Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies [1 ed.] 1433820897, 9781433820892

Table of contents :
Contents
Contributors
Series Foreword
Introduction
Part I Overview
Chapter 1 Components of Evidence-Based Practice in Psychology
Chapter 2 Psychotherapy Outcome Research With Ethnic Minorities: What Is the Agenda?
Part II
Measurement and Statistical Issues
Chapter 3 Threats to Cultural Validity in Clinical Diagnosis and Assessment: Illustrated With the Case of Asian Americans
Chapter 4 Statistical and Methodological Issues in Planning Randomized Clinical Trials With Ethnic Minorities
Chapter 5 Statistical Methods for Validating Test Adaptations Used in Cross-Cultural Research
Chapter 6 Testing Instrument Equivalence Across Cultural Groups: Basic Concepts, Testing Strategies, and Common Complexities
Part III
Methodological Challenges
Chapter 7 Mixed-Methods Research: Integrating Qualitative and Quantitative Approaches to the Psychological Study of Culture
Chapter 8 Cultural Adaptations in Psychotherapy for Ethnic Minorities: Strategies for Research on Culturally Informed Evidence-Based Psychological Practices
Chapter 9 Community-Based Participatory Research for Cocreating Interventions With Native Communities: A Partnership Between the University of New Mexico and the Pueblo of Jemez
Part IV Treatment and
Interventions
Chapter 10 A Culturally Informed Approach to American Indian/Alaska Native Youth Suicide Prevention
Chapter 11 Depression Prevention and Treatment Interventions: Evolution of the San Francisco Latino Mental Health Research Program
Chapter 12 Improving the Participation of Families of Color in Evidence-Based Interventions: Challenges and Lessons Learned
Chapter 13 Culturally Adapting Evidence-Based Practices for Ethnic Minority and Immigrant Families
Afterword: Some Culturally Informed Conclusions
Index
About the Editors

Citation preview

E VIDE NC E- BAS ED

PSYCHOLOGICAL P R A C T I C E W ITH ETHNIC MINORITIES

Cultural, Racial, and Ethnic Psychology Book Series Series Editor: Frederick T. L. Leong Qualitative Strategies for Ethnocultural Research Edited by Donna K. Nagata, Laura Kohn-Wood, and Lisa A. Suzuki Positive Psychology in Racial and Ethnic Groups: Theory, Research, and Practice Edited by Edward C. Chang, Christina A. Downey, Jameson K. Hirsch, and Natalie J. Lin The Cost of Racism for People of Color: Contextualizing Experiences of Discrimination Edited by Alvin N. Alvarez, Christopher T. H. Liang, and Helen A. Neville Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies Edited by Nolan Zane, Guillermo Bernal, and Frederick T. L. Leong

EV I DENCE -BA SE D

PSYCHOLOGICAL P R A C T I C E W IT H ETHNIC MINORITIES CULTURALLY INFORMED RESEARCH and CLINICAL STRATEGIES

Edited by NOLAN ZANE, GUILLERMO BERNAL, and FREDERICK T. L. LEONG

American Psychological Association • Washington, DC

Copyright © 2016 by the American Psychological Association. All rights reserved. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be reproduced or distributed in any form or by any means, including, but not limited to, the process of scanning and digitization, or stored in a database or retrieval system, without the prior written permission of the publisher. Published by American Psychological Association 750 First Street, NE Washington, DC 20002 www.apa.org

To order APA Order Department P.O. Box 92984 Washington, DC 20090-2984 Tel: (800) 374-2721; Direct: (202) 336-5510 Fax: (202) 336-5502; TDD/TTY: (202) 336-6123 Online: www.apa.org/pubs/books E-mail: [email protected]

In the U.K., Europe, Africa, and the Middle East, copies may be ordered from American Psychological Association 3 Henrietta Street Covent Garden, London WC2E 8LU England Typeset in Goudy by Circle Graphics, Inc., Columbia, MD Printer: United Book Press, Baltimore, MD Cover Designer: Mercury Publishing Services, Inc., Rockville, MD The opinions and statements published are the responsibility of the authors, and such opinions and statements do not necessarily represent the policies of the American Psychological Association. Library of Congress Cataloging-in-Publication Data Names: Zane, Nolan W. S., editor. | Bernal, Guillermo, editor. | Leong, Frederick T. L., editor. Title: Evidence-based psychological practice with ethnic minorities : culturally informed research and clinical strategies / Nolan Zane, Guillermo Bernal, and Frederick T. L. Leong. Description: Washington, DC : American Psychological Association, [2016] | Series: Cultural, racial, and ethnic psychology | Includes bibliographical references and index. Identifiers: LCCN 2015043406 | ISBN 9781433820892 | ISBN 1433820897 Subjects: LCSH: Cultural psychiatry. | Ethnopsychology. | Minorities—Mental health—United States. | Minorities—United States—Psychology. Classification: LCC RC455.4.E8 .E95 2016 | DDC 616.890089—dc23 LC record available at http://lccn.loc.gov/2015043406 British Library Cataloguing-in-Publication Data A CIP record is available from the British Library. Printed in the United States of America First Edition http://dx.doi.org/10.1037/14940-000

CONTENTS

Contributors.................................................................................................   ix  Series Foreword..........................................................................................   xiii Frederick T. L. Leong Introduction.................................................................................................. 3 Nolan Zane, Guillermo Bernal, and Frederick T. L. Leong I. Overview...............................................................................................  13 Chapter 1. Components of Evidence-Based Practice in Psychology...................................................................  15 Ronald F. Levant and Heather A. Sperry Chapter 2. Psychotherapy Outcome Research With Ethnic Minorities: What Is the Agenda?....................................  31 Anna S. Lau, Doris F. Chang, Sumie Okazaki, and Guillermo Bernal

v

II.  Measurement and Statistical Issues...................................................  55 Chapter 3. Threats to Cultural Validity in Clinical Diagnosis and Assessment: Illustrated With the Case of Asian Americans.........................................................  57 Frederick T. L. Leong and Zornitsa Kalibatseva Chapter 4. Statistical and Methodological Issues in Planning Randomized Clinical Trials With Ethnic Minorities...................................................  75 Carmen L. Rivera-Medina and José N. Caraballo Chapter 5. Statistical Methods for Validating Test Adaptations Used in Cross-Cultural Research..................................  103 Joseph A. Rios and Ronald K. Hambleton Chapter 6. Testing Instrument Equivalence Across Cultural Groups: Basic Concepts, Testing Strategies, and Common Complexities..........................................  125 Barbara M. Byrne III.  Methodological Challenges.............................................................  145 Chapter 7. Mixed-Methods Research: Integrating Qualitative and Quantitative Approaches to the Psychological Study of Culture............................................................  147 Marina Doucerain, Sylvanna Vargas, and Andrew G. Ryder Chapter 8. Cultural Adaptations in Psychotherapy for Ethnic Minorities: Strategies for Research on Culturally Informed Evidence-Based Psychological Practices.......  169 Nolan Zane, Jin E. Kim, Guillermo Bernal, and Catrina Gotuaco Chapter 9. Community-Based Participatory Research for Cocreating Interventions With Native Communities: A Partnership Between the University of New Mexico and the Pueblo of Jemez.... 199 Lorenda Belone, Janice Tosa, Kevin Shendo, Anita Toya, Kee Straits, Greg Tafoya, Rebecca Rae, Emma Noyes, Doreen Bird, and Nina Wallerstein

vi       contents

IV.  Treatment and Interventions.........................................................  221 Chapter 10. A Culturally Informed Approach to American Indian/Alaska Native Youth Suicide Prevention......... 223 Teresa D. LaFromboise and Saima S. Malik Chapter 11. Depression Prevention and Treatment Interventions: Evolution of the San Francisco Latino Mental Health Research Program............................................. 247 Adrián Aguilera, Jeanne Miranda, Sergio Aguilar-Gaxiola, Kurt C. Organista, Gerardo M. González, John McQuaid, Laura P. Kohn-Wood, Huynh-Nhu Le, Chandra Ghosh-Ippen, Guido G. Urizar, José Soto, Tamar Mendelson, Alinne Z. Barrera, Leandro D. Torres, Yan Leykin, Stephen Schueller, Nancy Liu, and Ricardo F. Muñoz Chapter 12. Improving the Participation of Families of Color in Evidence-Based Interventions: Challenges and Lessons Learned...................................................... 273 Norweeta G. Milburn and Marguerita Lightfoot Chapter 13. Culturally Adapting Evidence-Based Practices for Ethnic Minority and Immigrant Families................ 289 Wei-Chin Hwang Afterword: Some Culturally Informed Conclusions................................... 309 Nolan Zane, Guillermo Bernal, and Frederick T. L. Leong Index.........................................................................................................  315 About the Editors.....................................................................................  329

contents     

vii

CONTRIBUTORS

Sergio Aguilar-Gaxiola, MD, PhD, University of California, Davis Adrián Aguilera, PhD, University of California, Berkeley Alinne Z. Barrera, PhD, Palo Alto University, Palo Alto, CA Lorenda Belone, PhD, MPH, University of New Mexico College of Education and University of New Mexico Center for Participatory Research, Albuquerque Guillermo Bernal, PhD, University of Puerto Rico, Río Piedras Doreen Bird, MPH, University of New Mexico Department of Psychiatry and Behavioral Sciences, Albuquerque Barbara M. Byrne, PhD, University of Ottawa, Ottawa, Ontario, Canada José N. Caraballo, PhD, Pontifical Catholic University of Puerto Rico, Ponce Doris F. Chang, PhD, The New School for Social Research, New York, NY Marina Doucerain, PhD, University of Quebec at Montreal, Canada Chandra Ghosh-Ippen, PhD, University of California, San Francisco Gerardo M. González, PhD, California State University, San Marcos Catrina Gotuaco, BA, California State University, East Bay, Hayward Ronald K. Hambleton, PhD, University of Massachusetts, Amherst Wei-Chin Hwang, PhD, Claremont McKenna College, Claremont, CA ix

Zornitsa Kalibatseva, PhD, Stockton University, Galloway, NJ Jin E. Kim, PhD, University of California, San Francisco Laura P. Kohn-Wood, PhD, University of Miami, Coral Gables, FL Teresa D. LaFromboise, PhD, Stanford University, Stanford, CA Anna S. Lau, PhD, University of California, Los Angeles Huynh-Nhu Le, PhD, George Washington University, Washington, DC Frederick T. L. Leong, PhD, Michigan State University, East Lansing Ronald F. Levant, EdD, The University of Akron, Akron, OH Yan Leykin, PhD, University of California, San Francisco Marguerita Lightfoot, PhD, University of California, San Francisco Nancy Liu, PhD, University of California, San Francisco Saima S. Malik, EdM, Stanford University, Stanford, CA John McQuaid, PhD, San Francisco Veterans Administration, San Francisco, CA Tamar Mendelson, PhD, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD Norweeta G. Milburn, PhD, University of California, Los Angeles Jeanne Miranda, PhD, University of California, Los Angeles Ricardo F. Muñoz, PhD, Palo Alto University, Palo Alto, CA Emma Noyes, MPH, Washington State University College of Nursing, Spokane Sumie Okazaki, PhD, New York University, New York Kurt C. Organista, PhD, University of California, Berkeley Rebecca Rae, MCRP, MWR, University of New Mexico Center for Participatory Research, Albuquerque Joseph A. Rios, MA, Educational Testing Service, Princeton, NJ Carmen L. Rivera-Medina, PhD, University of Puerto Rico, Río Piedras Andrew G. Ryder, PhD, Concordia University, Montreal, Quebec, Canada; Jewish General Hospital, Montreal, Quebec, Canada Stephen Schueller, PhD, Northwestern University Feinberg School of Medicine, Chicago, IL Kevin Shendo, BA, Pueblo of Jemez Department of Education, Jemez Pueblo, NM José Soto, PhD, Pennsylvania State University, State College Heather A. Sperry, MA, The University of Akron, Akron, OH Kee Straits, PhD, University of New Mexico Robert Wood Johnson Foundation Center for Health Policy, Albuquerque Greg Tafoya, MPH, University of New Mexico Center for Participatory Research, Albuquerque Leandro D. Torres, PhD, University of California, San Francisco Janice Tosa, MA, Pueblo of Jemez Department of Education, Jemez Pueblo, NM x       contributors

