 Mandler, J. M., 464, 474

685

Mandler, G., 13, 46, 614, 624 Maner, J. K., 202,232 Manion, A. P., 410,424,510,539,540,548,572 Mann, S., 351,359,363,368, 369 Manning, C. G., 184, 190, 215, 217,228, 235, 244, 257,259 Manor, T., 424, 447 Marando, L., 158, 172 Marche, T. A., 406, 409, 423, 424, 480, 484, 486, 501, 502,506,589,601 Marcoen, A., 582, 603 Marin, B. V., 436,448 Markham, R., 214,227, 375, 394 Markman, K. D., 206,230 Marks, A. E., 520, 527, 535 Markus, E. J., 630, 644 Marlatt, G. A., 282, 283, 288, 294, 295 Marmelstein, L., 326, 336 Marquis, K. H., 214,232 Marsh, E. J., 151, 220,227, 232, 236 Marsh, R. L., 33, 35, 42, 48 Marshall, J., 214,232 Marshall, T. C., 207, 236 Marsil, D. F., 512, 519, 526,537, 541, 543,563,573 Marsolek, C. J., 26,45 Marten, E., 296 Martens, T. K., 376, 394 Martin, C., 263, 294 Martin, C. L., 436, 448 Martin, J., 29, 43 Martin, M., 467, 474 Martinez, D., 393 Marvel, M. K., 37, 48 Marxsen, D., 128, 155, 282, 297, 352, 353, 370 Marzolf, D. P., 408, 422 Masling, M., 26, 42 Mast, M.S., 198, 230 Mather, M., 34, 48, 205, 232, 245, 254, 258, 495, 505, 588,601,620,624 Matheson, D. J., 176, 190 Matsui, K., 49 Matthews, J. A., 317 Maughan, B., 123, 151 May, C. P., 591,601, 603, 636, 644 Mayhew, D., 26, 29, 47 Maynard, C., 511, 514, 516, 519,540 Mayo, R., 31,49 Mazur, D. J., 16, 49 Mazzoni, G., 108, 114, 184, 187, 191, 192,214,215,216, 232,235,244,251,252,258,325,326,336,337 McAdams, D.P., 207,232 McAllister, H. A., 136, 214,225 McAuliff, B. D., 425,433,441,448,525,540 McBride, W. J., 305,319 McCabe, D.P., 611,624 McCann, C. D., 219,229 McCann, D., 41,46

686

McCann,). T., 177,191 McCarthy, T. T., 178, 189, 583, 598 McCartney, K., 448 McCauley, M. R., 63, 65, 72, 77, 78, 378,394, 441,448, 511,519,540,592,593,599 McClelland, A. G. R., 112 McCloskey, L.A., 433,451 McCloskey, M., 132,151, 164,173, 178,184,191, 214,232 McClure, K. A., 372, 382, 383, 387, 390,391, 392,396, 432,451 McComas Collard, L., 412,424 McConkey, K. M., 178, 191, 214,234, 324, 325, 337 McCormick, C. T., 176, 191 McCoy, K., 511, 542 McCoy, M. L., 595,601, 645 McCranie, E. W., 93, 114 McDaniel, M., 112, 263,294 McDermott Sales, J., 122, 123, 150 McDermott, K. B., 60, 79, 123, 154, 178, 184, 192, 203,234,240,242,243,245,246,247,259,260, 263,267,283,291,295,457,477,480,483,486, 488,489 McDermott, K. B., 501, 505, 506, 592, 602, 605, 607, 624 McDonald, H. E., 206, 230 McDorman, T. A., 112 McDougall, J., 250, 257 McDowd, J. M., 631, 643 McEvoy, C., 213,229 McEwen, T., 160, 172 McFadzen, E., 136, 153 McFarlane, F., 412,424 McGaugh, J. L., 85, 94, 114, 115, 551,572 McGeoch, J. A., 306,318 McGhee, D. E., 34, 46, 205 McGorty, E. K., 97, 112, 120, 149 McGough, L., 418,424,429,448 McGovern, A., 512,537 McGuigan, F., 456, 458, 474 Mcintyre, J. S., 582, 583, 601 McKay, D. G., 306,316 McKee, D. H., 549, 571 McKinley-Pace, M. J., 213,231 McLean, J., 115 McLellan, K., 592, 600 McMahon,M., 70,77,78,378,395, 593,594,601 McNall, K., 47, 181, 190 McNally, R. J., 93, 94, 99, 105, 109, 114, 135, 142, 148, 151,202,207,232,279,280,292,294,310,318 McNichol, S., 458, 459, 474 McSpadden, M., 288, 295 Meade,M.L.,209,232,234, 249,259,260 Means, B., 140, 141, 151 Mecacci, L., 591, 601 Mega, C., 412,422, 457,472 Mehl, M. R., 17, 49, 122, 148 Meiser, T., 32, 47, 49

AUTHOR INDEX

Meissner, C. A., 124,151,180, 181,187, 191, 192,214, 235,286,294,363,368,376,395,586,601 Melinder, A., 456,471,554,572 Mello, E. W., 65, 79, 593, 601, 642, 644 Melnyk, L., 404,406,407,408,409,411,421,424,461, 466,470,549,558,570,668 Melton, G. B., 532, 538 Memon, A., 63, 64, 65, 69, 70, 78, 79, 128, 150, 184, 190, 191,209,211,216,228,229,232,244,249,257,258, 283,294,322,325,326,337,378,395,406,424,429, 448,460,478,584,585,586,587,588,593,599,600, 601,602,628,633,635,637,642,644,645 Merapol, N.J., 48 Merckelbach, H., 83, 97, 112, 116, 185, 190, 222,233 Merritt, K. A., 412,413,414,424,425,455,457,461, 462,467,470,474,475 Messer, D., 268, 294 Messo, J., 87, 114 Metcalfe, J. A., 306,317, 319 Metzger, R. L., 491, 493,505, 506 Meudell, P.R., 211,232 Michel, M. K., 264, 294 Michelli, J., 519,538 Mickes, L., 276, 279, 292 Migueles, M., 204, 228 Mikulincer, M., 21,49 Milberg, W. P., 605, 624 Miles, K. L., 412,424 Miller, B.S., 519, 541 Miller, D. G., 68, 78, 178, 191, 242, 258, 380, 395, 501 Miller, J. B., 24, 25, 49 Miller, L. M.S., 15, 49 Miller, M. B., 488, 505 Miller, N., 160, 172 Miller, R. D., 39, 49 Miller, R. L., 186, 190 Miller, W. R., 282, 294 Milling, L. S., 214,235, 337 Millis, K. K., 12, 46 Millward, C., 268, 270, 294 Milne, A. B., 207,236 Milne, R., 53, 64, 65, 69, 70, 72, 73, 79, 265, 294, 593, 600 Minakowska, !., 233, 602 Mineka, S., 218,232 Minshew, N.J., 266, 268, 294, 296 Mirani,J., 131, 151 Mitchell, K. J., 380, 383, 397, 501, 588, 601, 618, 620, 624,625 Mitchell, S., 443,447, 450, 456,473, 477 Moan, S. F., 412,426, 455,477, 559,574 Moffitt, T. E., 125, 150 Mohler, L., 104, 111 Mojardin, A. H., 314,317, 487,498,504 Monaco, G. E., 19,46 Montgomery, N., 177, 190 Moore, D., 288, 295 Moore, M., 578, 584, 601

AUTHOR INDEX

Moore-Russell, M., 570 Moores, L., 139, 152, 411,425, 456,476 Moran, P. B., 519,541 Morgan, C. A., 125, 154 Morgan, C. A. III, 93, 97, 98, 99, 100, 102, 114 Morgan, D. M., 276, 279, 292 Morgan, J., 574 Morison, S., 513,540 Moritz, S., 243, 259, 270, 272, 273, 274, 294, 295 Morrell, R., 583, 602 Morris, C., 542 Morris, L.A., 514,534 Morrison, A. P., 271,295 Morse, M. B., 460,474 Mortimer, A., 170, 173 Moscovici, S., 531,540 Mosmann, A. L., 346, 369 Mosteller, F., 15, 47 Moulin,C.J.A.,634,636,645 Mower, 0. H., 319 Moynan, S. C., 305, 308, 320 Mueller-Johnson, K., 213.232, 589, 601 Muente, T. F., 112 Muhlenbruck, L., 44, 363, 365 Mulder, M. R., 409,425, 557, 572 Muller, R. T., 524, 540 Multhaup, K. S., 583, 585, 601 Mundorf, N., 28, 49 Munsterberg, H., 178, 191, 241,259 Murachver, T., 457, 458, 468, 474 Murnane, K., 34, 42 Murphy, D. R., 23, 33, 43, 50 Murphy, G. H., 265, 293 Murphy, J. M., 306, 319 Murphy, K. j., 591, 603 Murphy, M. D., 248, 259, 582, 602 Murray, j. E., 120, 151, 410,426, 456,460,463,473, 477 Mutter, C. B., 335 Myers, B., 334, 336 Myers, D. G., 531, 540 Myers, D. M., 99, 111 Myers, J. E., 510, 511, 515, 518, 520, 521, 527,540, 541 Myers, J. E. B., 509, 540, 548,554, 562,563,564,571, 572, 573, 574 Myers, j. L., 22, 43 Myers, N. A., 454, 474 Myhr, R., 554, 572 Myles-Worsley, M., 436, 448, 459, 474 N

Nachson, 1., 117, 119, 122,126, 132, 133,149, 151, 154 Nadel, L., 82, 103, 114, 303,305, 308,309, 318, 319 Nader, K., 124, 153, 219,232, 233, 253,259 Nakamura, G. V., 34, 42, 203, 204, 205, 225, 228, 242, 259,459,470 Nakao, M.A., 16, 49