Anita Toya, Pueblo of Jemez Health and Human Services Department, Jemez Pueblo, NM Guido G. Urizar, PhD, California State University, Long Beach Sylvanna Vargas, BA, University of Southern California, Los Angeles Nina Wallerstein, DrPH, University of New Mexico Public Health Program; University of New Mexico Center for Participatory Research, Albuquerque Nolan Zane, PhD, University of California, Davis

contributors     

xi

SERIES FOREWORD

As series editor of the American Psychological Association’s (APA’s) Division 45 (Society for the Psychological Study of Ethnic Minority Issues) book series on Cultural, Racial, and Ethnic Psychology, it is my pleasure to introduce another volume in the series: Zane, Bernal, and Leong’s edited volume, Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies. This volume represents a major attempt to meet a challenge that has been facing Division 45 for the past two decades, namely, reconciling the cross-cultural competence movement with the evidence-based practice movement within psychology. In 1982, the Division of Counseling Psychology (APA Division 17: now the Society of Counseling Psychology) published a position paper that launched the cross-cultural competency movement in counseling and psychotherapy (Sue et al., 1982). Despite the significant impact of this movement, a lingering issue has remained: the “criterion problem.” This issue concerns the lack of research evidence that a culturally competent therapist produces better client outcomes than a therapist who is not deemed culturally competent. Indeed, much of the cross-cultural competence movement has focused on the therapist and her or his training rather than on therapeutic outcomes. At around the same time, APA Division 12 (now xiii

the Society of Clinical Psychology; formerly, Clinical Psychology) had adopted Cochrane’s (1979) call for evidence-based practice in developing guidelines for empirically validated treatments (EVTs). EVTs were eventually viewed as overly restrictive, but it nevertheless seemed reasonable that the evidencebased practice (EBP) approach should be adopted by clinical psychology. As Cochrane suggested, we need to be guided by a critical summary of the best available scientific evidence for how we approach our practice. The APA 2005 Presidential Task Force on Evidence-Based Practice (2006) addressed the issue of treatment for racial and ethnic minority groups. The Task Force report noted that a client’s characteristics, such as age, culture, race, ethnicity, gender, gender identity, religious beliefs, family context, and sexual orientation, need special attention, and that all of these attributes influence the client’s “personality, values, worldviews, relationships, psychopathology, and attitudes toward treatment” (p. 279). More important, the report concluded that “evidence-based practice in psychology (EBPP) is the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and preferences” (APA Presidential Task Force on Evidence-Based Practice, 2006, p. 273). However, the question remains, how do we reconcile these two movements? What do you do when there are no or limited studies to provide the evidence for psychotherapy with racial and ethnic minority clients and patients? Undoubtedly, withholding treatments for racial and ethnic minority patients until evidence can be accumulated cannot be the solution. Zane, Bernal, and Leong’s volume addresses many of these issues and challenges of reconciling the two movements in their Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies. The impetus for this book series came from my presidential theme for Division 45, which focused on “Strengthening Our Science to Improve Our Practice.” Given the increasing attention to racial and ethnic minority issues within the discipline of psychology, I argued that we needed to both generate more research and get the existing research known. From the supplement to the Surgeon General’s report on mental health (U.S. Department of Health and Human Services, 2001) to the Unequal Treatment report from the Institute of Medicine (Smedley, Stith, & Nelson, 2003)—both of which documented extensive racial and ethnic disparities in our health care system—the complex of culture, race, and ethnicity was becoming a major challenge in both research and practice within the field of psychology. To meet that challenge, Division 45 acquired its own journal devoted to ethnic minority issues in psychology (Cultural Diversity and Ethnic Minority Psychology). At the same time, a series of handbooks on the topic were published, including Bernal, Trimble, Burlew, and Leong’s (2003) Handbook of Racial and Ethnic Minority Psychology. Yet, we felt that more xiv       series foreword

coverage of this subdiscipline was imperative—coverage that would match the substantive direction of the handbooks, but would come from a variety of research and practice perspectives. Hence, the Division 45 book series was launched. The series on Cultural, Racial, and Ethnic Psychology was designed to advance our theories, research, and practice regarding this increasingly crucial subdiscipline. It will focus on, but not be limited to, the major racial and ethnic groups in the United States (i.e., African Americans, Hispanic Americans, Asian Americans, and American Indians) and will include books that examine a single racial or ethnic group, as well as books that undertake a comparative approach. The series will also address the full spectrum of related methodological, substantive, and theoretical issues, including topics in behavioral neuroscience, cognitive and developmental psychology, and personality and social psychology. Other volumes in the series will be devoted to cross-disciplinary explorations in the applied realms of clinical psychology and counseling, as well as educational, community, and industrial– organizational psychology. Our goal is to commission state-of-the-art volumes in cultural, racial, and ethnic psychology that will be of interest to both practitioners and researchers. Let me end by thanking the members of the editorial board who do the work of recruiting and reviewing proposals for the series: Guillermo Bernal, University of Puerto Rico, Rí­o Piedras Campus; Beth Boyd, University of South Dakota; Lillian Comas-Díaz, private practice, Washington, DC; Sandra Graham, University of California, Los Angeles; Gordon Nagayama Hall, University of Oregon; Helen A. Neville, University of Illinois at Urbana–Champaign; Teresa D. LaFromboise, Stanford University; Richard Lee, University of Minnesota; Robert M. Sellers, University of Michigan; Stanley Sue, Palo Alto University; Joseph E. Trimble, Western Washington University; and Michael Zarate, University of Texas at El Paso. They represent leading scholars in psychology who have graciously donated their time to help advance the field. —Frederick T. L. Leong Series Editor

REFERENCES American Psychological Association Presidential Task Force on Evidence-Based Practice. (2006). Evidence-based practice in psychology. American Psychologist, 61, 271–285. http://dx.doi.org/10.1037/0003-066X.61.4.271 series foreword     

xv

Bernal, G., Trimble, J. E., Burlew, A. K., & Leong, F. T. L. (2003). Handbook of racial and ethnic minority psychology. Thousand Oaks, CA: Sage. Cochrane, A. L. (1979). 1931–1971: A critical review with particular reference to the medical profession. In G. Teeling-Smith & N. Wells (Eds.), Medicines for the year 2000 (pp. 2–12). London, England: Office of Health Economics. Smedley, B. D., Stith, A. Y., & Nelson, A. R. (Eds.). (2003). Unequal treatment: Confronting racial and ethnic disparities in health care. Washington, DC: National Academies Press. Sue, D. W., Bernier, J. E., Durran, A., Feinberg, L., Pedersen, P., Smith, E. J., & Vasquez-Nuttall, E. (1982). Position paper: Cross-cultural counseling competencies. The Counseling Psychologist, 10, 45–52. U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity, a supplement to Mental Health: A Report of the Surgeon General. Washington, DC: Author.

xvi       series foreword

E VIDE NC E- BAS ED

PSYCHOLOGICAL P R A C T I C E W ITH ETHNIC MINORITIES

INTRODUCTION NOLAN ZANE, GUILLERMO BERNAL, AND FREDERICK T. L. LEONG

Ethnic and racial diversity in the United States is becoming an important and dynamic societal reality. By 2043, census officials indicate that ethnic/ racial minorities will outnumber Whites. Minority populations now account for 37.5% of the population and are continuing to grow. In contrast, those who identified as White alone experienced the slowest rate of population growth (U.S. Census Bureau, 2011). The ethnic/racial minority population will account for nearly 90% of the total growth in the U.S. population from 1995 to 2050. More than half of the growth in the U.S. population between 2000 and 2010 was due to the increase in the Latino population. During this period, the Asian American population grew 4 times faster than the general population and faster than any other minority group (U.S. Census Bureau, 2011). The two groups will more than triple their 1995 population sizes, reaching 97 million and 34 million people, respectively, in 2050 (Minority Development Business

http://dx.doi.org/10.1037/14940-001 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

3

Agency, 1999). Between 2000 and 2010, Texas joined California, the District of Columbia, Hawaii, and New Mexico in having a “majority minority” population, where more than 50% of the population was part of an ethnic or racial minority group. By 2025, 13 more states will be one third or more racial/ ethnic minority, and these include the populous states of Florida, New York, Illinois, Georgia, New Jersey, and Virginia (U.S. Census Bureau, 2011). This growing diversity has significant social, political, economic, and human services ramifications. By 2020, Hispanics (18.6%), Blacks (12.0%), Asians (5.7%), and all those belonging to the “all other groups” category (2.9%) will make up nearly 40% of the civilian labor force (U.S. Bureau of Labor Statistics, 2012). The ethnic minority marketplace in the United States now exceeds the gross national product of Canada. Moreover, in 2012, the $1.2 trillion Latino market was larger than the entire economies of all but 13 countries in the world (Selig Center for Economic Growth, 2012). Demographers predict that by the time the so-called baby boomers retire, the majority of those contributing to the Social Security and pension plans of primarily White workers will be racial/ethnic minorities. In California, White students in public schools now constitute the minority, with many other states of the country expected to follow suit by 2043 (Institute of Education Sciences, 2014). Clearly, the mental health field, as with other human services, must address and be responsive to this diversity. With respect to best practices in mental health care, the American Psychological Association (APA) has adopted a policy that defines evidencebased practice in psychology (EBPP), affirms the importance and usefulness of using EBPP to enhance health, and delineates the various principles that guide EBPP. EBPP is defined as “the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and preferences” (APA Presidential Task Force on Evidence-Based Practice, 2006, p. 273). Essentially, effective psychological treatment involves three critical processes: (a) applying the best available research evidence in the selection and application of treatments, (b) using clinical expertise that encompasses a number of competencies that have been found to promote positive therapeutic outcomes, and (c) being responsive to the patient’s characteristics, culture, and personal preferences (APA Presidential Task Force on EvidenceBased Practice, 2006). The first process is the focus of most clinical research. It involves using treatments that are effective according to randomized controlled trials (RCTs). However, the generalizability of such research to ethnic minority clients must be considered. The second and third processes involve—among other things—cultural competence. Cultural competence is the ability of the clinician to work with a client and provide treatment in a manner that is culturally meaningful and ecologically valid. Clinicians must have relevant skill sets as well as substantial cultural knowledge to deliver 4       zane, bernal, and leong

effective interventions to members of a particular culture. Given the aforementioned increase in diversity nationwide, mental health providers need cultural competence to work effectively with ethnic minority clients. This volume considers how all three EBPP processes relate to ethnic minority clients, with a particular emphasis on cultural competence. THE NEED FOR CULTURAL COMPETENCE From a multicultural perspective, it is clear that psychological services are most effective when responsive to the patient’s specific problems, strengths, personality, sociocultural context, and preferences. Thus, attending to ethnocultural aspects of the patient is a critical and essential component of EBPP. However, since 1978, various presidential commissions have documented ethnic disparities in mental health in terms of the unmet mental health needs of members of ethnic minority groups such as African Americans, American Indians, Asian Americans, and Latino/as (President’s Commission on Mental Health, 1978; President’s New Freedom Commission on Mental Health, 2003). These commissions concluded that the disparities were not so much due to racial and ethnic differences in rates of psychopathology but rather were due to inaccessible and ineffective treatment. Ethnic minority clients often saw therapists or were administered treatments that did not provide consideration of the clients’ lifestyles, cultural and linguistic backgrounds, and life circumstances. In view of the policy adopted by the APA on EBPP, those very considerations involving the patient’s culture and race are essential to effective psychological practice. Clearly, this is a major priority for EBPP and the mental health profession. The EBPP movement appears to provide some impetus to reduce ethnic and racial disparities in mental health. EBPP uses the best available evidence on patient characteristics, culture, and personal preferences to adapt the treatment to best serve a particular client. Nevertheless, the substantial promise of EBPP for addressing cultural diversity issues is largely tempered by the reality that there is little of this “best available evidence” with respect to ethnic minority populations. THE NEED FOR RELEVANT RESEARCH ADDRESSING ETHNIC MINORITY ISSUES The major problem in trying to use the EBPP model to guide treatment interventions with ethnic minority clients is that relatively little research has been conducted on these clients, especially research that satisfies rigorous introduction     

5

research criteria such as those involved in RCTs or empirically supported treatments (ESTs). In the case of ESTs (also referred to as empirically validated treatments), Chambless et al. (1996) could not find a single rigorous study that examined the efficacy of treatment for any ethnic minority population. The U.S. Department of Health and Human Services (2001) reported that the gap between research and practice is particularly acute for racial and ethnic minorities. Research involving controlled clinical trials that were used to generate professional treatment guidelines did not conduct specific analyses for any minority group. From 1986 to 2001, about 10,000 participants were included in RCTs evaluating the efficacy of treatments for certain disorders. For nearly half of these participants (n = 4,991), no information on race or ethnicity was given. For another 7% of participants (n = 656), studies reported only the general designation “non-White.” For the remaining 47% of participants (n = 4,335), very few minorities were included; not a single study analyzed the efficacy of the treatment by ethnicity or race. These earlier reviews did not include outcome studies conducted since the National Institutes of Health mandated that grant applicants include adequate samples of minorities, women, and children, nor did they explain why such samples could not be obtained. Some believe this mandate may have significantly increased treatment research on underserved populations, especially ethnic minorities. Findings from a study seem to indicate that even the most current treatment research programs are not producing “best available evidence” on minority populations or issues. The study involved a review of 379 National Institute of Mental Health-funded clinical trials published between 1995 and 2004 in the five leading mental health journals (Mak, Law, Alvidrez, & Perez-Stable, 2007). The investigators found that fewer than half of the studies provided information on the specific ethnic composition of their samples. Among those that specified their ethnic composition, most ethnic minority groups were underrepresented, notably Asian Americans, Hispanics, and Native Americans. White Americans continued to dominate as participants in clinical trials (61% in studies that provided specific ethnic information). Moreover, few studies analyzed for ethnic or cultural effects. This lack of research on culturally diverse populations is not confined to empirical work in the mental health field. Recently, Chen, Lara, Dang, Paterniti, and Kelly (2014) conducted a review of samples used in clinical trial studies funded by the National Cancer Institute. Of the research institutes at the National Institutes of Health, the National Cancer Institute has one of the largest annual budgets. Chen and colleagues (2014) found that of the 75,215 trial participants in studies of breast, colorectal, lung, and prostate cancer from 2000 to 2002, the proportion of African American trial participants declined from 3.7% to 3% of this total and the proportion of Hispanic trial participants decreased from 11% to 7.9%. Moreover, only 20% 6       zane, bernal, and leong

of studies published in high-impact oncology journals examined race and/or ethnic variations in their results. Cultural competence is an important and necessary condition of EBPP and, as such, EBPP can be a great catalyst for addressing ethnic and racial disparities in mental health treatment and services. However, researchers and funding agencies have not paid much attention to ethnic and cultural research that determines if these treatments are effective or, in other words, culturally and ecologically valid. The conclusions reached by the President’s Commission on Mental Health in the late 1970s are echoed today, some 35 years later, in the U.S. Department of Health and Human Services Surgeon General’s Supplement (U.S. Department of Health and Human Services, 2001) and the President’s New Freedom Commission on Mental Health (2003). Research is needed that is inclusive of ethnic minority populations but also explanatory in nature about the effects of cultural variables. PURPOSE AND OVERVIEW OF THE VOLUME Developing and generating this type of meaningful research involves many challenges. The overall purpose of this volume is to inform and stimulate research and evaluation efforts that lead to the development of evidencebased practices for ethnic minority populations. Experts in the field of ethnic minority mental health treatment examine the reasons why research on culturally informed psychological practices has not progressed as much as some had hoped as well as present tangible ways and substantive strategies for conducting more meaningful and effectual research in this area. The volume is divided into four parts. Part I (Chapters 1 and 2) discusses general challenges confronting research on cultural factors in EBPP. Part II (Chapters 3–6) discusses critical measurement issues, such as the use of culturally valid and acceptable measures, measurement equivalence, assessment of change in treatment, test translation and adaptation, and major statistical challenges (e.g., power and sample heterogeneity, moderated mediation, and testing for equivalence vs. differences). Part III (Chapters 7–9) covers important design and methodological issues, such as the use of meta-analytic strategies, mixed-method approaches, and participatory research in underserved communities. Finally, Part IV (Chapters 10–13) covers approaches to culturally adapting treatments. In short, the volume consists of chapters that discuss the challenges and highlight specific recommendations for conducting culturally informed EBPP. Below is a brief overview of individual chapters. Levant and Sperry (Chapter 1) provide an overview of EBPP and a summary of the report of the APA Presidential Task Force on Evidence-Based Practice (2006). The three integral components of EBPP are best available introduction     