687

Nance, N., 587, 598, 616, 622, 627, 643 Narby, D. J., 136, 147 Narchet, F. M., 181, 192 Narens, L., 633, 645 Nash, M., 244, 258 Nash, M. R., 325, 337 Nash, M. J., 324, 326, 337 Nathan, D., 73, 79, 178, 179, 191 Nathanson, R., 435,449, 550, 551, 552, 573, 574 Nauful, T., 555,571 Neely, J., 304,316 Neisser, U., 25, 49, 84, 94, 114, 123, 132, 133, 135, 141, 14~ 151,152,162,173,233,454,456,475 Nelson, K., 141, 150, 154, 436,448,454, 455,456,457, 458,459,471,473,474,475,477 Nelson, T. 0., 288, 295, 633, 645 Nemeroff, C., 82, 93, 113 Ness, H., 225 Neuberg, S. L., 232 Neufeld, P., 170, 173, 176, 192 Neuliep, j. W., 28, 46 Neuschatz, J. S., 204,231, 233, 243, 246, 247, 248, 254, 255,258,259,324,332,336,337,488,502,505,50~

587,601 Newcombe, N. S., 461, 471 Newcombe, P. A., 441,448 Newlands, P. J., 66, 67, 79 Newman, C., 225 Newton, j. W., 87, 112, 122, 149 Nicholas, E., 412,426, 455,477, 559,574 Nicholls, T., 128, 153, 372, 395 Nickerson, M., 511, 540 Nickerson, R. A., 86, 114 Nicolaou, A. L., 155 Nida, R. E., 461,470, 472, 571 Niedenthal, P. M., 203, 229 Niederhoffer, K. G., 17, 49 Niemann, Y. F., 518,542 Nightingale, N. N., 511,512,514,515,519,523,527, 534,536,540 Nijman, H., 93, 112 Nisbett, R. E., 479, 505 Nogrady, H., 325, 337 Noice, H., 119, 152 Noice, T., 119, 152 Nolde, S. F., 103, 113 Noll, j. A., 295 Noon, E., 72,76,525,536,550,565,570 Norman, K. A., 254, 259, 585, 602, 606, 607, 610, 624 Norton, G. R., 319,335 Norwick, R. j., 187, 191 Nasworthy, G., 600 Noyes, j., 635, 643 Nugent, K., 218,232 Nunez, N., 39, 52, 525, 542, 556, 563, 572, 574, 595, 596, 601,645 Nunn, J. A., 162, 174

688

Nunno, M., 426, 451 Nussbaum, J. F., 20, 22,49 Nysse-Carris, K. L., 511,524, 530, 531,535, 541

0 O'Sullivan, J. T., 350, 364 Oakes, M.A., 201, 229 Oates, K., 436, 437, 448 Ober, B. A., 296, 583, 598 O'Callaghan, J., 569, 569 O'Callaghan, M., 511, 525, 536 Ochsner, K. N., 224 Oesterlund, K., 28, 50 Offermann, J. T., 213, 237 Ofshe,R.J.,40,49, 176,177,178,179,182,187,191, 192,202,216 Ogden, D., 532, 538 O'Gorman, J. G., 85, 113 Ohnishi, M., 16, 49 Olde, B. R., 26, 46, 196,228 Oliva, P. S., 202, 233 Olson, E. A., 157, 161, 165, 167, 169, 173, 211,237, 385, 386,392,393,397,587,598 Olson, J. M., 206,234, 526, 541 Olsson, N., 373, 374, 375,395 Ondersma, S. J., 516,540 Oorsouw, K. V., 222, 223, 233 Oran, D., 550, 569 Oran, H., 549, 550, 569, 573 Orbach, Y., 113, 136, 152, 353,354,359,360,367, 415, 418,424,425,430,433,437,438,439,440,441,443, 444,445,446,447,448,450,456,460,464,465,466, 468,470,473,475,476,477,478,523,541,569 Orcutt, H. K., 512,525,531,538,540, 565,566,571,573 Orne, E. C., 189, 323, 324, 327, 328, 337, 338 Orne, M. T., 189, 323, 324, 327, 328, 330, 336, 337, 338 Ornstein, P. A., 105, 114, 118, 146, 148, 152, 214,233, 264,294,405,412,413,421,424,425,440,442,444, 448,454,455,456,457,458,459,461,466,467,468, 469,470,472,474,475,571 Orr, S. P., 94, 115 Orsillo, S.M., 142, 152 Oskamp, S., 214,232 Ost, J., 84, 114, 131, 133, 152, 182, 192 Ostrom, T., 209, 225 Owens, J. L., 457, 471 Owens, M., 82, 93, 113 Owens, R. G., 514,536 Owre, K. P., 589,601 Ozbek, I. N., 549, 574 p

Paas, F., 119, 154 Paddock, J. R., 244,259 Padilla, J., 65, 77 Paivio, A., 60, 79

AUTHOR INDEX

Pallier, G., 375, 395 Palmer, J. C., 48, 60, 78, 213,232, 615, 624 Palmore, E., 581, 597 Pansky, A., 103, 114, 240,258 Papa, A., 29,42 Papalia, D. E., 580, 581, 601 Papierno, P. B., 419,422 Papotto, C., 104, 111 Pardeck, J. T., 550, 573 Paris, S. G., 435,448, 481, 482,483, 505 Park,B.,200,201,226,237,377,396 Park, D. C., 582,601, 609,623 Parker, H. A., 310,318 Parker, J. F., 73, 79, 93, 111, 122, 123, 125, 132, 134, 150,152,462,469,471,475,511,540 Parkin, A., 284, 288, 295 Parkin, A. J., 29, 49, 630, 632, 645 Parks, T. E., 36, 49, 137, 138, 152 Parrott, D. J., 349, 369 Pasupathi, M., 32, 49, 123, 152, 220, 221, 233 Patel, S., 176, 190 Patrick, C. J., 286, 292, 344, 345, 368 Payment, K. E., 385, 397, 501, 507 Payne, B. K., 201, 233 Payne, D. G., 243, 246, 247,258, 259, 306, 319, 334, 336, 488,489,501,502,504,505 Payne, J.D., 82, 92, 95, 96, 97, 98, 99, 103, 114 Payne, M. F., 519, 521,536 Pazzaglia, P., 288, 295 Pea, R. D., 433, 448 Peace, K. A., 93, 96, 99, 114, 135, 152 Pears, K. C., 413,425 Pecchioni, L. L., 20, 49 Pecher, D., 59, 79 Pedersen, A. C. 1., 219, 233 Peeck, V. H., 119, 154 Pelham, B. W., 29, 45 Peluso, J.P., 213,237, 277,296 Pence, D., 40, 49 Pendergrast, M., 522,541 Pennebaker, J. W., 17, 18, 50, 122, 148, 152, 319 Pennington, B. F., 266,291 Pennington, N.H., 197, 199,229,233,513,538 Penrod, S.D., 67, 78, 97,112,121,149, 174,372,375, 376,377,378,394,395,396,511,513,528,531,536, 538,539,541,585,587,598,601 Pentland, J., 230, 243, 244, 249,258, 325, 336, 401 Percer, J. M., 488, 505 Pereira, J. K., 455, 477 Perfect, T. J., 629, 630, 632, 633, 634, 635, 638, 645 Perkowitz, W. T., 16, 44 Perlini, A. H., 333,338, 379, 396 Perner, J., 455, 466, 467,475 Perris, E. E., 454, 474 Perrott, M.A., 349, 369 Perry, C., 323, 324, 325, 328, 330, 331, 336, 337 Perry, A. R., 516, 536 Perry, D. J., 75, 76

AUTHOR INDEX

Perry, E., 635, 645 Perry, N. W., 405,415,425, 426, 429,432,433, 434,448, 449,451 Peruche, B. M., 201, 233 Pesta, B. J., 248, 259 Peters, D.P., 91, 114, 353, 369, 462,475, 552,573 Peters, T. J., 286, 292, 293 Peterson, C., 93, 114, 118, 138, 139, 140, 150, 152, 155, 405,407,410,425,454,455,456,461,462,463,473, 475,592,600 Petit, T. L., 630, 644 Pettersen, T., 554, 572 Pettibon, S., 188, 189 Petursson, H., 186, 190 Pezdek, K., 26, 27, 49, 67,76, 77, 93, 111, 114, 126, 127, 132,133,135,141,150,152,204,215,230,233,244, 248,249,250,251,255,257,258,259,260,410,425 Pfeifer, J. E., 166, 173 Pfeifer, R. L., 362, 365 Phelps, A.M., 491,506 Phelps, N. Z., 491,506 Philippot, P., 149 Phill,C.,469 Phillips, L. H., 63 7, 644 Phillips, P. L., 584, 600 Piaget, J., 482, 483, 506 Picariello, M. L., 461, 476 Pichert, J. W., 59, 76, 665, 668 Pichora-Fuller, M. K., 12, 23, 50 Pickel, K. L., 91, 115, 223,233, 376, 381,395 Pickering, A., 88, 116 Pickett, C. L., 26, 28, 45, 198, 228 Pickrell, J. E., 243, 244, 249, 250, 253, 259 Pierce, B. H., 320 Pihl, R. 0., 283, 284, 286,291 Pillemer, D. B., 121, 122, 132, 152, 454, 461, 475, 476 Pipe, M. E., 74, 76, 105, 111, 115, 118, 120, 141, 151, 217,234,407,410,412,417,422,425,426,427,433, 440,449,450,456,457,460,461,463,464,465,466, 467,468,470,471,472,473,474,476,47~478,482,

506,523,535,541,548,559,565,573 Piper, A., Jr., 202, 233 Pisoni, D. B., 246,259, 457,477 Pitman, R. K., 115, 280,292, 310,318 Planalp, S., 12, 22, 49 Plant, E. A., 201, 233 Platt, R. D., 277, 278, 295 Platz, S. J., 376, 395 Pleydell-Pearce, C. W., 121, 123, 149, 162, 172 Podell, D. M., 521,541 Pohl, R. F., 206, 230 Potage, D. C., 222, 233 Polczyk, R., 213,233, 589, 590,602 Polotsky, P., 82, 93, 113 Pontoski, K. E., 387, 388, 396 Poole, D. A., 74, 75, 79, 108, 115, 120, 124, 125, 127, 136,138,140,153,243,256,257,322,337,401,405, 406,408,410,411,412,413,415,421,422,425,429,

689

430,433,437,440,441,442,449,460,464,476,480, 482,504,506,529,538,560,573 Poole, J. H., 296 Poon, L. W., 582, 602 Pope, H. G., 202, 233 Pope, H. G. Jr., 202, 233 Port, L. K., 423 Port, L. P., 571 Porter, A., 128, 155 Porter, S., 93, 96, 99, 108, 114, 115, 128, 134, 135, 146, 152,153,155,167,173,184,192,214,233,244,249, 254,259,282,297,339,348,358,362,363,364,368 Posnansky, C. J., 482, 505 Postman, J., 14,42 Postman, L., 14, 40, 42, 304, 319 Potenza, M., 503, 505 Potter, R. F., 29, 42, 129, 147, 167, 172 Powell, G. J., 517,541 Powell, J. W., 189 Powell, M. B., 75, 79, 141, 152, 153,214,237,412,424, 458,459,460,461,468,476,477 Powell, S., 268, 294 Pradere, D., 260 Prado, L., 423, 571 Pratt, C., 466,476 Prawatt, R. S., 481, 482, 506 Pressley, M., 455, 477 Preston, E. L., 134, 204,233, 248,259, 601 Price, D. W. W., 440, 449, 465, 476, 548, 559, 573 Price, H. L., 117,129,137,141,147, 148,153 Price, J. D., 491, 493, 505, 506 Price, M., 455, 477 Prichert, J. W., 224 Priestley, G., 463, 476 Prince, S., 623 Principe, G. F. A., 105, 114, 456, 457, 468, 475 Prizmich, L. P., 510,540, 554,573 Pruett, J. C., 461,476 Prull, M., 26, 27,49 Prull, M. W., 614, 623 Pryce, J., 510,536 Pryor, J. B., 198, 225 Przeworski, A., 280, 297 Ptacek, J. T., 29, 49 Puglilsi, J. T., 582, 602 Purdie, V. J., 201,227 Putnam, F. W., 276, 284,291, 296, 302 Putnam, W. H., 327, 337 Putnick, M. E, 429, 445 Pynoos, R. S., 120, 124, 153, 253, 259 Q Qin, J. J., 213,227,410,413,422,442,446,466,476, 493,505,511,514,535,541,571,605,623 Quas, J. A., 105, 115,414,416,423, 425,455,460, 464, 466,467,472,476,509,511,514,516,530,535,538, 541,545,548,549,550,551,556,559

690

Quas, J. A., 567,570, 571, 572, 573 Quig, M. B., 636, 644 Quigley, C. A., 334, 338, 379, 397 Quigley, K. L., 67,77 Quinton, N., 587, 598, 616, 622, 627, 643

R Rabbitt, P.M. A., 630, 646 Rabinowitz, J. C., 583, 602 Racine, C., 243, 260, 270,295, 495,507, 605, 611, 624 Rackstraw, P., 237 Radvansky, G., 636, 646 Ragan, P., 284, 296 Rahhal, T. A., 591,599 Raju, S., 214,237 Ranalli, R., 404, 425 Range, L. M., 206, 228 Rarity, J., 225 Rashid, S., 587, 603 Raskin, D. C., 341,342,344,355,356,359,367, 368, 430,438,445,449 Rasmussen, C., 266, 268, 293 Rassin, E., 358, 359, 360, 364, 368 Ratner, H. H., 35, 45, 50,4 55, 476 Ray, S., 285, 295 Raye, C. L., 34, 216, 230, 254, 258, 271, 293, 361, 367 Raymond, D. S., 54, 59, 64, 77 Raymond, P., 198, 225 Raz, N., 611, 624 Read, J.D., 105, 108, 113, 115, 117, 118, 119, 126, 128, 134, 135, 136, 140, 142, 146, 148, 149, 151, 153, 184, 191,214,215,217,231,233,243,256,258,259,260, 283,287,288,295,339,367,372,392,395,396,486, 506,522,539 Ready, D. J., 327, 337 Reddy, C., 623 Reder, L. M., 214, 224,483, 504 Redfield, D., 516,519,537 Redlich, A. D., 186, 188, 192,450,510, 511,526,540, 541, 554, 571, 573 Reed, A. W., 491, 493, 505, 506 Reed, D. S., 436, 446 Reed, R. S., 408,423, 441,446 Reese, E., 455, 476 Regan,P.C.,521,522,539,541 Reich, P. A., 433, 449 Reich, R. R., 295 Reid, J. E., 160, 173, 177, 180 181, 190, 223,230 Reilly, K. D., 291 Reinhart, M., 79, 574 Reinitz, M. T., 204, 205, 229 Reinke, P., 24, 26, 28, 48 Reisberg, D., 84, 85, 87, 91, 94, 106, 109, 112, 113, 114, 115, 178, 189 Reiser, M., 322, 337 Reisman, S., 379,396 Reiss, S., 319

AUTHOR INDEX

Reiter-Lavery, L., 367 Relyea, M., 187, 192 Remen, R., 198, 200, 233 Renick, A., 405,421, 442,444,464,470, 559,570 Renkey, M. E., 231 Renner, B., 21,50 Renner, P., 266, 295 Reppucci, N.D., 532,535, 541, 555, 571 Resick, P. A., 319 Revelle, W., 104, 115 Reyna, V. F., 118, 127, 147, 153, 214,225, 256,257, 277, 295,314,316,437,442,444,457,458,470,479,480, 481,483,484,485,486,487,488,489,490,492,493, 496,497,498,499,500,501,502,504,505,506,507, 612,622 Reynolds, K., 204, 233, 259 Reynolds, K. D., 216,235 Reysen, M. B., 211,234 Rhodewalt, F., 21,50 Rholes, W. S., 219, 229 Rhue, J. R., 324, 336 Richards, J. M., 29,50 Richardson-Klavehn, A., 373, 393, 631,644 Richeson, J. A., 198,234 Riddlesberger, M. M., 414,423, 455,464,472, 559,571 Rideout, R., 454, 475 Ridley, A.M., 213, 214,234 Riggio, R. E., 24, 50 Riha, J., 242, 259 Rind, B., 516,540,541 Rips, L. J., 164, 172 Rizzo, L., 270, 292 Roberu,K.P., 141,153,352,369,410,416,425,433, 434,435,441,449,458,459,460,461,476,477, 484,506 Roberts, M.A., 243, 257 Roberts, S., 463, 476 Robertson, D. A., 384, 395 Robertson, S. W., 224 Robertson, T. E., 202, 232 Robertson, W. 0., 16, 50 Robinson, J.D., 20, 49, 442, 449 Robinson, K. J., 487, 494,506, 607,624 Robinson, M.D., 384, 391, 395 Robles, B., 381, 392, 396 Rocchetti, G., 591,601 Rocci, P., 592, 600 Rodd, Z. A., 305,319 Rodriguez, 462,470 Roe,C.,250,259,410,425 Roebers, C. M., 214,234, 381,391,392,395, 442,444, 456,466,467,470,471,477 Roediger, H. L., 60, 79, 123, 136, 137, 154, 178, 184, 192,203,209,232,234,240,242,243,244,245,246, 247,249,257,258,259,260,267,270,273,283,291, 295,457,477,480,483,486,487,488,489,494,501, 502,505,506,558,573,592,602,605,607,624 Roemer, L., 142, 154

AUTHOR INDEX

Roese, N.J., 206, 234 Rogers, C. M., 514, 541 Rogers, M., 255, 257 Rogers, S. J., 266, 291 Rojas, E., 283, 295 Romanczyk, A., 410,424, 510,519,539, 548,572 Romeo, R., 214,237 Roncolato, W., 464, 477 Roper, R., 214,234 Rose, R., 585,601,628,645 Rosen, M. J., 585, 603, 605, 606, 612, 624, 625, 633, 646 Rosenberg, E., 348, 349, 368 Rosenfeld, P., 346, 347, 368 Rosenhan, D. L., 202, 234 Rosenhow, D. J., 283, 294 Rosenthal, R., 23, 48, 440, 449 Ross, C. A., 319 Ross, D. F., 184, 189, 192,213,214,226, 241,257, 406, 421,441,445,450,457,470,511,519,522,525,538, 539,541,563,573,594,602,627,645 Ross, L., 479, 505 Ross, M., 36, 43, 123, 132, 134, 154, 207, 218,237 Rossiter, L., 216, 224 Rotella, C. M., 500, 506 Rothbart, M., 202,234 Rothbaum, B., 317 Rothman, M. B., 577,602 Rovee-Collier, C., 214, 234, 454, 463,472, 473, 477 Rowe, D. M., 515,535 Rowland, J. H., 48 Royal, D., 582,601 Rozelle, R. M., 516,540 Rubin, D. C., 118, 120, 123, 126, 130, 132, 133, 135, 147,154,162,173,207,234,373,374,396 Ruby,C.L.,353,354,355,356,360,368 Ruderman, A. J., 33, 51 Rudnicki, A. G., 530, 535 Rudolph, W., 19, 50 Rudy, L., 108, 112,408,412,423,425, 441,446, 457, 462,468,472,477,556,571 Ruff, C. C., 272, 295 Ruffin, M. T., 16, 51 Ruffman, T., 466, 467, 475 Rumpel, C., 167, 172 Russano, M. B., 181, 192 Russell, D., 515,537 Russell, J. A., 22, 43, 202, 225 Rutter, M., Ill, 123, 125, 150, 151 Ryan, C. S., 377, 396 Ryan, E., 570 Ryan, R. H., 380, 387, 392, 396 Rybarczyk, B. D., 582, 602 Rydell, S. M., 68, 80