7

research evidence, clinical expertise, and patient characteristics, including cultural values and treatment preferences. The authors emphasize how EBPP involves an integration of these multiple aspects of best practices in psychological treatment. Lau, Chang, Okazaki, and Bernal (Chapter 2) provide a critical and selective review of the treatment outcome research with ethnocultural groups. Because ethnic minorities are underrepresented in clinical trials, questions remain about the applicability of EBTs for ethnic minority groups and the best strategies for improving their outcomes in care. The authors present three distinct lines of investigation, focusing on generalizability of EBTs for ethnic minorities, cultural adaptation of EBTs, and innovation with culturally sensitive therapies. To parse the priorities for continued research, the authors review emerging research within each area with attention to both methodological and epistemological issues. Leong and Kalibatseva (Chapter 3) propose a conceptual model to evaluate threats to cultural validity in clinical diagnosis and assessment among racial and ethnic minorities and exemplify it with Asian Americans. Minimizing threats to cultural validity in clinical diagnosis and assessment is essential for culturally informed EBPP. Cultural validity refers to the efficacy of an instrument or the accuracy of a diagnosis to incorporate important cultural factors. The authors examine five threats to cultural validity: pathoplasticity of psychological disorders, cultural factors influencing symptom expression, therapist bias in clinical judgment, language capability of the client, and inappropriate use of diagnostic and personality tests. Rivera-Medina and Caraballo (Chapter 4) discuss statistical and methodological issues that investigators need to consider in planning an RCT. They focus on the type of variables to be used as outcomes (dimensional vs. dichotomous), sample size and its impact on the power to detect the effectiveness of the intervention, handling of missing data, and consequences of adding secondary outcomes or conducting multiple testing. Key methodological concerns and analytic strategies that impact the results from an RCT with ethnic minority groups are examined. For example, they discuss the limitations of the intent-to-treat analytic strategy in clinical trials within ethnic minority communities. Rios and Hambleton (Chapter 5) discuss the statistical methods for validating test adaptations, which are used in cross-cultural research. Because unresolved methodological issues can lead to invalid references from data in cross-cultural research, the authors of this chapter seek to provide researchers with a basic knowledge of the statistical procedures that can be implemented to evaluate measurement equivalence within a cross-cultural context. The authors examine three potential sources of measurement bias in cross-cultural and cross-ethnic assessment: construct bias, method bias, and item bias. They 8       zane, bernal, and leong

also discuss the specific conditions or circumstances that result in these types of bias (e.g., method bias can occur from sample, administration, or test bias). Byrne (Chapter 6) focuses on the testing of instrument equivalence across cultures. She provides an overview of the basic concepts associated with both the measurement and structural equivalence of a measuring instrument. Multigroup comparisons of mean scores assume that both the instrument and the construct being measured operate the same across the populations of interest, and this assumption must be tested. She also discusses methods for testing instrument equivalence. Finally, she discusses the complexities in testing for instrument equivalence across cultural groups. For example, she notes that the issue of bias does not necessarily relate to the intrinsic properties of an instrument but rather to the characteristics of the respondents from each cultural group. Doucerain, Vargas, and Ryder (Chapter 7) discuss mixed-methods research, which involves integrating qualitative and quantitative approaches to the psychological study of culture. They first define mixed-methods research and then explain its philosophical underpinnings. They then discuss the reasons for using mixed methods in research, focusing on five main rationales: triangulation, complementarity, development, initiation, and expansion. They also present a typology of mixed-methods research designs involving three dimensions: level of mixing, time orientation, and emphasis of approaches. The design chosen should be governed primarily by the research question at hand and by which integration would ensure optimizing the benefits of both qualitative and quantitative methods. Zane, Kim, Bernal, and Gotuaco (Chapter 8) critically review the work done to culturally adapt psychotherapies for culturally diverse populations. They examine efforts to culturally adapt ESTs for use with culturally diverse populations. They focus only on adaptations made to ESTs or empirically validated treatments because these interventions are often considered as the treatments of choice for clients, including ethnic minority clients. They also conduct a content analysis to determine what domains have been addressed by these adaptations and what other domains may need to be addressed to advance culturally competent treatment. Finally, they propose an alternative approach that may be especially helpful in advancing this work. Belone and colleagues (Chapter 9) focus on the implementation of community-based participatory research (CBPR) with American Indian/ Alaska Native (AI/AN) communities. Because of historic genocide and federal institutional policies perpetuated against AI/AN communities, CBPR is a particularly appropriate approach. Based on a 13-year tribal–academic partnership, indigenous and Western theory were blended to cocreate and implement a culturally centered prevention curriculum, the Family Listening Program. This program integrated three perspectives: cultural centeredness introduction     

9

and indigenous theory, a public health model of risk and protective factors, and empowerment theory based on the educational philosophy of Paulo Freire. CBPR partnering processes provide needed data on culturally centered interventions in improving access and quality care for ethnocultural groups. LaFromboise and Malik (Chapter 10) provide a compelling example of the highly delicate balance between maintaining scientific rigor and preserving culture and tradition within an intervention. They have developed the American Indian Life Skills program, which is a suicide prevention intervention for AI/AN youth. AI/AN youth have the highest rates of suicide among all ethnic minority adolescents; culturally unique risk factors for suicide include historical trauma, stress from acculturation, community violence, and substance abuse. These factors are addressed in the development of the prevention intervention. Furthermore, the authors emphasize that it is necessary to use a community-driven approach that actively engages members in intervention design, evaluation, and implementation. Aguilera and colleagues (Chapter 11) demonstrate how a training program located in a public-sector hospital can contribute to the dissemination of culturally appropriate evidence-based practices. The Clinical Psychology Training Program of the University of California, San Francisco, sponsors 12 programs that involve innovative prevention or treatment interventions to manage specific mental health conditions such as depression and alcohol abuse. Within all of these projects, data are gathered to allow for treatments or interventions to be enhanced. For example, in light of the information that there is lower rate of adherence in disadvantaged populations, the text messaging study used text messages to complement cognitive–behavioral therapy. In essence, it is important to note that these types of training programs are appealing to many researchers who are committed to conducting research with these populations in their workplaces. Milburn and Lightfoot (Chapter 12) note how family-based interventions can reduce high-risk behaviors in vulnerable adolescent populations such as homeless youth and substance-using youth. However, a major challenge in implementing family-based interventions centers on overcoming obstacles to family involvement in these interventions. Using a case example of an evidence-based family-based intervention for delinquent African American adolescents and their parents or guardians, the authors provide specific strategies for increasing family participation. They also discuss using formative research to develop a more tailored strategy for enhancing family involvement. They distinguish between recruitment, retention, and engagement issues when designing and evaluating the effectiveness of family-based interventions for culturally diverse communities. In the final chapter, Hwang (Chapter 13) discusses a strategy to culturally adapt an empirically supported treatment using two approaches. The 10       zane, bernal, and leong

psychotherapy adaptation and modification framework (PAMF) is a theorydriven and top-down approach, whereas the formative method for adapting psychotherapy (FMAP) represents a community-based and bottom-up approach. He outlines the five major phases of the community-based FMAP approach, of which the theory-driven PAMF constitutes one particular phase. These approaches were used to guide the cultural adaptations to cognitive–behavioral therapy for Chinese Americans with depression. This project demonstrates how salient cultural issues can be systematically addressed to tailor the intervention to the needs of ethnic minority and immigrant communities. It is our hope that this volume will provide a rough but instructive road map for conducting more systematic and high-yield research on culturally informed interventions. Like most maps, this one will eventually become outdated and obsolete as more “roads” and “pathways” (in this case, to culturally competent mental health care) are built. We eagerly look forward to when this occurs. REFERENCES American Psychological Association Presidential Task Force on Evidence-Based Practice. (2006). Evidence-based practice in psychology. American Psychologist, 61, 271–285. http://dx.doi.org/10.1037/0003-066X.61.4.271 Chambless, D. L., Sanderson, W. C., Shoham, V., Bennett-Johnson, S., Pope, K. S., & Crits-Christoph, P. (1996). An update on empirically validated therapies. The Clinical Psychologist, 49, 5–18. Chen, M. S., Lara, P. N., Dang, J. H. T., Paterniti, D. A., & Kelly, K. (2014). Renewing the case for enhancing minority participation in cancer clinical trials. Cancer, 120, 1091–1096. http://dx.doi.org/10.1002/cncr.28575 Institute of Education Sciences. (2014). Racial/ethnic enrollment in public schools. Retrieved from http://www.nces.ed.gov/programs/coe/ Mak, W. W. S., Law, R., Alvidrez, J., & Perez-Stable, E. J. (2007). Diversity representation in NIMH-funded clinical trials: Review of a decade of published research. Administration and Policy in Mental Health, 34, 497–503. http://dx.doi. org/10.1007/s10488-007-0133-z Minority Development Business Agency. (1999). The emerging minority marketplace. Retrieved from http://www.nces.ed.gov/programs/coe President’s Commission on Mental Health. (1978). Report to the President. Washington, DC: U.S. Government Printing Office. President’s New Freedom Commission on Mental Health. (2003). Achieving the promise: Transforming mental health care in America. Report of the President’s New Freedom Commission on Mental Health. Rockville, MD: Author. introduction     

11

Selig Center for Economic Growth. (2012). Hispanic consumer market in the U.S. is larger than the entire economies of all but 13 countries in the world, according to annual UGA Selig Center Multicultural Economy study. Retrieved from http:// www.terry.uga.edu/news/ U.S. Bureau of Labor Statistics. (2012). Labor force projections to 2020: A more slowly growing workforce. Monthly Labor Review. Retrieved from http://www. bls.gov/opub/mlr/2015/home.htm U.S. Census Bureau. (2011). Overview of Race and Hispanic origin: 2010. Retrieved from http://www.census.gov/library/publications.html U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity, a supplement to Mental Health: A Report of the Surgeon General. Washington, DC: Author.