s Sable, H. J., 305, 319 Sachsenmaier, T., 571

691

Saenz, D. S., 28, 48 Saetermoe, C. L., 408, 422, 549, 570 Safer, M.A., 28, 38, 43, 50, 132, 135, 136, 137, 138, 148, 151,154,207,230,234,463,477 Saks, M., 330, 335 Sales, J. M., 122, 123, 150, 462,471 Salmon, K., 111, 217,234, 407,425, 426, 455, 456,458, 461,464,466,467,474,476,477,478 Salovey, P., 202, 234 Salthouse, T. A., 631, 644 Sameroff, A. J., 490, 506 Samuels, C., 216,236, 253,260 Sanchez, R. P., 511, 512,519,522, 523,525,537, 563,571 Sanders, G. S., 327, 337, 379, 396 Sanders, J., 330, 335 Sanders, R. E., 159, 173, 248, 259, 582, 602 Sands, S. H., 319 Sandys, M., 531,541 Sanson-Fisher, R. W., 15, 46 Sarfati, D., 571 Sargant, W., 300, 319 Sas, L. D., 549, 550, 567, 568, 573, 574 Sattler, J., 430, 449 Saucedo, C., 282, 294 Saults, J. S., 631, 644 Saunders, F. F., 568,572 Savitsky, K., 169, 172, 180, 190 Sawyer, J.D., 27, 46, 199,228 Saxe, G. N., 319 Saxe, L., 345, 347, 369 Sayette, M. A., 285, 294, 349, 369 Sayfan, L., 456,471 Saywitz, K. J., 65, 74, 76, 79, 412, 426, 434, 435, 436, 437,444,449,455,464,465,468,477,549,550,551, 552,555,556,559,560,572,573,574 Scarberry, N.C., 136, 147,463,469 Scarpa, A., 413, 426 Schaaf, J. M., 150, 412, 413, 420, 422, 423, 460, 461, 466,472,477,538 Schachter, D. L., 209, 272, 296 Schacter, D. S., 5, 23, 32, 34, 47, 50, 82, 93, 115, 123, 154,184,191,206,207,216,217,231,234,243,245, 254,258,259,260,274,275,276,280,283,284,291, 292,294,295,310,318,495,500,505,507,583,585, 600,602,605,606,607,610,611,613,623,624,632, 645 Schafe, G. E., 219,233 Schaller, M., 43 Schank, R. C., 199, 235 Scheck, B., 170, 173, 177, 192 Scheflin, A. W., 129, 148,307,316,324, 329,330, 335, 337 Schein, E. H., 178, 192 Scheiner, J. L., 511,531, 536,539,541 Scherer, K. R., 351,366 Schewe, P., 511,540 Schick, K., 410,426,456,477 Schieber, F., 580, 602

692

Schimossek, E., 353, 367 Schlagman, S., 131, 151 Schmidt, C. W., 541 Schmidt, E., 511,512,535 Schmidt, H. G., 119, 154 Schmidt, J., 342, 366 Schmidt, S. R., 29, 50, 85, 115 Schmitt, F. A., 582, 602 Schmitter-Edgecombe, M., 38,50 Schmolck, H., 84, 115, 131, 133, 154 Schneider, B. A., 50, 114 Schneider, W., 455, 466, 467,477 Schneier, I., 179, 192 Schnell, S. V., 550, 574 Scholnick, E. K., 481, 482, 505 Schooler, J. W., 106, 115, 124, 151, 214, 221,235, 253, 260,286,294,301,307,309,319,381,396 Schrauf, R. W., 123, 154 Schreiber, N., 73, 79 Schreiber, T. A., 213, 214,227 Schul, Y., 31,49 Schuller, R. A., 511, 526, 541 Schultheis, J. A., 376, 393 Schuman, H., 120, 147 Schutte, J. W., 511,541 Schuurmans, E., 221,235 Schwartz, B. L., 214, 235, 306, 315 Schwartz, J. L. K., 35, 46, 205 Schwartz, R., 571 Schwartz-Kenney, B. M., 109, 112, 441,446, 556,571 Schwartzman, D. F., 216,235 Schwarzin, H. J., 517, 542 Schwarzmueller, A., 118, 140, 150, 413,422, 454,461,471 Scialfa, C. T., 580, 600 Scoboria, A., 187, 192, 214,235, 324,325,332,333,337 Scogin, F., 375, 376, 396, 587,602, 635, 645 Scott, E., 547,571, 574 Scott, K. L., 135, 148 Scrivner, E., 136,137,154,463,477 Scullin, M. H., 412,419,426,470 Sealy, A. P., 528, 542 Seamon, J. G., 243, 256,257, 260 Searcy, J. H., 584, 585, 589, 602, 635, 636, 637,642,645 Searle, J. R., 16, 50 Sedikides, C., 22, 43, 220, 235 See,S.T.K.,23,50, 100,210,230,595,600,627,644 Seelau, E. P., 68, 80 Seelau, L., 633, 646 Seemungal, F. V., 75, 79 Sego,S.A.,511,512,522,525,537,563,571 Seifert, C. M., 369 Seitz, A., 108, 114, 184, 191, 215,232, 326, 336 Self, G., 249, 260, 385, 397 Semin, G. R., 19, 44, 212,235 Semmler, C., 211,235 Sepler, F., 514, 542 Sergent-Marshall, S., 213, 214,227, 275,296

AUTHOR INDEX

Sesco, B., 78, 408, 412,424 Seying, R., 510, 536 Seymour, T. L., 346, 369 Shaffer D. R., 528, 536, 542 Shafto, M. G., 346, 369 Shank, R., 437, 450 Shannon, L., 162, 173 Shapiro, C., 509, 538, 540, 548, 571 Shapiro, D. A., 15, 47 Shapiro, E., 509, 543 Shapiro, L. B., 118, 152 Shapiro, L. R., 413, 426, 461, 475 Sharkansky, E. J., 150 Sharkey, W., 15, 51 Sharma, K., 372, 393, 395 Sharman,S.A.,214,236 Sharman, S. J., 216, 217,235, 244,258 Shaughnessy, M. F., 305,314,319 Shaver,P.R.,21,29,45,49,510,514,535,538 Shaw, J. I., 531, 542 Shaw,J.S., 74,77, 209,222,235,372,381,382,384, 387,388,389,390,391,392,396,404,422,442,

446,553,571 Shaw, L.A., 595,601, 645 Sheard, D. E., 126, 154 Sheehan, P. W., 179, 184, 192, 324,325,336, 337, 379,396 Sheen, M., 215,230 Sheesley, L.A., 213,231 Sheffield, E. G., 463, 477 Sheibe, R., 438, 445 Shelton, J., 505 Shenaut, G. K., 295 Shepherd, E., 170, 173 Sherman, D. A., 20, 44, 215, 227 Sherman, J. W., 205, 235, 244, 257 Sherman, S. J., 184, 190, 216, 235 Shevell, S. K., 164, 172 Shewan, D., 214,234 Shilling, V. M., 630, 646 Shimamura, A. P., 222, 226 Shinozuka, M., 49 Shobe, K. K., 93, 105, 115, 135, 154, 280,291 Shrimpton, S., 436, 437,448 Shum, M.S., 121, 122, 154 Shuman, D. W., 458, 474 Shute, R., 458, 474 Siegal, M., 441, 448 Sifonis, C. M., 605, 624 Sigurdsson, J., 186, 192 Sillars, A. L., 28, 30, 50 Silva, P. A., 125, 150 Simcock, G., 454, 455, 477 Simmons, W. L., 203, 328, 337 Simon, E., 582, 602 Simon, H. A., 30, 41, 44, 456,470 Simons, D. J., 203, 235

AUTHOR INDEX

Simpson, A. L., 38, 50 Simpson, J. A., 21,51 Simpson, M. A., 264, 294 Sinclair, G. P., 119, 150 Singer, S., 28, 50 Sinisi, C. S., 514,534 Sitarski, J., 275,291 Sivec, H. J., 326, 337, 338 Sivers, H., 300, 309, 319 Skolnick, P., 531,542 Skowronski, J. J., 121, 155, 162, 163, 174,209,225 Slackman, E., 141, 154, 459,477 Slater, E., 300,319 Slaughter, S. J., 34, 46 Slawinski, E. B., 581, 602 Siemon, A., 29,43 Slobogin, C., 189, 192 Small, M. Y., 174, 481, 482,507 Smiley, S. S., 482, 504 Smith, A. D., 582,601, 607,611,624 Smith, B. W., 39, 52,291, 525,542, 563,574 Smith, C., 82, 115, 515, 537 Smith, E., 73, 76, 148,413,421,442,445 Smith, H., 514, 516,542 Smith, M. C., 126, 132, 133, 154 Smith, M. J., 270,291 Smith, N.C., 29, 43 Smith, P. J., 244, 258 Smith, S.M., 21, 51, 68, 79, 104, 115, 303, 305, 306, 307, 308,31~320, 605,624 Smith, V. L., 204, 214,235, 511, 526,540, 541 Smolin, J. M., 29,47 Snedeker, M., 78, 178, 179,191 Snow, B., 548, 574 Snyder, H. N., 509, 542 Snyder, L. D., 202, 205,226, 235, 393,435,437,449, 465,468,477,560,574 Snyder, M., 169, 173,236 Sobus, M. S., 532, 536 Socall, D., 31,47 Soraci, S. A., 262, 291 Sorensen, T., 548,574 Sorenson, E. R., 348, 366 Souchay, C., 634, 646 Soukara, S., 255,260,357, 361,369 Souphasith, S., 248, 257 Southwick, S.M., 114, 125, 134, 142, 154 Spaniol, J., 34, 50, 205, 236 Spanos, N. P., 216, 236, 253, 260, 326, 333, 338, 379, 396, 511,537 Spaulding, K., 387, 394 Speaker, C. J., 454, 474 Spellman, B. A., 302,304,316 Spence, M. J., 214,236,410,426,460,466,478 Spencer, C., 557, 574 Spencer, J. R., 429, 450 Spencer, L., 184, 192, 214, 233