12       zane, bernal, and leong

I Overview

1 COMPONENTS OF EVIDENCE-BASED PRACTICE IN PSYCHOLOGY RONALD F. LEVANT AND HEATHER A. SPERRY

Psychologists have long been involved with evidence-based approaches to patient care, based on the scientist–practitioner model that is at the heart of professional psychology. Evidence for mental health interventions can be defined in many ways and has been the subject of major debates (Norcross, Beutler, & Levant, 2006). Some psychologists believe that psychological interventions should be based solely on randomized clinical trials (RCTs), whereas others believe that other forms of evidence have their value. Building consensus on the definition of evidence and ensuring that evidence-based practice in psychology (EBPP) recognizes not only the research but also the clinician’s expertise and the patient’s preferences, values, and culture are important to the future of the profession and to quality patient care. When Ronald Levant was president of the American Psychological Association Portions of this chapter are adapted from “Evidence-Based Practice in Psychology,” by the American Psychological Association 2005 Presidential Task Force on Evidence-Based Practice, 2006, American Psychologist, 61, pp. 271–285. Copyright 2006 by the American Psychological Association. http://dx.doi.org/10.1037/14940-002 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

15

(APA) in 2005, he believed it was vital that APA speak with one voice on the issue to avoid potential confusion among members, the public, media, legislators, health officials, and third-party payers. Hence he sponsored an APA Presidential Initiative on Evidence-Based Practice in Psychology in 2005. This presidential initiative aimed to affirm the importance of attending to multiple sources of research evidence and to affirm that good psychological practice based on evidence is also based on clinical expertise and patient values. The mission of the APA 2005 Presidential Task Force on Evidence-Based Practice in Psychology was three-fold (referred to as a three-legged stool), corresponding to the three components of the Institute of Medicine (2001) definition of evidence-based practice (“Evidence-based practice is the integration of best research evidence with clinical expertise and patient values,” p. 147): 1. to consider how a broader view of research evidence, one that inclusively considers multiple research designs, research in public health, health services research, and health care economics, should be integrated into a definition of EBPP; 2. to explicate the application and appropriate role of clinical expertise in treatment decision making, including a consideration of the multiple streams of evidence that must be integrated by clinicians and a consideration of relevant research regarding the expertise of clinicians and clinical decision making; and 3. to articulate the role of patient values in treatment decision making, including a consideration of the role of ethnicity, race, culture, language, gender, sexual orientation, religion, age, and disability status and the issue of treatment acceptability and consumer choice. This task force included a diverse group of 18 scientists and practitioners. Areas of expertise included clinical expertise and decision making, health services research, public health and consumer perspectives, treatment outcome and process research, full-time practice, clinical research and diversity, health care economics, and evidence-based practice research, training, and applications. The task force brought together people who would not have been likely to attend each others’ programs at the APA annual convention and those who worked in mixed groups to hear and understand the interlaced nuances involved in every issue. The task force developed two documents. The first was a policy statement for APA governance action. The second document was a report that elaborated on the policy statement, supporting a broad conceptualization of EBPP. Both documents offered the following definition: “Evidence-based practice in psychology (EBPP) is the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and 16       levant and sperry

preferences” (American Psychological Association 2005 Presidential Task Force on Evidence-Based Practice, 2006, p. 273). These efforts succeeded. At its August 2005 meeting, the APA Council of Representatives adopted the policy statement and received the report of the task force. In addition, the report was published in American Psychologist (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). The policy statement and report have been quite influential. Right after it was adopted by APA, the report was adopted verbatim by the Norwegian Psychological Association (Norsk Psykologforening, 2007). Furthermore, for the past 9 years, APA has used the policy statement and report as a basis for examining various types of initiatives, including proposed legislation in the area of EBPP to ensure consistency with this foundational APA policy. One of the most expansive examples of EBPP’s impact has been its role in APA’s development of clinical practice guidelines, which are a widely accepted way of synthesizing evidence for decision making. This effort is built squarely on the foundation of the APA policy on EBPP. Not only is “best available research” one of the key elements considered in guideline development, but best practices in guideline development have a place for consideration of patient values and preferences and it is understood that end users will use such guidelines in the context of other existing evidence, clinical expertise, and patient values and preferences. Moreover, the APA has referenced the EBPP policy in activities related to outcomes assessment and accountability and frequently references the policy when members of the public ask APA questions such as “What is APA’s policy on such-and-such treatment?” and APA staff respond more generically about supporting evidence-based practice (L. Bufka, personal communication, January 9, 2015). Finally, the APA Practice Organization (APAPO) has used the EBPP policy to advocate for Medicare payment reform. In an effort to provide some level of certainty regarding Medicare payment rates and ensure psychologists are paid appropriately and fairly, APAPO promoted EBPP efficiencies in advocacy communications to lawmakers during the development of bipartisan legislation to permanently repeal the Medicare sustainable growth rate formula and replace it with a new payment and delivery system to improve patient care and restrain cost growth. Though Congress failed to enact such legislation in 2014, the deep (24%) Medicare cut that psychologists, physicians, and other health care professionals faced was averted until March 31, 2015, as a result of successful advocacy by APAPO and the broader health care community. APAPO will continue to stress the value of integrating best research evidence with clinical expertise and patient values as they seek to address the deficiencies in Medicare’s payment system (L. Stine, personal communication, January 9, 2015). Finally, it is important to point out that the report (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006) has components of evidence-based practice     

17

been cited more than 500 times in the scholarly literature and has been widely disseminated among APA divisions and state provincial and territorial psychological associations, and the definition has been widely employed in the training community. In short, it stands as the definition of EBPP. This chapter provides a general overview of the components of EBPP, summarizing the major points in the task force report (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). Key themes are illustrated by recent literature on culturally informed EBPP for members of ethnic minority groups. But first, it is important to distinguish EBPP from other concepts. DISTINGUISHING EVIDENCE-BASED PRACTICE IN PSYCHOLOGY FROM OTHER CONCEPTS EBPP is thought to be likely to promote effective psychological practice, improve patient outcomes, and enhance public health. EBPP applies empirically supported principles of psychological assessment, case formulation, therapeutic relationship, and intervention. It is important at the outset to clarify the relation between EBPP and evidence-based treatments (EBTs, formerly known as empirically validated treatments, then as empirically supported treatments, and occasionally as research-supported psychological treatments). EBPP is the more comprehensive concept. EBTs refer to specific treatments for a specific problem under specific circumstances. EBPP starts with the patient and asks what research evidence (not limited to results from RCTs) will assist the psychologist in achieving the best outcome. In addition, EBTs are specific psychological treatments that have been shown to be efficacious in RCTs, whereas EBPP encompasses a broader range of clinical activities (e.g., psychological assessment, case formulation, therapy relationships). As such, EBPP articulates a decisionmaking process for integrating multiple streams of research evidence—including but not limited to RCTs—into the intervention process. The following sections explore in greater detail the three major components of this definition—best available research evidence, clinical expertise, and patient characteristics—and their integration. BEST AVAILABLE RESEARCH EVIDENCE A large body of evidence drawn from studies using a variety of research methods attests to the effectiveness of psychological practices. Psychological interventions have been found to be safe and effective for children, youth, adults, and older adults across a wide range of psychological, addictive, 18       levant and sperry

health, and relational problems, through a vast number of individual studies and meta-analyses since the 1970s. Indeed, the effect sizes of psychological interventions aimed at patients of any age are often comparable to widely accepted medical treatments. Further, recent research has shown that compared with medications and other alternative approaches, the effectiveness of psychotherapy is particularly persistent. Research suggests that psychotherapy can and often does pay for itself in terms of medical-cost offset (the reduction of medical and surgical expenses in health care settings that integrate behavioral with physical health care), increased productivity in the workplace, and quality of life. Researchers and practitioners tend to agree that psychological practice needs to be based on evidence and that research needs to have both internal and external validity. Research will not always address all practice needs. One major issue in integrating research into day-to-day practice is that the research has generally been conducted with primarily White samples and it is unclear whether these efficacy and effectiveness results can be generalized to minority and marginalized populations. Sue and Zane (2006) examined the ways in which EBTs aim to reduce disparities yet also recognized the ways in which EBTs have failed to do so. With regard to ethnic minorities, little research has been conducted to determine the efficacy and effectiveness of EBTs with people of color. (People of color refers to anyone who is not White, reflecting a common experience of racism.) In addition, as previously mentioned, there exists a need to broaden the definition of evidence so it that does not solely rely on RCTs, which often do not include a large sample of ethnic minority people. Finally, research on culturally competent interventions is new, and thus studies on the efficacy and effectiveness of interventions that have actually employed minority and marginalized groups as participants have not usually utilized these newer interventions. This problematic underrepresentation of ethnic minority persons in EBT research may be due in part to complex systemic problems. Historically, research in American psychology has relied upon White participants, as noted by Guthrie (2004) in his book titled Even the Rat Was White. This may be in part because it often is costly and challenging to reach out to ethnic minorities in research. Hence, convenience samples are too often used in research, particularly when the research is time-sensitive, which often results in few ethnic minorities in the sample. In addition, research on ethnic minorities can be challenging because of the lack of culturally sensitive and valid measures for use with ethnic minorities. On a related point, there is a significant question about the extent to which measures that are used in efficacy or effectiveness trials have been examined for measurement invariance across racial and ethnic groups. To not examine for measurement invariance leaves the research open to the possibility of construct bias, which implies that an outcome measure components of evidence-based practice     

19

measures something different in one group (e.g., Whites) than it does in another group (e.g., Latinos; Meade & Lautenschlager, 2004). Finally, research on ethnic minorities is often complicated by disparities, prejudice, inequities, and differential treatment, which can be difficult to partial out or control when examining variations within and across cultures and ethnicities. Hence it came as no surprise when the U.S. Surgeon General (U.S. Department of Health and Human Services, 2001) stated that a large gap existed between research and practice in regard to racial and ethnic minorities. Nonetheless, research on practice has made significant progress in investigating a broad range of issues and is providing evidence that is more responsive to day-to-day practice. Because EBPP can provide valuable information in regard to cultural competency, particularly when attention is paid to ethnic and cultural research, there is sufficient consensus to move forward with the principles of EBPP. Best available research evidence refers to scientific results related to intervention strategies, assessment, clinical problems, and patient populations in laboratory and field settings as well as to clinically relevant results of basic research in psychology and related fields. Many different kinds of research designs contribute to EBPP, including clinical observation, qualitative research, systematic case studies, single-case experimental designs, epidemiological research, ethnographic research, process–outcome studies, cost-effectiveness, cost–benefit analysis, treatment utilization, effectiveness research, efficacy research, and meta-analysis. Each of these types of research makes its own unique contribution (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). With respect to evaluating research on specific interventions, the APA policy statement titled “Criteria for Evaluating Treatment Guidelines” (American Psychological Association, 2002) identified two widely accepted dimensions. The first dimension is treatment efficacy, which refers to the systematic and scientific evaluation of whether a treatment works. The criteria listed the major types of research evidence in ascending order as to their contribution to conclusions about efficacy as follows: clinical opinion, observation, and consensus among recognized experts; systematized clinical observation; and sophisticated empirical methodologies, including quasi experiments and RCTs. The second dimension is clinical utility, which refers to the applicability, feasibility, and usefulness of the intervention in the local or specific setting where it is to be offered. This dimension also includes determination of the generalizability of an intervention to the community whose efficacy has been established in the laboratory. Evidence on clinical utility includes attention to generality of effects across diverse patients, therapists, settings; the robustness of treatments across various modes of delivery; the feasibility with which treatments can be delivered to patients in realworld settings; and the costs associated with treatments. 20       levant and sperry

It is a requirement of EBPP that psychologists recognize the strengths and limitations of evidence obtained from a variety of types of research. Research indicates that multiple factors contribute to the success of practice: the treatment method, the individual psychologist, the treatment relationship, and the patient. EBPP will consider all of these determinants and their optimal combinations. Not only is psychological practice a complex relational and technical enterprise, but clinical and research attention must be applied to multiple, interacting sources of treatment effectiveness. Empirical data remain limited for many disorders, problem constellations, and clinical situation. When the data are lacking, it becomes necessary for clinicians to use their best clinical judgment and knowledge of the best available research evidence that may exist to develop coherent treatment strategies. It would be beneficial for researchers and practitioners to join together to ensure that any research that is available on psychological practice is clinically relevant and internally valid (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). CLINICAL EXPERTISE The task force report (APA 2005 Presidential Task Force on EvidenceBased Practice, 2006) noted that clinical experts are not infallible, as they are subject to the same errors and biases that all humans are, stemming from cognitive heuristics that are generally adaptive as well as from emotional reactions, which generally guide adaptive behavior as well but can also lead to biased or motivated reasoning. Nonetheless, the report noted that there is consistent evidence of enduring and significant differences between experts and novices undertaking complex tasks. With regard to the patient’s characteristics and preferences, clinical expertise is vital. Clinicians identify and integrate the best research evidence that will best serve patients and assist them in achieving their therapeutic goals. Because psychological training uses the scientist–practitioner model, psychologists develop a clinical expertise informed by scientific expertise. This expertise allows the psychologist to understand and integrate scientific literature as well as utilize hypothesis testing in the clinical setting. Clinical expertise encompasses an array of competencies, including diagnostic competence, case formulation, and treatment planning; clinical decision making, treatment implementation, and monitoring of progress; interpersonal skills; continual self-reflection and incorporation of new knowledge and skills; and understanding the influence of individual and cultural differences on treatment. Clinical expertise is manifested by a well-planned approach. Although eclectic or integrative approaches are often used in a therapeutic setting and components of evidence-based practice     