693

Spencer, W. D., 611, 624 Sperber, D., 22, 51 Spieker, B., 412,424, 455,474 Spiecker, S.C., 197,236 Spiegel, D., 306,318,320,329,337 Spiegel, H., 329, 337 Spiegel, J.P., 300,317 Spiegel, R., 460,474 Spitz, H. H., 262, 295 Spjut, E., 65, 78 Sporer, S. L., 67, 79, 124, 151, 286,294,361,362,364, 369,372,396 Spritz, B., 467, 469 Squire, L. R., 84, 115, 131, 154 Srull, T. K., 31, 47, 196, 197, 198, 199, 203,221,237 Stafford, J., 336 Stafford, L., 15, 24, 28, 51 Stahelski, A. 1., 169, 172 Stahl, T., 342, 366 Stanny, C. J., 87, 115, 128, 154 Stapelton, J., 348, 368 Statham, D., 184, 192 Steblay, N.J., 87, 88, 91, 116 Steblay, N. M., 68, 80,324,325,327,338,379, 396 Steele, C. M., 286, 295 Stein, N. L., 464, 474 Steinberg, A.M., 121, 124, 153 Steinberg, E. P., 581, 602 Steinberg, L., 547, 571, 574 Steinberg, M., 320 Steketee, G., 317 Steller, M., 352, 353, 354, 355, 356, 359, 360, 369, 441,450 Stellwagen, L., 509,543 Stephenson, G. M., 136, 137, 149, 211,236, 384, 391, 396,406,422,437,442,445 Stem, C., 411,417,426 Stem, L. B., 136, 137, 149, 463, 471 Stern, W.,239,260,402,403,411,417,426 Sternberg, K. J., 105, 113, 152, 353, 354,359, 360,367, 408,416,417,418,423,424,425,430,432,434,435, 437,438,439,440,441,443,444,445,446,447,448, 449,450,456,464,465,468,473,475,47~569

Sterns, H. L., 580, 601 Stevenage, S. V., 70, 75, 79, 585, 600 Stevenson, M., 443, 513, 531, 535 Steward, D. S., 405, 407, 426, 559, 574 Steward, M.S., 405,407,426, 436,450, 559,574 Stewart, H., 447, 473 Stewart, T. L., 511, 522, 525, 529,537, 542 Stiles, W. B., 24,51 Stitzer, M. L., 288, 293 Stockwell, T., 286, 293 Stokes, M.A., 412,424 Stollery, B., 634, 645 Stoltzfus, E. R., 591, 601 Stone, J. I., 213,231

694

Stracciari, A. S., 288, 295 Strack, F., 237 Strater, L., 614, 622 Strawbridge, C. P., 291 Street, R. L., 15, 51 Strier, F., 532, 542 Stromwall, L. A., 363, 366, 369 Strong, J. W., 8, 51 Stroud, J. N., 376, 397, 586, 603, 637, 646 Studebaker, C. A., 204, 235 Stuen, C., 580, 602 Sturgill, W., 373, 393 Stuss, D. T., 591, 603, 630, 635, 646 Stuve, T. E., 372,375, 377,394, 395, 587,601 Suengas, A. G., 216,230, 361,367 Sugarman, D. B., 529, 542 Sullivan, A. L., 275,291 Sullivan, E., 516,540 Sullivan, K., 263, 295 Sullivan, T. P., 188, 192 Sunshine, N., 450 Supska, M., 233, 602 Surtes, L., 69, 77 Sutherland, R., 214,236, 410,426, 456,458,460, 461,477 Swales, M., 630, 643 Swanson, K., 584, 602, 636, 645 Swartwood, J. N., 462, 478 Sweeney, D. F., 282, 296 Sweeny, B., 454, 473 Swim, J., 511, 513, 525,539, 542 Sykes, R.N., 375, 396

T Tafarodi, R. W., 207, 236 Tajfel, H., 517,542 Talarico, J. M., 126, 132, 133, 154, 373, 374, 396 Talwar, V., 556, 572 Tam, P., 433, 448 Tanaka, J. W., 214,235 Tarn, P., 425 Tarran, M. J., 513,539 Tarrant, A., 436,445 Taska, L., 549, 551,570 Tate, C. S., 549, 574 Tattersall, M. H., 15, 48 Taub, E. P., 423, 571 Taylor, B., 93, 111 Taylor,)., 114, 135,141,152 Taylor, S., 120, 135, 146, 148, 359, 370 Taylor, S. E., 33, 51, 196, 197, 199, 203,228, 236 Tedesco, J. F., 550,574 Tekcan, A. I., 132, 133, 154 Telfer, K., 437,446 Tellegen, A., 277,296 Terr, L. C., 93, 116, 120, 154,311,320,322,338,454,477 Terranova, S., 259

AUTHOR INDEX

Terry, T., 514, 541 Tessler, M., 457, 477 Thierry, K. L., 236, 410,426, 429,433,447,450,460, 464,465,466,477,478 Thissen, D., 26, 27, 47 Thomas, A. K., 184, 192, 216,236 Thomas, K. G. G., 103, 114 Thomas, M., 415,420,426, 451 Thomas, R. L., 248, 257 Thomas, S. W., 412, 424, 442, 447, 538, 540, 565, 571,573 Thomassin, L., 198,236 Thompson, B. E., 456, 459 Thompson, C. P., 37, 48, 121, 155, 162, 173 Thompson, D. M., 41, 51, 592,602 Thompson, S. A., 274, 292 Thompson, T. L., 20, 49 Thompson, W.C., 442, 450, 652, 668 Thomsen, C. J., 202, 235 Thomson, D. M., 59, 80, 141, 152, 153, 306,320,458,459, 460,476 Thorpe, L. A., 558, 573 Tidwell, R. P., 568, 572 Tilden, J., 325, 337, 379, 396 Tindell, D. R., 605, 624 Titcomb, A. L., 488, 507 Tizzard-Drover, T., 138, 152, 155 Tobey, A. E., 409, 412,426, 457, 468, 478, 525,538, 540, 565,571,573 Tobin, T., 323,336 Todorov, A., 220,236 Toglia, M.P., 184, 189, 192, 204, 213, 214,226, 233, 241,243,247,248,256,257,258,259,260,406,421, 441,450,457,470,488,500,507,588,595,600,601, 602,619,623,627,645 Toichi, M., 266,296 Tollestrup, P. A., 88, 89, 90, 109, 116, 127, 136, 138, 153,155,282,287,295,297 Tomes, J. L., 213,236 Toth, S. L., 491,505 Towell, N. A., 66, 79 Townsend, M.A. R., 482,504 Tracy, J. L., 284, 296 Travis, D. M., 306,319 Tredoux, C. G., 637, 646 Trevisan, M.S., 72, 77, 379, 394 Treyens, J. C., 204, 225, 242, 248, 257 Trivelpiece, J. W., 516, 542 Tromovitch, P., 516, 541 Tubb, V. A., 283,294, 511, 530, 542,628, 644 Tubbs, A., 557, 574 Tuchin, M., 380, 387, 397 Tucker, A., 458, 474 Tuckey, M. R., 125, 127, 155, 204,236 Tully, B., 359, 369 Tulving, E., 41, 51, 59, 71, 80, 260,306,320,440,465, 478,592,602,631,646

AUTHOR INDEX

Tun, P. A., 581, 585, 602, 603, 604, 606, 612, 613, 618, 624,625,633,646 Turk, D. C., 202, 234, 236 Turnbull, W., 16,51 Turner, T. J., 199,225,317, 375,397, 605,622 Turnure, J. E., 263, 296 Turtle, J. W., 88, 116, 119, 127, 136, 138, 155, 157, 161, 164,166,167,171,172, 173,236,382,39~646 Tversky, B., 220,227, 232, 236, 380, 387, 397 Tweed, J., 525,535 Tyda, K., 511,531,535 Tye, M. C., 354, 356, 369 Tymofievich, M., 352, 370 Tzeng, 0. C. S., 517,542

u Underwager, R., 331,338 Underwood, B. J., 241,260, 304,319,507, 607,624 Undeutsch, U., 353, 369 Uranowitz, S. W., 205, 236 Urquiza, A. J., 514,542 Uttl, B., 113, 584, 603

v Vaillant, G. E., 123, 155 Valdiserri, M., 632, 645 Valentine, P., 442, 445 Valentine, S., 511,540 Valentine, T., 88, 89, 90, 116, 588,598, 616, 617, 623 Van Abbema, D., 469 van den Bos, K., 218, 236 van den Hout, M., 278, 279, 294 Vander Hart, 320 van der Linden, M., 149 Van Koppen, P. J., 209, 226 van Stegeren, A., 85, 112 Van Tassel, G., 500,506 Van Voorhees, E., 413,426 VanderVeen, S., 283, 295 Vandierendonck, A., 221,235 Vanous, S., 21, 51, 203,227 Varner, L. J., 104, 116 Vela, E., 306, 320, 378, 396 Verfaellie, M., 260, 605, 624 Verhaeghen, P., 582, 603 Vicari, S., 263, 296 Vidmar, N., 518,542 Vieth, V., 509, 542 Villa, M. F., 291 Villaume, W. A., 22,51 Vinogradov, S., 271,296 Viswesvaran, C., 441,448 Vogel-Sprott, M., 285, 293 Vornick, L.A., 214,236

695

Vosburg, S. K., 289,293 Vrij, A., 84, 114, 131, 152, 180, 192, 201,236, 253, 255, 260,351,352,353,354,357,358,359,361,362,363, 364,365,368,369,409,425 Vrij, A., 557,572

w Wade,E.,220,236,306,316 Wade, K. A., 108, 113, 134, 151, 184, 191, 215, 217,231, 244,250,258,260 Wadkins, T. A., 186, 190 Wagenaar, W. A., 37, 51, 84, 112, 119, 133, 149, 155, 209,226,253,260 Wagner, D., 237 Wagner, J. T., 74, 76, 435, 444 Wagner, W. G., 211,236, 391,396, 511,514,516,521, 524,535 Wagstaff, G. F., 53, 70, 72, 78, 322, 325, 336, 338, 376, 378,379,391,394 Wakefield, H., 331, 338 Wakeford, Y., 263, 295 Waldron, V. R., 24,51 Walker, A. G., 417,426, 434,450 Walker, L., 531,539 Walker, N. E., 416, 417, 426, 434,450 Wall,A.,550,570 Waller, C., 112, 302 Walsh, K. K., 582, 602 Walter, B. M., 630, 632, 645 Walther, E., 210, 237 Ward, T., 513,534 Ward, T. B., 605, 624 Warden, D., 437, 446 Warhaftig, M. L., 64, 77 Wark, L., 406, 424 Warner, D., 379, 394 Warren, A. R., 39, 40, 51, 256,257,405,406,410,415, 418,420,421,426,432,433,434,437,440,443,450, 451,462,478,491,505,506,522,525,529,535,538, 539,542,556,557,563,572,574,618,625 Warren-Leubecker, A., 549, 574 Warrington, E. K., 266,291 Wason, P. C., 202, 237 Waterman, C. K., 512,515, 516,518,519,542,557,574 Watson, G. S., 273, 294 Watson, J. M., 243, 259, 260, 267, 275, 294, 295, 296, 483,488,506 Watters, E., 178, 179, 192, 202,216 Watzlawick, P., 17,51 Waung, M., 51 Wearing, M. H., 600 Weaver, C. A. III., 133, 155 Weaver, J., 28,49 Weber, A., 53, 54, 76 Weber, J. G., 34, 42 Wegener,l.,33,34,47,205,230