21

nonspecific engagement such as building rapport can impact psychological treatment, psychologists must often rely on their own judgment within the therapeutic setting. They need to use well-articulated case formulations, knowledge of relevant research, theoretical conceptualizations, and clinical experience to identify or develop interventions designed to attain desired outcomes. Although some patients may present to therapy with a welldefined and well-researched concern or diagnosis regarding which there is evidence for using a particular treatment, this is not always the case. When a body of evidence supports a particular treatment for a particular patient concern or diagnosis, this evidence should be considered in formulating a treatment plan and used to provide a clear rationale for the desired course of treatment. However, often treatment evidence is sparse for patient problem constellations, patient populations, and clinical situations. Under such circumstances, EBPP can be helpful in identifying clinical expertise, understanding and using best available evidence, and monitoring and modifying treatment as appropriate. Practitioners were excited about the explicit recognition of clinical expertise in the policy statement and report of the task force on EBPP. For instance, Brooke (2006) stated that “the importance of context and clinical expertise are once again recognized by APA as essential to professional competence in an evidence based practice” (p. 23). As a psychologist in independent practice, Hunsberger (2007) argued that “by virtue of daily personal interactions with clients, clinicians are arguably the true experts on the nature and accessing of psychological evidence” (p. 614). PATIENT CHARACTERISTICS, VALUES, AND CONTEXT Research findings on what treatment works best for whom (Nathan & Gorman, 2002) can guide effective practice. However, psychological services are most likely to be effective when responsive to the patient’s specific problems, strengths, personality, sociocultural context, and preferences. Because psychology has been affiliated with a growing amount of literature on diversity, individual differences, and developmental change, those in the field are in a good position to effectively integrate research and clinical experience to best understand varying patient characteristics in EBPP (Sue, Zane, & Young, 1994). As stated in the task force report, “It is important to know the person who has the disorder in addition to knowing the disorder the person has” (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006, p. 279). EBPP aims to consider the goals, values, religious beliefs, worldviews, and treatment preferences in accordance with the experience of the psychologist 22       levant and sperry

as well as an understanding of the available research. Current debates about the role of patient characteristics in EBPP are based on the question of how best to approach the treatment of patients whose sociocultural characteristics (e.g., gender, gender identity, sexual orientation, race, ethnicity, social class, religion, family structure, and ability/disability) and problems (e.g., comorbidity) may differ from those of samples studied in research. This question is a matter of active conversation within psychology, and attention is increasingly being paid to the generalizability and transportability of psychological interventions. According to the task force report (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006), specific concerns revolve around four major questions: 1. To what extent do social factors and cultural differences necessitate different forms of treatment or can interventions widely tested in majority populations be readily adapted for patients with different ethnic or sociocultural backgrounds? 2. To what extent do widely used interventions adequately attend to developmental considerations, both for children and adolescents and for older adults? 3. To what extent do cross-diagnostic patient characteristics, such as personality traits or constellations, moderate the impact of empirically tested interventions? 4. To what extent do comorbidity and polysymptomatic presentations moderate the impact of treatment? Available data indicate that a variety of cross-diagnostic patient-related variables, such as functional status, readiness to change, and level of social support, influence outcomes. Other patient characteristics that may also influence outcomes include variations in presenting problems or disorders, etiology, concurrent symptoms or syndromes, and behavior; chronological age, developmental status, developmental history, and life stage; the sociocultural characteristics listed above; current environmental context, stressors (e.g., unemployment, a recent life event) and social factors (e.g., institutional racism, health care disparities); and personal preferences, values, and preferences related to treatment (e.g., goals, beliefs, worldviews, treatment expectations). Available research indicates that strategies and relationships need to be matched to each particular patient. Many presenting symptoms are similar across patients. However, symptoms or disorders that appear to be similar often have significant differences with respect to the psychological processes that create or maintain them as well as prognosis. Furthermore, the majority of patients present with a variety of symptoms or comorbidity rather than a single, uniform disorder. An intervention intended to treat one symptom may likely impact others, and the components of evidence-based practice     

23

presence of multiple conditions moderates treatment response. More recent research also suggests that individual differences in personality often underlie diagnoses and symptoms and can account for a majority of the comorbidity among syndromes widely documented in research (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). EBPP also requires that psychologists pay attention to factors as they relate to lifespan development and the current life stage of the patient. Research suggests that attachment, socialization, cognitive, social–cognitive, gender, moral, and emotional development are just some of the developmental processes that are crucial in understanding psychopathology, particularly in the treatment of children, adolescents, families, and older adults (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). Finally, EBPP requires attention to the sociocultural characteristics listed above (see also American Psychological Association, 2003; Division 44/ Committee on Lesbian, Gay, and Bisexual Concerns Joint Task Force on Guidelines for Psychotherapy with Lesbian, Gay, and Bisexual Clients, 2000). These characteristics require attention because of their impact on personality, values, worldviews, relationships, psychopathology, and attitudes toward treatment. The patient’s culture may not only influence the nature and expression of psychopathology but also inform her or his understanding of psychological and physical health and illness. In addition, cultural values, beliefs, and social factors such as implicit racial biases may also influence help-seeking patterns, using and receiving help from others, the presentation and reporting of symptoms, fears and expectations about treatment, and desired outcomes. Through their training, psychologists are also able to recognize and reflect on the ways their own personality variables, values, and context may interact with those of the patient. Although mental health fields often emphasize the importance of culturally competent practices, the literature remains lacking in research on culturally adapted treatments (Griner & Smith, 2006). In addition to the EBPP movement, multicultural psychologists have noted that a monocultural bias exists within the field (Gone, 2009). Within the field of psychology, procedural and research norms are not typically culturally sensitive, which causes problems for clinicians who do not have adequate research on work with diverse clients to inform their practice. One major concern regarding a lack of culturally informed EBPP is that current empirically based practice and treatments are based on the life experiences and needs of European Americans, which can be assimilative, alienating, or harmful for racial and ethnic minority persons (Gone, 2009). More psychologists should inform themselves of and advocate for services that are adequate and appropriate for work with diverse clients. EBPP recognizes that culture is made up of shared values, history, knowledge, rituals, and customs that provide a sense of community and meaning, 24       levant and sperry

yet even within the same culture, these values are held by individuals who may interact with their culture differently (La Roche & Christopher, 2009). In addition, EBPP recognizes that assumptions about culture can be problematic because of individual differences and heterogeneity within a cultural group, which may also include sociocultural factors such as unemployment, lack of insurance, and other situational factors. Therefore, “psychologists must attend to the individual person to make the complex choices necessary to conceptualize, prioritize, and treat multiple symptoms” (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006, p. 279). Attending to the individual person involves being aware of a variety of characteristics: gender, gender identity, sexual orientation, social class, religion, family structure, ability/ disability, ethnicity, and race. Race is a social construction that is used to group people into categories based on perceived physical attributes, appearance, ancestry, and other factors. Race is also tied to power, status, and opportunity. Although societal attitudes and public policy have advanced to reinforce social equality, in Western cultures, Whiteness is distinctly related to advantage and opportunity. Therefore it is clear that race is an interpersonal and political process and intimately impacts clinical practice and health care quality (Smedley & Smedley, 2005). Patients and clinicians may identify with racial groups, but the importance of race in clinical practice is relational rather than solely an individual attribute. Considerable evidence from many fields (Institute of Medicine, 2003) shows that racial power differentials, systemic biases, and implicit stereotypes based on race or ethnicity contribute to inequality in health care services that many people of color experience. The treatment process, relationship, and outcome may be affected by race, ethnicity, and culture, so it is imperative that clinicians carefully consider their impact. As an example, Gone (2009) examined Native American historical trauma to expand current perspectives on healing. The author identified four insights in regard to bridging the gap between EBTs and culturally sensitive treatment through EBPP. First, he intended to depathologize individual community members in the study by identifying that their distress was at least in part due to historical oppression rather than personal failure. Second, outcome measures were based on a more robust state of holistic wellness rather than typically used indicators in Western culture (e.g., reduction in posttreatment symptom scores). A third insight concerned the nature of the allowable evidence for appraisal of outcomes. Whereas EBTs gather evidence on outcomes via strict procedural and quantitative methods, within Native American culture the testimony of the counselors or patients is considered to meet a high standard of validity, contrary to how such evidence is viewed in EBT studies. The fourth insight pertains to the political agenda behind treatment. EBT uses rigorous outcome evidence to assist professionals in components of evidence-based practice     

25

providing psychological services that reduce distress, but within this Native American culture, the aim of treatment was decolonization, which is defined as the intentional, collective, and reflective self-examination by persons who were once colonized for remedial action (Gone, 2009). Although this is merely one example of culturally diverse mental health treatment, it shows how systemic bias could have negatively impacted Native American clients. The clinicians in this study were also Native American, so racial power differentials and implicit stereotypes were likely not large factors in treatment. However, psychologists who attempt to use this treatment or a variation of it with Native American clients who do not also identify as Native American should check their implicit biases about Native American culture and should address racial power differentials within treatment. Further research on indigenous forms of healing would assist psychologists in their ability to provide appropriate and efficacious treatment to Native American persons. Research has shown that people of color utilize therapy only when their problems become severe, drop out of treatment prematurely, and overall underutilize therapy (Griner & Smith, 2006). This situation may occur for a few reasons. First, cultural values of people of color may clash with traditional mental health practices, and people of color, like most people, are less likely to seek out treatment when their values and beliefs are incongruent with available interventions. Second, people of color may be mistrustful of mental health services because of historic racial disparities or having limited access to therapists of their native language or own ethnic background. Finally, there tends to be limited access to therapy in communities where people of color reside. Services should be provided within the community, respecting the spiritual traditions and family values of people of color. Both recent and chronic stressors within the patient’s social and environmental context are important in case formulation and treatment planning. Numerous factors can have an enormous influence on mental health, adaptive functioning, treatment seeking, and psychological, social, and financial patient resources. These may include sociocultural and familial factors, social class, and broader social, economic, and situational factors (e.g., unemployment, family disruption, lack of insurance, recent losses, prejudice, immigration status). For psychotherapy to be effective, patients and clinicians must negotiate their relationship and work together in a way that will lead to positive outcomes. Because patients and clinicians must work together, the values and preferences (e.g., goals, beliefs, preferred modes of treatment) of the patient are a central component of EBPP. Because of culture and individual preferences, patients can have strong opinions on the types of treatment and desired outcomes. The psychologist must therefore ensure that the patient is informed in regard to the cost and benefits of the different treatment choices. 26       levant and sperry

EBPP aims to provide patients with informed choice with regard to effective alternative interventions. A combination of patient preferences and the psychologist’s judgment, based on evidence and clinical experience and expertise, help to enhance treatment and provide effective practice (APA 2005 Presidential Task Force on Evidence-Based Practice, 2006). Whaley and Davis (2007) highlighted the complementary nature of cultural competence and EBPP in mental health services for the psychological treatment of people of color and supported the use of cultural adaptations of the existing research base for people of color. CONCLUSION EBPP is a combination of best available research, clinical expertise, and the recognition of and integration of patient characteristics, culture, and treatment preferences. Its goal is to promote more effective psychological practice and enhance public health. EBPP requires the appreciation of the value of multiple sources of scientific evidence. Psychologists should collaborate with patients to make appropriate clinical decisions and should make patients aware of probable costs, benefits, resources, and options. Untrained persons who are uninformed about a particular case should never make treatment decisions. It is necessary to recognize that the research has not yet been able to address all decisions relevant to client concerns, so continued monitoring of patient progress and adjustment as needed are essential to EBPP. Wampold, Goodheart, and Levant (2007) clarified and elaborated the task force report and policy statement. These authors addressed some of the concerns noted in this chapter and clarified that the main goal of the task force was “to create a scheme that would suggest how evidence should be used to design and offer services that will benefit patients and to assure the public and the health care system that psychologists are providing evidence-based services” (p. 618). Wampold et al. also stated that although not all psychologists will agree with every aspect of EBPP, the task force work represented a very significant accomplishment for the field. REFERENCES American Psychological Association. (2002). Criteria for evaluating treatment guidelines. American Psychologist, 57, 1052–1059. http://dx.doi.org/10.1037/ 0003-066X.57.12.1052 American Psychological Association. (2003). Guidelines on multicultural education, training, research, practice, and organizational change for psychologists. American Psychologist, 58, 377–402. components of evidence-based practice     

27

American Psychological Association 2005 Presidential Task Force on EvidenceBased Practice. (2006). Evidence-based practice in psychology. American Psychologist, 61, 271–285. http://dx.doi.org/10.1037/0003-066X.61.4.271 Brooke, R. (2006, Summer). Controversial discussions II: The return of expertise in evidence based practice. Psychologist-Psychoanalyst, XXVI(3), 23–26. Division 44/Committee on Lesbian, Gay, and Bisexual Concerns Joint Task Force on Guidelines for Psychotherapy with Lesbian, Gay, and Bisexual Clients. (2000). Guidelines for psychotherapy with lesbian, gay, and bisexual clients. American Psychologist, 55, 1440–1451. http://dx.doi.org/10.1037/0003066X.55.12.1440 Gone, J. P. (2009). A community-based treatment for Native American historical trauma: Prospects for evidence-based practice. Journal of Consulting and Clinical Psychology, 77, 751–762. http://dx.doi.org/10.1037/a0015390 Griner, D., & Smith, T. B. (2006). Culturally adapted mental health intervention: A meta-analytic review. Psychotherapy: Theory, Research, Practice, Training, 43, 531–548. http://dx.doi.org/10.1037/0033-3204.43.4.531 Guthrie, R. V. (2004). Even the rat was White: A historical view of psychology (2nd ed.). Boston, MA: Allyn & Bacon. Hunsberger, P. H. (2007). Reestablishing clinical psychology’s subjective core. American Psychologist, 62, 614–615. http://dx.doi.org/10.1037/0003-066X62.6.614 Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academies Press. Institute of Medicine. (2003). Unequal treatment: Confronting racial and ethnic disparities in health care (B. D. Smedley, A. Stith, & A. R. Nelson, Eds.). Washington, DC: National Academies Press. La Roche, M. J., & Christopher, M. S. (2009). Changing paradigms from empirically supported treatment to evidence-based practice: A cultural perspective. Professional Psychology: Research and Practice, 40, 396–402. http://dx.doi.org/10.1037/ a0015240 Meade, A. W., & Lautenschlager, G. J. (2004). A comparison of item response theory and confirmatory factor analytic methodologies for establishing measurement equivalence/invariance. Organizational Research Methods, 7, 361–388. http:// dx.doi.org/10.1177/1094428104268027 Nathan, P. E., & Gorman, J. M. (Eds.). (2002). A guide to treatments that work (2nd ed.). London, England: Oxford University Press. Norcross, J. C., Beutler, L. E., & Levant, R. F. (Eds.). (2006). Evidence-based practice in mental health: Debate and dialogue on the fundamental questions. Washington, DC: American Psychological Association. http://dx.doi.org/10.1037/11265-000 Norsk Psykologforening, N. (2007). Prinsipperklaering om evidensbasert psykologisk praksis [Declaration of principles of evidence-based practice in psychology]. Tidsskrift for Norsk Psykologforening [Journal of the Norwegian Psychological Association], 44, 1127–1128.

28       levant and sperry

Smedley, A., & Smedley, B. D. (2005). Race as biology is fiction, racism as a social problem is real: Anthropological and historical perspectives on the social construction of race. American Psychologist, 60, 16–26. Sue, S., & Zane, N. (2006). Ethnic minority populations have been neglected by evidence-based practices. In J. C. Norcross, L. E. Beutler, & R. F. Levant (Eds.), Evidence-based practice in mental health: Debate and dialogue on the fundamental questions (pp. 329–374). http://dx.doi.org/10.1037/11265-008 Sue, S., Zane, N., & Young, K. (1994). Research on psychotherapy with culturally diverse populations. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 783–817). New York, NY: Wiley. U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity, a supplement to Mental Health: A Report of the Surgeon General. Washington, DC: Author. Wampold, B. E., Goodheart, C. D., & Levant, R. F. (2007). Clarification and elaboration on evidence-based practice in psychology. American Psychologist, 62, 616–618. http://dx.doi.org/10.1037/0003-066X62.6.616 Whaley, A. L., & Davis, K. E. (2007). Cultural competence and evidence-based practice in mental health services: A complementary perspective. American Psychologist, 62, 563–574. http://dx.doi.org/10.1037/0003-066X.62.6.563

components of evidence-based practice     

29

2 PSYCHOTHERAPY OUTCOME RESEARCH WITH ETHNIC MINORITIES: WHAT IS THE AGENDA? ANNA S. LAU, DORIS F. CHANG, SUMIE OKAZAKI, AND GUILLERMO BERNAL

In light of the underrepresentation of ethnic minorities in clinical trials (Mak, Law, Alvidrez, & Pérez-Stable, 2007), the push to adopt evidence-based treatments (EBTs) has generated debate concerning their relevance, applicability, and value for diverse communities (e.g., Bernal & Scharrón-del-Río, 2001; Hall, 2001). Since efforts began to shepherd the movement of clinical practice toward the uptake of EBTs, some have questioned the wisdom of producing lists of treatments with evidence of efficacy based predominantly on tests with White, middle-class, Englishspeaking participants. The Task Force on Promotion and Dissemination of Psychological Procedures (1995) acknowledged that no treatment could be said to meet the basic criteria for established efficacy for an ethnic minority group (Chambless et al., 1996). At the time of the 1995 report, the few existing trials with minority inclusion had sample sizes that were deemed

http://dx.doi.org/10.1037/14940-003 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

31

too small to be included in the evidence base. However, the task force recommended that treatments identified as efficacious with Whites be offered to ethnic minorities. Bernal and Scharrón-del-Río (2001) argued that these lists of treatments must at least be accompanied by caveats regarding the limited external validity of the evidence base. They opposed insistence on generalization in the absence of data, stating that “to the extent that a specific theory of psychotherapy is developed, constructed, and tested in a particular cultural group, packaged as empirically sound, and imposed on another there may be a new form of cultural imperialism” (p. 333). Sue and Zane (2006) asserted that assuming generalization reduces the pressure to conduct research on ethnic minorities to empirically assess external validity of EBTs. In contrast, others have emphasized the value of the extant evidence regarding treatment outcomes for ethnic minorities. Miranda et al. (2003) observed that increased suffering from untreated mental illness would result from asserting that there is insufficient data to recommend that ethnic minorities seek evidence-based care. The recommendation to encourage access to EBTs across groups is based on four sources of support: clinical evidence that EBTs can be used effectively with ethnic minorities, a small international literature on treatment outcomes, a lack of evidence that EBTs are not beneficial for ethnic minorities, and emerging findings that EBTs are effective for diverse samples. On both sides of the debate, there is agreement regarding the dearth of efficacy research with ethnic minorities. Our goal in this chapter is to parse priorities for continued research on treatment outcomes for ethnic minorities. We structure our review around three paradigms in treatment outcome research: generalizability of EBTs for ethnic minorities, cultural adaptation of EBTs for ethnic minorities, and innovation with culturally sensitive therapies (CSTs). In Figure 2.1, we highlight a set of parallel research questions regarding treatment outcomes for ethnic minorities across these three areas. We present examples of trials designed to test these questions. Although we present this framework to encapsulate some primary methodological questions facing the field, it is a challenge to organize research neatly into these distinct categories. As we note along the way, the classification of various interventions as EBTs, adapted EBTs (A-EBTs), and CSTs is problematic. These decisions are subject to one’s interpretation of the degree to which cultural responsiveness was central in either modifying an existing intervention or conceiving a novel intervention. We also discuss the extent to which each of these research questions represents differing agendas advanced from differing philosophical and epistemological perspectives, with distinct implications for intervention science. 32       lau et al.

psychotherapy outcome research     

Generalizability of Treatment Effects EBT Efficacy & Effectiveness

Adaptation of Treatment A-EBT Efficacy & Effectiveness

Innovations in Treatment Development CST Efficacy & Effectiveness

1. Do EBTs work for minorities? Single ethnic group trials

4. Do A-EBTs work for targeted minorities? Treatment–Control trials

8. Do CSTs work for minorities? Treatment–Control trials

Clinical outcomes: statistical & clinical significance Engagement: recruitment, attrition, participation, satisfaction

Clinical outcomes: statistical & clinical significance Engagement: recruitment, attrition, participation, satisfaction

Clinical outcomes: statistical & clinical significance Engagement: recruitment, attrition, participation, satisfaction

2. Do EBTs work as well for minorities? Multiethnic trials

5. Do A-EBTs work better than EBTs? Treatment–Treatment trials

9. Do CSTs work as well as EBTs? Treatment–Treatment trials

Testing treatment by ethnicity interactions Meta-analytic studies

Clinical outcomes & engagement Empirical test of cultural competence principles Power considerations

Clinical outcomes & engagement Power considerations

3. What are the mechanisms of action in EBTs with minorities?

6. Do A-EBTs work better across groups? Multiethnic Treatment–Treatment trials

10. What are the mechanisms of action in CSTs with minorities?

Is effective therapy process generalizable? Are core treatment components equally salient?

Adaptation for some optimizing treatment for all? Testing treatment by ethnicity interactions

Is effective therapy process generalizable? Culturally focused theory testing

7. What are the mechanisms of action in A-EBTs with minorities? Is effective therapy process generalizable? Are core treatment components equally salient? Empirical test of cultural competence principles

33

Figure 2.1.  Central research questions in treatment outcome research with minorities. CST = culturally sensitive therapy; EBT = evidencebased treatment; A-EBT = adapted evidence-based treatment.

GENERALIZABILITY OF EVIDENCE-BASED THERAPIES FOR ETHNIC MINORITIES In this section, we address two fundamental questions regarding treatment outcomes with minorities. 77

77

Question 1: Do EBTs work with ethnic minority groups? When tested with ethnic minority samples, can significant improvements in symptoms and functioning be attributed to the effects of treatments that were developed and validated with majority samples? In answering this question, we reference the randomized controlled trial (RCT) literature from studies with inclusive or minority samples. Question 2: Do EBTs work as well with ethnic minorities as they do with Whites? Here, we reference studies that examine whether treatment outcomes are moderated by race/ethnicity. Recent reviews provide some affirmation in response to both important questions.

In identifying relevant studies, we relied initially on two published reviews and expanded to include more recent trials. Miranda and colleagues (2005) reviewed published efficacy, effectiveness, and prevention trials as well as naturalistic treatment studies for ethnic minority adults and youth. Huey and Polo (2008) conducted a meta-analysis of EBT effects with ethnic minority children and adolescents. Both reviews concluded that a growing literature supports generalized effects of EBTs. Huey and Polo’s (2008) review was circumscribed to interventions meeting Chambless and Hollon’s (1998) efficacy criteria and to trials meeting Nathan and Gorman’s (1998) methodological criteria. Treatments were classified as well-established, probably efficacious, or possibly efficacious for minority youth if supporting studies met one of the following additional criteria: (a) at least 75% of participants in the sample were ethnic minorities, (b) separate analyses with ethnic minorities showed superiority to control conditions, or (c) analyses indicated that ethnicity did not moderate key treatment outcomes or that treatment was effective with ethnic minorities despite moderator effects. With these definitions, no EBT met criteria as well-established for ethnic minority youth primarily because effects had not been replicated by independent investigators. However, under criterion (a), several interventions could be classified as probably efficacious for ethnic minorities. The two reviews converge on the conclusion that the most rigorous literature supports cognitive–behavioral therapy (CBT) for depression with African American adults and Latino adults and youth (e.g., Miranda et al., 2003; Wells, Miranda, Bruce, Alegría, & Wallerstein, 2004) and Latino youth 34       lau et al.

(e.g., Rosselló & Bernal, 1999; Rosselló, Bernal, & Rivera-Medina, 2008) and interpersonal therapy (IPT) for Latino youth (e.g., Mufson, Weissman, Moreau, & Garfinkel, 1999; Rosselló & Bernal, 1999). Also probably efficacious are multisystemic therapy, brief strategic family therapy, multidimensional family therapy, and the Coping Power program for conduct problems among Latino and African American youth (Henggeler, Melton, & Smith, 1992; Liddle, Rowe, Dakof, Ungaro, & Henderson, 2004, Lochman & Wells, 2004; Santisteban et al., 2003). More limited evidence supports the possible efficacy of CBT and exposure therapies for anxiety disorders among African Americans (Ginsburg & Drake, 2002). In contrast, there are no RCT data on the effects of EBTs for Asian American and Native American adults, with only meager support from nonrandomized wait-list controlled trials for CBT for depression with Chinese American and Native American samples (Eap & Hall, 2007; Gone & Alcántara, 2007). However, not all studies have demonstrated the superiority of EBT over control groups in RCTs. For example, a pilot RCT study comparing the effectiveness of CBT delivered in school settings versus usual care (UC) among predominantly African American inner-city adolescents with anxiety disorders found no significant difference in effectiveness between CBT and UC (Ginsburg, Becker, Drazdowski, & Tein, 2012). Another RCT of a schoolbased intervention to reduce violence and substance abuse with predominantly (more than 75%) Latino high school students found no significant effects of the intervention on school grade, substance use, or violence (Shetgiri, Kataoka, Lin, & Flores, 2011). Shetgiri et al. (2011) suggested that this intervention, which had been shown to be effective with White and African American students, may not have attended sufficiently to the needs of Latino students. In answer to Question 2, evidence from inclusive trials with Whites and ethnic minorities has yielded mixed results regarding parity in EBT outcomes. Under criterion (c), trauma-focused CBT qualified as probably efficacious given no significant treatment by ethnicity interaction in samples including Whites, African Americans, and Latinos (Cohen, Deblinger, Mannarino, & Steer, 2004). A number of interventions were deemed possibly efficacious given single findings that outcomes were not moderated by ethnicity in samples including Whites, African Americans, and Latinos, including IPT for depression (Brown, Schulberg, Sacco, Perel, & Houck, 1999), CBT for trauma-related problems (e.g., Stein et al., 2003; Zoellner, Feeny, Fitzgibbons, & Foa, 1999), group CBT for anxiety (e.g., Silverman et al., 1999), and parent training (e.g., Reid, Webster-Stratton, & Beauchaine, 2001). In contrast, ethnicity by treatment interactions have been reported for exposure therapy for agoraphobia (Chambless & Williams, 1995) and CBT for depression (e.g., Cardemil, Reivich, & Seligman, 2002; Miranda et al., 2003), with African Americans showing poorer response than have Whites and Latinos. psychotherapy outcome research     

35

A complex example of moderation emerged from the Multimodal Treatment of ADHD study. Main effects indicated an overall ethnic disparity in efficacy disfavoring Latinos and African Americans when children were randomized to medication management (MM), behavior therapy (BT), combination treatment (CT), or UC. However, analyses suggested that African American children fared better than did White children with behavioral substitution (BT vs. UC), and Latinos showed an enhanced multimodal superiority effect (CT vs. MM) compared with Whites (Arnold et al., 2003). A focus on main effect disparities in a multiarm trial can obscure implications for ethnic minorities. Huey and Polo (2008) concluded that data on ethnicity by treatment interactions offer no clear support for either the disparities or the equivalence position on EBT efficacy in response to Question 2. To date, the most compelling evidence of EBT generalizability exists for CBT and IPT for depression and behavioral treatments for conduct problems among Latino and African American adolescents and adults. There is more limited but growing support for CBT and exposure therapies for anxiety disorders and CBT models for trauma-related problems in these two minority groups. There is a stark absence of data on the efficacy of EBTs with Asian American and Native American adults and children. Notwithstanding these gaps in the evidence base, Miranda and colleagues (2005) concluded that EBTs are “likely appropriate for most ethnic minority individuals” (p. 134) and Huey and Polo (2008) recommended that EBTs be used as first-line interventions for ethnic minority youth. However, unguarded optimism about EBT generalizability may be premature. First, even with the limited number of tests to date, examples of negative findings with ethnic minority groups can be found particularly for African Americans in CBT. Second, many tests of generalizability have been composed of weak tests of treatment by ethnicity interactions and questionable subgroup analyses. Subgroup analyses testing treatment effects separately by ethnicity are criticized because of inflated probability of Type I error associated with multiple tests. Moreover, these analyses do not determine whether subgroup effect sizes differ from one another (Kraemer, Frank, & Kupfer, 2006). Tests of interactions are often inadequate statistically (e.g., relying on examination of simple effects rather than single degrees-of-freedom contrasts; Huey & Polo, 2008). Even when formal tests of moderation are carried out, trials are usually severely underpowered to detect all but very large moderator effects (Hohmann & Parron, 1996). Third, ethnic minorities in existing ethnically inclusive RCTs likely represent more highly acculturated individuals whose social context, language capacity, and exposure to Western mental health practice may make them more amenable to EBTs. Finally, not all the trials reviewed may be relevant to the question of generalizability of standard EBTs. The strongest data support CBT and IPT for depression for Latinos 36       lau et al.