696

Wegner, D. M., 168, 173,302,304,318, 320 Weickert, T. W., 273, 292 Weinberg, M.S., 516, 517,542 Weinberger, D. R., 273, 274,292, 293 Weingardt, K. R., 178, 192,522,540 Weingartner, H. J., 282, 283, 284, 286, 292 Weinstein, B., 581, 603 Weinstein, E., 183, 192 Weisberg, J., 28, 30,50 Weiss, A. P., 273,296 Weiss, E. H., 550,574 Weiss, R. L., 582,599 Weiten, W., 510, 530,542 Welch-Ross, M. K., 466,478 Welker, J., 574 Wellman, H., 437, 445 Wells, G. L., 53, 68, 80, 157, 159, 160, 165, 167, 169, 172,173,174,211,232,235,237,328,338,372,377, 382,385,386,387,392,393,395,397,587,596,598, 603,633,641,642,644,646 Wenzel, A. E., 130, 154 Wenzlaff, R. M., 168, 173, 302, 320 Werth, L., 237 Wertsch, J., 31,51 Wertz, J. M., 349, 369 Wesolowska, B., 233, 602 Wessel, I., 87, 116, 151 Wesson, M., 464, 478 West, J., 263, 294 West, P., 282, 293 West, R., 591, 603 West, R. L., 582, 603 West, T., 592, 600 Westby, S., 437,445 Westcott, D. R., 127, 150, 204, 229 Westcott, H. L., 118, 155, 432,438,445, 450, 520,543, 567, 568, 572, 574 Westphal, M., 29,42 Wetzel, C. D., 288, 296 Wewerka, S. S., 454, 455, 469 Whalen, N., 93, 114, 118, 152, 455, 461,462, 463,474 Wheeler, M. A., 628, 646 Whetstone, T., 204, 233, 259 Whitcomb, D., 406,413,427, 509,515,536,543, 550, 562,574 White, B., 44, 149,227 White, G., 139, 152, 425, 456, 475 White, L. T., 74, 79, 121,124, 125, 127, 140, 153, 406, 410,413,425,437,449,482,506 White, R., 122, 155 White, S. H., 454, 476 Whitehouse, W. G., 189, 324,328,337, 338 Whiteside, R., 513,536 Whiteside, S., 522, 542 Whitfield, C., 316 Whitman, J. C., 273,294 Whittenberger, G., 393 see Martinez

AUTHOR INDEX

Whittlesea, B. W. A., 136, 155 Wible, C. G., 132, 151, 164, 173 Wicke, C., 286, 292 Widner, R. L. Jr., 306, 320 Wiebe, S., 469 Wiederman, M., 511, 514,516,519,540 Wieringa, B. M., 112 Wigmore,}. H., 9,51 Wiley, J. G., 584, 600 Wiley, R., 326, 337 Wiley, T., 513,531,535 Wilhelm, F. H., 29, 43 Wilkenfeld, M. J., 605, 624 Wilkens, C. E., 382, 383, 387, 390, 396 Wilkie, J., 15, 24, 30, 41,42 Wilkinson, C., 278, 296 Wilkinson, J., 440, 451 Will, A. P., 167, 174 Willey, R., 517, 539 Willham, C. F., 206,226 Williams, A. R., 85, 116 Williams, C. J., 200,227, 516, 517,542 Williams, C. S., 600 Williams, D. L., 266, 268, 296, 491, 505, 506 Williams, J. M.G., 152 Williams, L. M., 307,309,320, 515,517,537, 543 Williams, R. B., 632,645 Williams, T. L., 320 Williamson, N. L., 586,598 Willis-Shore, J., 296 Willoughby, L., 609, 623 Wilsnack S. C., 320 Wilson, A. E., 207, 218,234, 237 Wilson,C.,40,49,440,449,465,476,559,573 Wilson, C. L., 21,51 Wilson, D., 22, 51 Wilson, G. T., 283, 296 Wilson, J. C., 214,237,407,410,425, 427,482,506 Wilson, M. E., 408, 423, 442, 446 Wilson, T. C., 517,543 Windmann, S., 103, 116 Wingfield, A., 581, 585,603,605, 606, 612,624, 625, 633,646 Winkel, F. W., 200, 237 Winman, A., 206, 237, 373, 374, 394, 395 Winningham, R. G., 131, 132, 133, 155 Winograd, E., 152, 213,227, 237, 277,296 Winzenz, D., 68, 76 Witt, C. J., 588, 598, 617, 622 Witte, K. L., 582, 603 Wittenbrink, B., 200, 201,226, 237 Wittlinger, R. P., 119, 147 Wixted, J. T., 119, 120, 122, 130, 140, 155 Wolfe, D. A., 567,574 Wolfe, J., 150 Wolford, G. L., 488,505 Wolfskeil, M.P., 393

AUTHOR INDEX

Woloshin, K. K., 16,51 Wood, J. M., 73, 74, 77, 210,228, 209, 210,228, 381, 396,404,406,422,432,442,446,451,511,542, 553, 571 Wood, J. V., 207,226 Wood, T. L., 595, 600, 626, 644 Woodall, C. E., 39, 40, 51,405,415,418,420,426,432, 451,563,574 Woods, M.G., 93, 114 Woodward, T. S., 243,259, 272, 273,294, 295 Woodworth, M., 146, 362,368 Woolard, J., 571 Worthington, D. L., 197,226 Worthley, J. S., 306,316 Woythaler, K. A, 387,388, 389,390,396 Wright, A M., 593, 594, 603 Wright, D. B., 209, 232, 233, 237, 249, 257, 260, 376, 385,397,586,603,635,637,638,646 Wright, L., 178, 179, 192 Wright, R., 75, 79, 314,316, 487,498,504, 505 Wrightsman, L. S., 177, 178, 187, 191, 192,429,433, 434,449 Wyatt, G. E., 517, 541, 543 Wyer, R. S., 27, 52, 196, 197, 198, 199,203, 221,237 Wyer, R. S.Jr., 19, 26, 32, 36, 38, 52, 196, 197,199,221, 223, 237 X Xu, G., 219,224

y Yandell, B., 169, 174 Yang, C. L., 34, 42, 205, 225 Yarmey, A D., 127, 128, 131, 155, 371,373,374,375, 397,519,543,581,584,587,594,603 Yarmey, M. ]., 375,397 Yeager, C. A, 293 Yeates, G. N., 286, 292 Yehuda, R., 82, 93, 116 Yerkes, R., 92, 116 Yerys, B. E., 291 Yi-Ting, H., 48

697

Yoga, M., 151 Yonelinas, A P., 631, 646 Yoon, C., 591, 603 Youtz, C., 15,47 Yozwiak, J. A., 511, 519,520, 523, 537, 543 Yudilevitch, L., 424, 450 Yuille,]. C., 58, 62, 74, 80, 82, 84, 88, 92, 93, 100, 108, 112, 115, 116, 122, 126, 127, 128, 129, 130, 131, 134, 135, 136, 138, 149, 153, 155, 167, 173,236, 244,259, 278,282,287,289,295,296,29~ 339,340,352,353, 354,355,358,359,365,368,369,370,382,397,430, 434,441,442,447,449 Yzerbyt, V. Y., 169, 172

z Zacks, R. T., 284, 293, 297, 581, 600, 635, 636, 644, 646 Zajac, R., 414,427, 552,563,574 Zani, A, 591,601 Zanni, G. R., 213, 214,232, 237 Zapamiuk, J., 58, 80, 289, 296, 359, 370 Zaragoza, M.S., 178, 184, 191, 213, 214,226,227, 232, 237,380,383,385,397,406,421,442,444,460,478, 501,507,618,624,625 Zavalloni, M., 531,540 Zeelenberg, R., 59, 79 Zelig, A, 122, 126, 132, 133, 151 Zelig, M., 379, 397 Zeltzer, L.K., 549, 567, 570, 574 Zeman, B. R., 456, 469 Zerbe Enns, C., 311, 320 Zerr, T. K., 387, 388, 389, 390, 396 Zhao, Z., 85, 112 Ziesel, H., 531,539 Zietlow, P. H., 28,51 Zillmann, D., 28, 49 Zimmerman, J. A., 242,259 Zlochower, A J., 349, 365 Zoellner, L. A, 280, 297 Zuckerman, M., 230,339,340, 351, 363,370, 602 Zuhova, ]., 45, 228 Zwaan, R. A, 12, 46

Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

Subject Index A Accessibility, 59, 119-120, 127, 301-302, 305, 312, 313, 314,315,325,463,484-485,495,497,498-499, 503,592 Age differences (in memory; in reliability of testimony), 213,375-376,455-456,494-495,498,581-583, 584,585,587,588-589 Aging, 274, 577-578, 579-582, 584, 590-59, 595-596, 597,606,627-631,632,634,635,636, 637-638,643 Alcohol (effects on memory), 69, 93, 94, 105, 182, 186, 214,261-262,281-289,305,638,660 Alibi, 157-171, 185, 187-188, 358, 661, 662-663, 668 and provider, 158-160, 161, 162, 163-164, 165, 166, 167-169 and witness, 158, 159, 161, 164, 165, 170-171 Alzheimer's disease, 261, 274, 290, 634, 660 American Society of Clinical Hypnosis guidelines, 322,331 Amnesia, 93, 94, 104, 105-106, 179, 222-223, 276, 277, 278,299,300,302-303,305,307,308-310, 325-326,454,632 Amygdala, 98, 309 Anatomically detailed dolls, 407, 412, 417, 464, 557,658 Archival events, 126 Archival studies (of police records), 124, 136-137 Arousal, 28-29, 81, 83, 85, 86, 91, 92, 94, 96, 98, 110, 120,143,146,202,283-284,286,287,311,342, 343,345,362,630 Assimilation, 11 Attention, 14, 21, 22, 23-30, 32, 35, 40, 85, 86-87, 90-92, 95, 97, 101-102, 110-lll, 122, 135, 146, 164,169,196,198,199,201,213,222,240,262, 270,278,279,283-284,286,304,371,376,390, 411,431,432,551,581,582,593,619,630,631, 635,656 Autism, 261, 265-269, 660 Autobiographical knowledge, 121,303 Autobiographical memory, 37, 121, 123, 136, 146, 157, 160,162,164,167,184,216,217,401,454-455 Automatic processes, 390, 63 7 Availability, 22, 119-120, 127, 134, 135, 137, 146, 264

B Behavioral confirmation, 169-170, 171, 181 Best practices (in forensic interviews), 144, 503 Biases, 25, 70, 94, 103, 135, 196, 198, 201-202, 206-207, 218,281,345,359,420,437,510,516,518,531, 532-533,568,586-587,652 Blocking, 308

c Central details, 107, 120, 125, 127-128, 135, 375 Children/Childhood Child abuse, 30, 39-40, 41, 69, 110, 125-126, 215, 352,353,355,403-404,430,463,509-534,659 Child Advocacy Center (CAC), 553-555, 556, 557, 558,561,569,659 Child sexual abuse (CSA), 117, 146, 199, 247-248, 253,300,340,354,358,360,403,404,414,459, 509-534,546,548,550,551,553,562,564-565, 566,666-667 Child victims, 356,418, 486,497,513,520,521,527, 534,550,554 Child witnesses, 56, 73-75, 224,355-357,405, 410-411,413,414-415,417-418,441,453,460, 469,480,500,503,511,519-520,524,525,527, 530,532,545-569,658,660 Childhood sexual abuse, 248, 250, 253, 291,310,458, 523,656 Childhood events, 117-118, 120, 134, 187, 216, 217, 249,277,278,279,325-327 Children's memory, 105, 107, 118, 401-421 Children's source monitoring, 460 Children's testimony, 74, 403, 415, 417-418, 429, 433,468,510,530,546,560,566 Children's eyewitness testimony, 548 Circadian rhythm, 591 Closed circuit television (testimony), 524, 562, 564-566, 658,668 Coerced confessions, 20, 30 Cognitive Impairment, 261-262, 637-638 Cognitive Interview, 53-54, 58-73, 287, 288,353, 357, 377-379,383,387,583,592-594,597,641,663, 664-665

699

700

SUBJECT INDEX

Cognitive mechanisms, 302, 304-305, 306, 308, 409 Confabulate, 185, 245, 250, 254, 312, 331,361,385, 594 Confirmatory biases, 652 Confrontation Clause, 9 Consistency (of witness reports), 20, 21, 123, 126, 129, 130, 131-133, 135-136, 138-139, 142, 144, 145, 164,167,201,206,207,209,217,219-220,345, 354-355,361,363,402,410,482,483,522, 523,632 Consolidation, 94, 109-110, 132, 138, 140, 143, 145, 219 Constructive memory, 127-128, 134, 205-207, 218, 240-241 Controlled processing, 14, 196,390, 614 Conversation, 3-41,56,57, 61, 66, 74, 117, 122-123, 124,182,183,196,207,219-221,224,314,352, 356,401,409,418,433,460,560,563,581,653 and Litigation, 4-11 Corroboration, 66, 160, 161, 211, 280, 351, 403, 527-528,662-663 Corroborator, 158, 165, 663 Courtroom testimony, 387, 521, 553, 565, 628, 638 Co-witness information, 73, 210, 384-385, 387,391 Co-witnesses, 184, 209-211, 220, 377 Credibility, 6, 10, 35, 55, 136, 159, 166-169, 170, 204, 210,211,213,285,328,331,339-366,386,414, 460,510,511,517,519-527,530,533,566,568, 594-597,627,638,641,650,652,666-667 and Assessment, 339-365 Creeping determinism, 206 Criminal justice system, 252, 372,649-651,660,668 Criterion-Based Content Analysis (CBCA), 255, 353-362 Cross-race effect, 586 Cross-race identifications, 376-377 Cryptomnesia, 33, 35

Discovered memories, 299-300,301,312 Dissociation/Dissociated state, 104, 105, 176, 182, 213, 276-279,284-285,299,300,302,308-309,310, 313-314,496, 660 Dissociative Identity Disorder (DID), 119, 141-142 DNA/DNAevidence,160,161, 166,176-177,181,187, 239,527,548 Dual process theory, 489, 585, 614

E Easterbrook hypothesis, 87 Ecological validity, 128, 245, 411, 414, 419, 429, 510, 530 Emotional events, 28, 82--85, 92, 94-96, 98, 100, 103, 107-109, 122, 126, 134-135, 143, 146, 312, 656 Empathy,91, 213,512,513,526 Episodic memory, 159, 163, 266-267, 272, 274, 581-583,606 Event-specific knowledge, 121 Expert testimony, 3, 19, 211, 224, 328, 522, 528-529, 533-534,628 Extrinsic evidence, 176, 178 Eyewitnesses and confidence, 371-393 and evidence, 53, 98, 159-161, 345, 386, 545, 548 and memory, 53-76, 103, 111, 124, 249, 282, 321-335,339-365,372,375,380,382,392, 545-569,581-582,605-606,615,618,620-621, 627-643 and overconfidence, 392 and Statement, 167, 195-224,339,548, 582 and Testimony, 124, 159, 261-290,327-328, 333, 372,469,479,527-528,546,547,548,555,559, 587-590,622,628-629,631,633-643,657-660

D

F Daubert standard, 83,330-331,347,358 Deception, 10, 57, 160, 167-168, 171, 180, 185, 188, 219,255,339-341,344,348,349-352,358,361, 362-365,566,666-667 andCue,340,348,351-352,358,362-365 and Detection, 57, 168, 339,350,358,361-362, 363 Delay (effects on memory), 35, 40, 67, 85, 89, 103-104, 117-146,186,212,214,249,256,266,270, 271-272,287,308,332,354,374,375,385,386, 413,440,454,458,459,461-463,468,484-485, 492-493,497-498,499,522-523,528,548,558, 562,584-585,628,656 Demeanor, 4, 11, 521-522 Direct questions, 403,405-406,407,410,416,551,554, 557,566,641 Directed Forgetting, 305, 636 Directive utterances, 431 Disclosure, 18, 420, 432, 458, 522, 523, 528, 562, 655, 656,658,665-666

Facial Action Coding System (FACS), 348-350 Facilitators, 431 False confession, 175-189, 217, 344, 661 False memories/False memory susceptibility, 31, 35, 36-37, 106, 109, 134, 177-178, 182, 184, 185, 187, 196, 209, 210, 212, 214-217, 239-257, 261-290,299,301,307,308,313,315-316,322, 323-324,329,332,381,385-386,479-503,552, 583,592,605-622,641,653,654,659,663,664 False recognition/recollection, 127, 221, 240, 243, 247, 248,252,269,272-276,280,437,481-483, 485-486,488,490,492-496,499-500,605, 606-608,610-614,618,620,621 False recognition reversal, 485, 499-500 Familial informant false-narrative procedure, 215 Familiarity, 36, 103, 245,253,273,306,415,482-483, 486,488,490,585,588,607,610,614,616,617, 619-620,621,631-632,633,637,638,640,643

SUBJECT INDEX

Federal Rules of Evidence, 7, 8 Feedback effects, 69-70, 73, 179, 186, 188, 2ll, 219, 252,363-364,377,384-387,391-392,587 Flashbulb memories, 29, 37, 126, 129-133, 164, 373-374 Forensic interviews (also see investigative interviewing), 217,420,430,434-435,437-440,443,463,502, 524,547,551,553-555,557-558,562,564-565, 566,569,597 Free narrative, 54, 72, 126, 145, 403,405,416, 418,456, 461,593,651 Freezing effect, 210 Fronta1Lobes,629-630,643,659 Frye standard (and scientific research), 328, 330-331, 628 Fuzzy trace theory, 256, 277, 484-485, 487-490, 492, 495-496,499-500,502,612

G Gender Ouror gender, victim gender, defendant gender), 22,33,86, 198,199,204,275,436,511-516, 518-519,524,526,532,558,658,666 Gender differences, 5ll-516, 524 Gist Memory, 98, 133, 274-276, 485, 489, 497-499, 503 Gist-based processing, 607,612-614, 621 Glucocorticoids, 97 Ground truth, 124,126,131,145,344,345,347,414 Gudjonsson Suggestibility Scale, 186, 310, 412, 589-590

H Hearsay rule, 5-8, 658-59 Hearsay testimony, 5-10, 39-40, 520, 524-26, 562-65, 568-69,650,658-59 Hindsight biases, 25, 28-29, 206-07, 218, 653 Hippocampus, 93,97-98, 309 Hurd rules, 332 Hypermnesia, 136-39, 303, 305-06, 463 Hypnoanalysis, 321 Hypnosis and the courts, 328-29, 330-32 and memory, 299, 300, 306-07, 321-334, 378-79, 387,663,664 and misleading questions, 178, 184, 214 and confidence in memory reports, 324, 326-27 and memory illusions, 183, 214, 216, 280, 323 Hypnotic age regression, 326-27

I Identifications by eyewitnesses, 87, 98, 103, 159-60, 202, 211,221,224,374,584-86,663,665,667 Illusory truth effect, 219, Imagination inflation, 184, 215-16, 222, 240, 244, 245