(Huey & Polo, 2008; Miranda et al., 2005); these data were based on years of development with ethnic minorities. For example, the RCTs by Miranda et al. (2003) and Wells et al. (2004) employed manuals that underwent innumerable modifications based on research on depression treatment with lowincome Latinos (Muñoz & Mendelson, 2005). Likewise, the CBT and IPT treatments in the Rosselló and Bernal (1999) and Rosselló et al. (2008) trials were based on the Muñoz et al. manuals and led by pioneers in cultural adaptation research. Last, the brief strategic family therapy tested by Robbins et al. (2008) was based on years of translational research and practice with Cuban American families with adolescents with substance and conduct problems. Thus, these EBT trials may better represent culturally adapted EBTs or even culturally sensitive therapies. EFFICACY OF CULTURALLY ADAPTED EVIDENCE-BASED TREATMENTS FOR ETHNIC MINORITIES Although many experts agree that it is a major public health priority to ensure that ethnic minorities have access to EBTs and evaluate parity within inclusive trials, many maintain that there is also a need to examine strategies for the effective adaptation of EBTs to ensure they are culturally responsive. Lau (2006) argued that adaptation of EBTs should be evidence based, both in indication and in design. Adaptation efforts should be undertaken when there is a plausible threat of EBT generalization failure. Adaptation is indicated when data suggest a poor fit between an EBT and a particular cultural group. Obvious generalization failure occurs when there is inequity in clinical outcomes when the EBT is administered with fidelity and at the optimal dose. Direct evidence would be a finding of a negative clinical outcome (e.g., lack of efficacy with a minority sample, lack of parity in subgroup analyses, significant treatment by ethnicity interaction disfavoring a minority group). Another type of generalization failure can occur when reduced engagement of minorities renders it difficult to deliver what would otherwise be an effective dose of the EBT. Thus, data suggesting differential recruitment, attrition, participation, and satisfaction may all indicate that cultural adaptation could improve outcomes for minorities. The need for adaptation may also be indirectly indicated. For example, psychopathology research suggesting culturally specific phenomenology of disorder would indicate the need to focus intervention on targeting relevant symptom patterns. Alternatively, data suggesting unique patterns of risk and resilience regarding disorder in a minority group could indicate the need to tailor the treatment to address factors that cause or maintain disorder in that group. Consistent with the criteria for EBTs, A-EBTs should also be wellspecified and preferably manualized. Specific modifications to EBT procedures psychotherapy outcome research     

37

or content should be described to permit replication. The interventionist should design adaptations in an a posteriori manner, guided by data showing how the fit of the EBT could be improved for the target group. Adaptations can generally be classified into one of two major categories: those designed to enhance minority engagement in the EBT and those meant to contextualize the therapy content to ensure the fit with needs of the target group. Other modifications to treatment delivery are important for engagement and efficacy with minorities but may not represent cultural adaptations to treatment process or content, per se. These include structural changes, such as providing treatment in the preferred language of the patient, providing supportive services to permit attendance (e.g., child care, transportation), and locating treatment in credible and accessible community sites. Other modifications may be surface-level changes (e.g., depicting ethnic minorities in graphic material, changing examples to reflect culturally familiar activities) that do not fundamentally alter treatment process or core content. Many well-described A-EBTs have appeared in the literature. Foremost among these are the major trials of demonstrating the efficacy of CBT and IPT for depression with Latino and African American participants (e.g., Muñoz & Mendelson, 2005). Although the current review focuses on examples in which the adaptations to the original intervention are described within a published controlled trial, we note that a growing number of carefully constructed, culturally adapted interventions have also been described in the literature but have not yet been evaluated in a published controlled trial (e.g., D’Angelo et al., 2009; Parra Cardona et al., 2012). As an example of adaptation to ensure effective engagement, McCabe, Yeh, Garland, Lau, and Chavez (2005) developed guiando a ninos activos (GANA), a culturally modified version of parent–child interaction therapy (PCIT) for Mexican American families with young children with disruptive disorders. A telephone engagement protocol reduces barriers to attendance, and treatment is framed as an educational program to combat stigma. Pretreatment orientation explains the roles of the teacher, parents, and the child through video examples, and misconceptions about therapy are addressed. An initial assessment of parental beliefs and values dictates how treatment components are presented. Parents complete a questionnaire and an algorithm generates tailored recommendations to flexibly accommodate their values. For example, depending on parent attitudes toward discipline, time-out can be framed as a punitive practice, invoking the term punishment chair (silla de castigo), or as a self-control intervention called a thinking chair (silla de pensativa). Fifty-eight Mexican American parents with a child with a disruptive disorder were randomly assigned to PCIT, GANA, or UC. McCabe and Yeh (2009) found that PCIT and GANA outperformed UC, and although no significant difference in efficacy was associated with 38       lau et al.

cultural adaptation, the mean effect sizes associated with the GANA–PCIT planned contrast were encouraging (Cohen’s d = 0.28 for conduct symptoms, Cohen’s d = 0.42 for parenting). Of note, Matos, Torres, Santiago, Jurado, and Rodríguez (2006) undertook a similar process to arrive at a version of PCIT adapted for Puerto Rican families that included additional time devoted to establishing rapport, a modified time-out procedure for resistant children, and modifications to handouts to increase their appeal. In their wait-list controlled trial, Matos, Bauermeister, and Bernal (2009) found significant treatment effects on attention-deficit/hyperactivity disorder (ADHD) symptoms, conduct problems, parenting stress, and practices. Huey and Pan (2006) evaluated a culturally adapted version of onesession treatment (OST) for Asian Americans with simple phobia. OST is a modeling and exposure-based therapy completed within a single 3-hour treatment session. The investigators designed a set of seven culture-responsive adaptations that promoted the congruence of OST procedures with expert recommendations for engaging Asian American clients in psychotherapy (e.g., using a directive approach emphasizing therapist commands, framing objectives within the culturally congruent goal of developing emotional control, motivating change by aligning with face concerns). Results of their small but tightly controlled trial revealed advantages of adapted OST over standard OST delivered with fidelity (Huey & Pan, 2006). Adapted OST was superior to self-help control on six of eight clinical outcomes, whereas OST was efficacious on four. Effect sizes suggested the relative advantage of adapted OST versus standard OST was often fairly large. An expanded 6-month follow-up, which included additional participants, confirmed that adapted OST was superior to standard OST and the self-help control in reducing general fear and catastrophic thinking in Asian Americans (Pan, Huey, & Hernandez, 2011). Moderator analyses showed that low-acculturated Asian Americans benefited more from the adapted OST, whereas more acculturated participants responded well in both conditions. In a randomized wait-list controlled trial of a clinician-assisted, Internetdelivered CBT (iCBT) for 55 Chinese Australians with depression, Choi et al. (2012) evaluated a modification of the iCBT Sadness program (Perini, Titov, & Andrews, 2009). Cultural adaptations included primarily structural and surface-level changes to promote engagement, including changing the name of the intervention to avoid stigma, translating materials into Chinese, using images of East Asian individuals, modifying terms to be consistent with Chinese values and terminology, and addressing myths about depression and its treatment. Chinese-speaking clinicians provided weekly e-mail or phone contacts to encourage patients to complete six online lessons within the 8-week trial. Moderate to large effect sizes from pre- to posttreatment and pretreatment to 3-month follow-up on self-reported depressive symptoms and psychotherapy outcome research     

39

clinical diagnostic status indicated that this culturally adapted iCBT program was effective for Chinese Australians. Likewise, Cooley-Strickland, Griffin, Darney, Otte, and Ko (2011) described a shallow structural adaptation of a school-based group CBT anxietyprevention intervention originally developed for Australian schoolchildren (Barrett & Turner, 2001) for use with urban African American youth exposed to community violence. The modifications included changing examples to be relevant to urban African American children (e.g., types of animals, community violence) and treatment delivery by African American group leaders. In an RCT with 93 African American students (ages 8–12), both intervention and wait-list control groups showed reductions in anxiety symptoms and exposure to community violence as well as improvement in standardized reading test scores. The intervention group also showed additional gains (e.g., increased math test scores, decreased life stress, and reduced victimization). Other adaptation examples illustrate the expansion of therapy content to address culturally relevant experiences related to the onset or maintenance of disorder. Martinez and Eddy (2005) developed and evaluated the efficacy of parent management training (PMT) for immigrant Latino families adapted to mitigate culturally specific risk factors for youth behavior problems and to mobilize culturally protective factors. In designing the adaptation, the team first trained Latino family therapists in the community in PMT, and together the community experts and investigators adapted the intervention to marshal content that was culturally relevant, informed by prevention science research, and likely to engage the Latino families. To ensure fidelity, they preserved core components associated with PMT efficacy in trials across settings and populations. In a randomized prevention trial of 73 Spanish-speaking Latino parents, adapted PMT produced benefits in parenting outcomes and youth behavior problems relative to no-treatment control. Taking a similar approach, Kohn, Oden, Muñoz, Robinson, and Leavitt (2002) adapted CBT to include both structural and content modifications to address the context of depression among African American women. Structural adaptations included changing some CBT terminology to promote acceptability and adding experiential meditative exercises and a treatment termination ritual. Content additions included relationship management skills, exploration of faith-based coping, and combating negative images of African American women. In a small matched-control evaluation study, the depressed African American women who received adapted CBT showed a two-fold decrease in depressive symptoms compared with the matched comparison sample who received standard CBT. Another approach to contextualizing therapy content is illustrated by Hinton et al. (2005), who translated data on descriptive psychopathology among trauma-exposed Cambodian American refugees into specific 40       lau et al.

modifications of CBT for posttraumatic stress disorder (PTSD) and panic. Hinton et al. (2006) identified neck-focused panic and orthostatically triggered panic as two culturally specific presentations in traumatized Cambodian refugees, then adapted CBT extensively to the symptoms and needs of survivors of the Khmer Rouge (Hinton et al., 2005). In an RCT of 40 treatment-resistant Cambodian Americans with comorbid PTSD and panic disorder, they found impressive improvements in symptoms (Cohen’s d = 2.17–3.78) in immediate treatment relative to delayed treatment. Following crossover, 55% of patients in this initially severe sample no longer met criteria for PTSD. Hinton, Hofmann, Rivera, Otto, and Pollack (2011) used similar procedures to adapt their CBT approach for treatment-resistant Latinos with PTSD. Cultural adaptations were undertaken to emphasize concepts easily understood by Latinos with little formal education, target prominent somatic complaints, and address common idioms of distress, including nervios and ataque de nervios. In addition, key CBT techniques were adapted to improve their acceptability and applicability to Latinos, including the use of Christian imagery and family-based analogies to illustrate concepts and culturally specific reinterpretations of somatic sensations during interoceptive exposure. In an RCT involving 24 treatment-resistant Latinas with PTSD, the culturally adapted CBT produced superior improvements in PTSD, anxiety, idioms of distress, and emotion regulation ability compared with a control group trained in applied muscle relaxation, with differences persisting at 12-week follow-up. The foregoing examples are A-EBTs with significant and substantive changes to therapy process and content to promote fit for specific ethnic groups. Although highly selected, these studies provide some preliminary evidence to answer two primary research questions shown in Figure 2.1. Question 4: Do A-EBTs work for ethnic minorities? Despite a range of effect sizes and small samples, all studies reviewed found evidence of efficacy relative to control. Moreover, investigators in each case show effects at least in the range of those found in RCTs of similar design with standard EBTs. Question 5: Do A-EBTs work better than standard EBTs for ethnic minorities? Only three small RCTs and one nonrandomized study included direct comparisons with the standard EBT. In one case, simple examination of group differences in pre- to postchange suggested an advantage of adaptation (Kohn et al., 2002). In two small RCTs evaluating the same treatment, analysis of variance indicated significant separation between A-EBT and EBT, favoring the adapted version on clinical outcomes (Huey & Pan, 2006; Pan et al., 2011). In the fourth study, statistical significance was not reached but results showed small to medium incremental effects of cultural adaptation even given strong efficacy in the standard EBT (McCabe & Yeh, 2009). These studies examined the effects of adaptation on clinical outcomes (i.e., symptom reduction); however, many psychotherapy outcome research     