701

Implantation paradigm, 215-16, 243, 245, 249-51 Implicit associative responses (IARs), 240--41 Implicit causaliry, 19 Incontrovertible evidence, 181 Incorrect inference dodge, 340 Indirect speech act, 16-17 Infantile amnesia, 325-26, 454 Inference, 13, 16, 18-21,30,38, 108, IIO, 178, 181,212, 223,256,316,340,361,479-89,497-98,628 Inhibition, 299,302-05,314,630,635-36,640 Inquisitorial justice system, 650, 658 Intellectual disability, 534 Interviewer bias, 73,405-08, 652, 659 Interrogation interviews, 39, 89, 97, 99, 168, 170, 175-189,288,351,382,650,662 and tactics, 20, 40, 176-88, 324 and police training, 55-57, 180-81, 188 Investigative interviews, 54, 64, 76, 356, 405, 412, 417, 430,433,439,480,486,566,590 Invitation utterances, 430, 432, 435, 438-40, 465

Jost's Law, 122 Judicial opinion on hypnosis, 328, 332 Jurors and gender, 5ll-14, 518, 532 and race, 516-18 Jurors' perceptions of eyewitness testimony, 328, 522, 525-26,622,653 Juror decision making, 19, 39, 136, 159, 170, 197,332, 510,519,526,533,662-666 Jury decision making, 96, 168, 170, 297-98, 206,510, 526,652-53,657 Jury selection and mock juries, 211, 524-25, 532, 530-31 Jury instructions, 510-ll, 533

K Kulechov effect, 202

L Leakage, 340 Legal accommodations, 546-4 7 Legal reforms, 54 7-69 Leniency bias, 572 Leveling, 11 Lie Detection-see credibility assessment, 340-41, 347-50,359,362-64 Life Events Inventory (LEI), 244, 251-52

M Marijuana, 281, 283-84, 286, 288-89, 660

SUBJECT INDEX

702

Memory distrust syndrome, 179 reconstruction, See reconstructive memory representation, 122-23, 132, 136, 143-44, 458, 461, 488,497,562,612 trace,94, 121-22,140,247,254,256,409,413,457, 495,607,614 triggers, 299 Mental context reinstatement, 59, 67, 72, 104, 306, 325,378-79,440-41,464-65,583,592-93,642, 652,660. Mental Retardation, 261-69, 290, 409,660 Mere-memory-testing effect, 485, 493, 502 Meta-analysis, 64-65, 68, 88, 87-99, 107, 120, 128, 270, 346,352,358-59,531 Metamemory, 306, 630, 633-35 Misinformation effects, 35-36, 68, 178, 184, 195-224 Misleading postevent information, 242, 380-81,387, 391-92,588,595,617-18,633 Misleading questions, 73, 186, 210, 214, 219, 265, 278-80,332-33,379-81,384,401,406-09,413, 441-42,457-58,460,550,615-16 Mnemonic aids, 553, 557, 561, 569 Mood dependence, mood-dependent, 299, 316 N Narrative elaboration technique (NET), 465 Narrative processing, 197 Narrowing of attention/memory, 90-92, 95, 146 Neil v. Biggers (1972), 371, 391, 667 NICHD interview protocol, 415-16,438-39,465,468-69 Normal forgetting, 299,302,305

0 Open-ended questions, 54-55, 58, 61-64, 72-74, 264-65, 356,374,405-07,410,416,418,432,435, 438-39,455,468-69,593,665 Opponent processes, 496 Option-posing utterances, 431,438-39 Overconfidence, 380, 383, 392, 634-35

p Participant vs. observer roles, 128 Per se exclusion, 328,331-34 Perception and aging, 580-81, 596-97, 654 Perception of speech, 12 Periphery of visual field, 22, 639 Peripheral details/information, 28-29, 38, 81, 87, 91-92, 95,107,110,120,128,133-35,143-46,214, 278-79,287-88,375,411-12,552,591,596,621 Permastore, 119 Person evidence and alibis, 161, 165-66 Physical abuse, 516, 524, 534 Physical context reinstatement, 217, 583

Physical evidence and alibis, 158, 160-62, 165-66, 170, 176-77 andCSf\,253,356,510,516,527,548,559 Plausibility and false memories/confessions, 187, 213, 216, 250-52 and recovered memories, 106, 109 and sexual abuse, 356, 516 Police interview training, 53, 55-57, 66-67, 70-75, 160 Polygraph Countermeasure, 343, 359 Post-stimulus elaboration, 122 Post-traumatic stress disorder (PTSD), 82, 119, 149, 276, 280-81,304,308,310,321 Pre-trial publicity, 35 Pseudomemory, 323, 329 Pychogenic amnesia, 93-94, 106, 300 Psychological coercion, 39-40, 183, 501, 661-62, 668 Psychophysiological Credibility 1\ssessment, 340, 342, 347 Public confidence ratings, 387

R Race (victim race, defendant race, juror race) and bias/stereotyping, 45, 94, 199, 204 and CSf\, 31,516-18,666 and person identification, 89-90,376, 586, 637, 640,642 Rapport-building, 32, 35, 38, 74,416 Reality Monitoring, 36, 37, 58, 61,268, 271,310,352,607 in conversation, 36, 37 Recall-to-reject, 500 Reconstruction, 95, 110, 123, 129-30, 133-34, 145-46, 220,307,312,314,349,418,497,664 Reconstructive memory, 127-28, 134, 205-06,218, 240-41 Recovered memories, 105-07, 250, 253, 279-80, 299-316,512 Recovered memoty therapy, 202, 207, 215-16, 224,242 Referential situation model, 13 Rehearsal, 85, 122, 129, 131-32, 143, 218, 220, 222, 413, 458,463 Remember/know, 245 Reminiscence, 136-39, 299, 305-06 Repetition of events/experiences, 118-19, 126, 140-42, 223, 284, 312,453,457-58,460,468,486 of interviewing, 136, 145, 244, 406, 413, 463, 550 of questioning, 74, 265, 377, 381-84,387,391, 405-06,413, 658 of recall, 134-35 Report consistency, 129, 132 Repressed memoty, 280, 515, 522, 529 Repression, 104-06, 182,299,301-05,308,313,315, 522-23,529 Resistors, 186 Retrieval induced forgetting, 219,305,636 Retrieval practice, 123, 142 Retrospective bias, 207, 218

SUBJECT INDEX

Retrospective reports, 125, 272, 457 Rule against hearsay, 5, 6, 7, 8

s Schema/Schematic Schema, 11,15,20-21,24-25,31-32,34-35,40-41, 98, 110, 127, 134, 164, 169, 171,196-208,218, 220-23,240-42,248,251,314,480-83,558,581, 588,592,653 Schema-consistent, 25, 31, 127, 199, 203,205, 242,248 Schema-inconsistent, 127, 199, 248 Schematic memory, 481 Schematic Processing, 20, 25,34-35,40, 196, 198-99,206,208,218 Schizophrenia, 261, 269-74,660 Scripts, 127, 197, 204, 251, 435-37, 458, 460, 592 Sexualabuse,8,49, 74,113,117,140,151,198,207, 216,248,250,253,279,300,307,309-12,322, 325,340,352,354-56,358,360,401,403-04, 407,414,438,440,458-60,463,465,497,509, 511-534,550-51,553,558,562 Sharpening, 11 Singleevents,117,121,126,134-35,137,140,142,144,312 Social mechanisms, 409 Source attributions, 204, 246, 258, 460, 609 Source Memory, 32-33,35, 37, 40, 46, 204, 220, 229, 231,270,466,583,598,630,632-33 Source Monitoring Framework, 183, 190, 229, 247, 254, 484,502,505,607 Source Monitoring, 32,34-5, 37-9, 103--04,180, 183, 190,205,214,216,223,247,268,270-73,277, 292-96,409-10,449-50,460,466,480,484, 488-89,494,502,559-60,583-85,607,610-11, 620,652,659-60 Specificity principle, 59, 465 Spontaneous free recall, 438 Spreading Activation, 267, 273, 608 Statement Validity Analysis (SVA), 352 Stereotype induction, 405, 407 effects on memory, 20-21, 34, 40, 74, 198, 200--01, 205,436,442,543 and/of Aging, 590-93, 627-29 Structured interview, 64-65,67, 71,378,415,438,443, 468,558-61,593-94 Substance Abuse, 105, 261, 282 Suggestibility, 178-79, 183, 186, 213, 231-34, 237, 265, 278,281,308-10,322-24,329,331,360,374. 401-27,430,432,441,456,464,492,520, 532-35,549,553,557,559-60,588-90,599,602, 616,619,635,638,641,651-53 Suggestive interviewing techniques, 339, 402,410-12, 418-19,424,443,450,464,530,542,668 Suggestive utterances, 431, 439, 663, 665

703

Suppression, 21, 29, 43, 50, 224, 296, 299, 302-04,308, 312,320, 611

T Terrorist attacks, 374 Training and credibility assessment, 243, 255, 355, 360, 362, 372,667 and interviewing skills, 53, 55-56, 66-67, 70-75, 160, 256,420,438,443,555,559,594 and source monitoring, 410,438,460,465-66, 559-60 Transportation-imagery model, 216, 218 Trauma/Traumatic, 81-83, 92-95,99-105, 110, 119, 122,135,141-42,178,215,247,250,252-54, 261,276,279-81,299-323,357,411,414,440, 442,454-55,461-63,520,524,545-46,549,555, 559,562,565-66,570-74,656,660 Traumatic abuse, 322 Trickle-down categorization, 200 Truth maxim, 17 Truth-lie discussion, 553, 555, 561, 565

u Undeutsch Hypothesis, 353-357

v Verbal Credibility Assessment, 348, 361 Verbal overshadowing, 77, 219, 221 Verballing, 674 Verbatim memory, 11-12, 16, 31, 41, 122, 241, 254,256, 481-506,563,585,612-619,653,658 Victim age, 510,519-20,531, 534 Victimization, 261,320,374,413,513-14,533,542-43, 577-79,600 Videotaping, 145,418,554,562,662

w Warnings, 4, 36, 206, 240,243, 245, 247,332-33, 533, 557,611 Weapon focus, 87-92, 95, 639 Witness feedback,384-85,387 interviewing, 54-57, 66, 69, 378 event,23, 117-121,133,137,143,145,159,184, 203,212,220,265,283,289,372-377,381,384, 411,434,579,606,615-622 y Yerkes-Dodson function, 92-93 Yes/No questions, 138, 406, 410, 438