41

A-EBTs operate to improve engagement (e.g., Breland-Noble & AAKOMA Project Adult Advisory Board, 2012), so their effects on treatment entry retention and adherence are paramount. In sum, there is encouraging evidence of the efficacy of A-EBTs with ethnic minorities; however, more studies with larger samples are required to evaluate whether adaptation results in superior outcomes in terms of engagement and efficacy. At this stage in the research, we are unaware of data bearing on Question 6: Do A-EBTs have generalized efficacy across ethnic groups? The results of four meta-analytic reviews have arrived at somewhat similar conclusions regarding the efficacy of A-EBTs. Griner and Smith (2006) conducted their meta-analysis using a very inclusive definition of A-EBTs for adults and youth (e.g., incorporation of cultural values into intervention content, client–therapist ethnic match, services in client’s primary language, provision of child care). Across the 76 studies reviewed, the average effect size was in the moderate range (Cohen’s d = 0.45), with effects being higher when therapists delivered treatment in the client’s (non-English) language. However, the analysis did not evaluate the relative advantage of A-EBTs versus EBTs. Huey and Polo’s (2008) meta-analysis compared effect sizes associated with treatments purported to include “culture-responsive elements” and with standard EBTs for minority youth and found no significant difference (Cohen’s d = 0.38). However, as in Griner and Smith (2006), the A-EBTs were not well-specified or superficial in nature and thus the researchers were unable to distinguish treatments on the basis of the content or quality of the cultural adaptation. Smith, Rodríguez, and Bernal (2011) performed a meta-analysis of 65 studies (experimental and quasi-experimental) of 8,620 patients. The studies were coded for the number and type of cultural adaptation conducted based on Bernal, Bonilla, and Bellido’s (1995) framework for cultural adaptations. The culturally adapted treatments were found to be moderately effective over traditional nonadapted treatments (Cohen’s d = 0.46). The study reported significantly better outcomes for treatments that employed a greater number of cultural adaptations. Also, treatments that focused on specific ethnocultural groups were more effective than were those produced for heterogeneous ethnically and culturally diverse target populations. A direct test comparing culturally adapted to conventional psychotherapies (Benish, Quintana, & Wampold, 2011) found culturally adapted psychotherapies superior to unadapted conventional treatments (Cohen’s d = 0.32) on primary outcomes. The meta-analysis involved 21 studies that met inclusion criteria and included 909 clients. In addition, Benish et al. (2011) examined the role of treatments addressing illness myth (i.e., the degree to which the therapeutic explanation offered to the client about his or her condition is congruent with the client’s culture and context) as a 42       lau et al.

moderator. These authors reported that illness myth was the only moderator of treatment effects (Cohen’s d = 0.21) among culturally adapted treatments. More recently, a fifth meta-analysis of cultural adaptations was conducted examining ethnic minority adult outpatients with depression and anxiety conditions (van Loon, van Schaik, Dekker, & Beekman, 2013). Nine studies met the inclusion and exclusion criteria of the review. The authors reported a large and significant effect size (Cohen’s d = 1.06) for culturally adapted depression and anxiety treatments. Two of the studies showed effectiveness for the population-specific cultural adaptations. Although the number of studies is relatively small, the findings provide additional evidence in support of population-specific cultural adaptations. Future meta-analysis should consider reducing heterogeneity by narrowing the diagnostic and developmental domains under study. In summary, four of the five meta-analyses (Benish et al., 2011; Griner & Smith, 2006; Smith et al., 2011; van Loon et al., 2013) found significant effects for cultural adaptation, with one study reporting a nonsignificant effect size (Huey & Polo, 2008). However, when the magnitude of the effect sizes is examined, it is clear that absolute efficacy of Cohen’s d ranges from 0.38 to 1.06 and relative efficacy from 0.32 to 0.40. Although more research in this area is needed, emerging evidence supports the incremental benefit of culturally adaptation in treatment outcomes. INNOVATION WITH CULTURALLY SENSITIVE THERAPIES By invoking the notion of tailoring treatments, Hall (2001) used the term CST, which parallels our view of A-EBTs. However, in our framework, CSTs may be distinguished from A-EBTs as interventions centered on healing traditions and cultural practices (e.g., alternative cultural healing philosophies, heritage folk practices) or designed to address group-specific protective factors or racial disparities not typically represented in the repertoire of extant EBTs. Hall (2001) argued that CSTs may be most valuable when fundamental conflicts arise between cultural values and the worldview within which EBTs have evolved. In this section, we describe four randomized trials of CSTs. An example of a CST emerging from a heritage cultural practice is cuento therapy derived from Puerto Rican culture and the use of cuentos or folktales to promote children’s healthy moral and emotional development (Costantino, Malgady, & Rogler, 1986). This group intervention involved teachers and parents reading, dramatizing, and discussing folktales—with adapted versions reflecting low-income urban settings in the United States— that illustrate child transgressions and moral lessons. In a randomized psychotherapy outcome research     

43

indicated prevention trial of 210 Puerto Rican children in New York, two versions of cuento therapy (using traditional or adapted cuentos) were compared with play therapy and a no-treatment control group. The cuento therapies significantly reduced children’s trait anxiety and aggression compared with play therapy and control with gains maintained after 1 year. The Strong African American Families (SAAF) program, aimed at rural African American families, was created to deter preadolescents’ engagement in HIV-related risk behaviors. The seven-session, group-based program was designed to foster regulated, communicative parenting behaviors, including consistent discipline, monitoring, and involvement; adaptive racial socialization strategies; general communication; and the establishment of clear expectations about sexual behavior and drug use. These parenting processes have been shown to have strong protective benefits for African American youth in particular (Brody, Murry, Kim, & Brown, 2002). A clustered randomized prevention trial involving 332 families (n = 150 in the SAAF intervention, n = 172 in the control condition) found that SAAF-induced increases in regulated, communicative parenting and youth self-pride mediated the effects of the intervention on the onset and escalation of risky sexual behaviors, even 65 months posttreatment. The SAAF-Teen (SAAF-T) program (Kogan et al., 2012) was designed to extend the protective benefits of African American parenting practices through the high school years, when risk behaviors tend to increase. Kogan et al. (2012) conducted a randomized controlled prevention trial of the SAAF-T program specifically designed to rule out nonspecific factors that could explain intervention effects. A total of 502 rural African American families with a 10th-grade student were randomly assigned to receive five sessions of either SAAF-T or a similarly structured attention controlled comparison intervention that focused on health and nutrition. Youth and families receiving SAAF-T reported greater increases in protective family management skills compared with the control group. In addition, the community engagement protocols resulted in 77.7% of families attending at least four of five intervention sessions, suggesting the transportability of SAAF-T into real-world settings. Gonzales et al. (2012) created the Bridges to High School Program/ Projecto Puentes a la Secundária (Bridges/Puentes), a family-focused prevention program for Mexican adolescents and their families. The 9-week group intervention combines adolescent coping, parenting, and family-strengthening intervention strategies to target ecological risk factors (e.g., exposure to drugs, deviant peers, poverty) as well as cultural risk factors (e.g., acculturation stress, diminished parenting effectiveness) for school disengagement and mental health problems during the transition to high school. A randomized trial involving 516 Mexican American adolescents found that Bridges/Puentes 44       lau et al.

was superior to a one-session control workshop in reducing adolescent substance abuse, internalizing and externalizing symptoms, and school discipline actions as well as improving school grades. Intervention effects were mediated by posttest changes in effective parenting, adolescent coping efficacy, adolescent school engagement, and family cohesion. Moderation analyses showed improved effects for families receiving the intervention in Spanish and for those with poorer baseline functioning. Thus, these four examples provide initial support for an affirmative response to Question 8: Are CSTs efficacious with minorities? However, data from treatment–treatment trials are not available to address Question 9: Do CSTs work as well as EBTs? or Question 10: What are the mechanisms of action in CSTs? However, in the case of interventions emerging from the prevention science research cycle, the absolute efficacy of culturally grounded interventions is often sufficient to make the case that, if they are implemented at the population level, there is the potential to reduce mental health disparities among high-risk ethnic minority communities. CONCLUDING REMARKS AND FRUITFUL DIRECTIONS FOR NEW RESEARCH The increasing maturity of research on cultural adaptation has led some to conclude that next steps should prioritize examinations of the relative efficacy of A-EBTs over standard EBTs in improving outcomes for ethnic minorities (e.g., Barrera, Castro, Strycker, & Toobert, 2013). Despite accumulating evidence of absolute efficacy of A-EBTs over control conditions, some argue that it is insufficient to justify the expense and effort of undertaking cultural adaptations. However, others have cautioned that research designs privileging questions of relative efficacy are misguided for several reasons. Methodologically, a number of concerns have been well articulated. First, trials to establish relative efficacy must be very large, and therefore very costly, given the need to show incremental gains over a well-established EBT. With expected gains in effect size in the 0.30 to 0.50 range, required sample sizes from what may be small clinical populations within specific ethnocultural groups may render such power-intensive trials logistically untenable (Domenech Rodríguez & Bernal, 2012). Second, Cardemil (2010) noted that investigators generally need to take considerable care to engage ethnic minorities to field successful controlled trials of EBTs. Any such engagement practices put in place to enable the trial would likely obscure differences between the standard EBT arm and the A-EBT. Third, Domenech Rodríguez and Bernal (2012) argued that most head-to-head comparisons of EBTs versus A-EBTs conducted by cultural adaptation researchers would psychotherapy outcome research     

45

violate the principle of equipoise, wherein the investigators maintain a position of uncertainty about which condition will render superior outcomes. A related concern is that it may be unethical for controlled trial researchers to withhold elements of culturally responsive care in the standard EBT arm in relative efficacy trials. These considerations render it difficult to conduct tests of relative efficacy, and resulting tests are likely to be conservative in nature. Relative efficacy trials directly comparing an A-EBT with a standard EBT must recruit large samples, maintain assessors blind to condition (or the overall design), and implement fidelity assessment that guards against contamination by monitoring a standard EBT arm to assess for spontaneous tailoring for cultural sensitivity. It is laudable that researchers have managed to conduct trials with this level of rigor and have demonstrated relative efficacy of A-EBTs (as reviewed in Question 5 above). However, the most fruitful next steps in treatment outcome research with ethnic minorities do not necessarily involve questioning A-EBT superiority in efficacy. As our review suggests, cultural adaptations range in form and function as well as in intended and likely impact. Some adaptations (e.g., translation of materials) are essential in extending the reach of EBTs to underserved groups but do not change the nature of the intervention per se; some (e.g., using images of ethnically similar faces) are relatively superficial in nature, intended perhaps to promote a sense of inclusion; and some heighten the relevance of the intervention content to the target population. Furthermore, some adaptations encompass augmented content to address culturally salient risk factors for a particular disorder (e.g., acculturation conflicts) or leverage culturally protective factors (e.g., emotion restraint values). Further still, other adaptations alter therapy process or frame interventions in culturally congruent ways to promote engagement (e.g., rationale for parent training). Undergirding these adaptations should be a theory of change that drives the purported effects of the adaptations, yet these theories are often not well articulated or addressed in clinical trials. A further limitation of the state of the science is that broad racial/ ethnic communities are the target of focus for A-EBTs and CSTs. There is enormous demographic and psychosocial heterogeneity within groups such as Latinos, African Americans, Native Americans, and Asian Americans that limits the value of this unit of analysis for the purpose of tailoring care. For example, the Asian American/Pacific Islander classification includes more than 50 distinct ethnic groups with more than 30 different languages spoken (Iwamasa, 2003), with great variations across and within each subethnic group with respect to their history of immigration and settlement, social capital, class, pre- and postmigration trauma and adaptation, transnational family formations, documentation status, and so on. This heterogeneity confers a great deal of complexity in mounting interventions designed to be responsive 46       lau et al.

to whole groups and calls into question a group-focused emphasis. It may be more fruitful to identify cultural dimensions relevant to the tailoring of interventions that may apply across groups. Indeed, to best inform intervention science, future studies should test theory-driven cultural adaptations and aim to understand mechanisms of action that enhance engagement or symptom outcomes. This approach is consistent with priorities outlined by the National Advisory Mental Health Council Workgroup (2010), which indicated that research must target adaptations that modify factors shown to be associated with EBT nonresponse, partial response, or patient nonengagement. Such work should be premised on a well-specified mechanism by which such moderator variables function to disadvantage a cultural group in an EBT. For example, minority status experiences (e.g., overt and covert discrimination, microaggressions) are relevant across racial minority groups as well as sexual minority and other disadvantaged groups. This risk factor may be the target of intervention adaptation across communities. Furthermore, immigrant groups across ethnicities share experiences such as acculturation-related stress and intergenerational relationship strain and may also be the focus of research questions about patient nonresponse to standard EBTs. Some documented adaptations have addressed these common elements of the experiences of ethnic minorities and may thus provide the basis for examining what dimensions of diversity are crucial in designing adaptations. Accordingly, the design of controlled trials must attend to the question of what are the purported mechanisms of change for ethnic minorities in EBTs, A-EBTs, and CSTs. Thus, the most promising advances will integrate inquiry into the heretofore neglected Questions 3, 4, and 10.

REFERENCES Arnold, L. E., Elliot, M., Sachs, L., Bird, H., Kraemer, H. C., Wells, K. C., . . . Wigal, T. (2003). Effects of ethnicity on treatment attendance, stimulant response/ dose, and 14-month outcome in ADHD. Journal of Consulting and Clinical Psychology, 71, 713–727. http://dx.doi.org/10.1037/0022-006X.71.4.713 Barrera, M., Jr., Castro, F. G., Strycker, L. A., & Toobert, D. J. (2013). Cultural adaptations of behavioral health interventions: A progress report. Journal of Consulting and Clinical Psychology, 81, 196–205. http://dx.doi.org/10.1037/a0027085 Barrett, P., & Turner, C. (2001). Prevention of anxiety symptoms in primary school children: Preliminary results from a universal school-based trial. British Journal of Clinical Psychology, 40, 399–410. http://dx.doi.org/10.1348/014466501163887 Benish, S. G., Quintana, S., & Wampold, B. E. (2011). Culturally adapted psychotherapy and the legitimacy of myth: A direct-comparison meta-analysis. Journal of Counseling Psychology, 58, 279–289. http://dx.doi.org/10.1037/a0023626 psychotherapy outcome research     

47

Bernal, G., Bonilla, J., & Bellido, C. (1995). Ecological validity and cultural sensitivity for outcome research: Issues for the cultural adaptation and development of psychosocial treatments with Hispanics. Journal of Abnormal Child Psychology, 23, 67–82. Bernal, G., & Scharrón-del-Río, M. R. (2001). Are empirically supported treatments valid for ethnic minorities? Toward an alternative approach for treatment research. Cultural Diversity and Ethnic Minority Psychology, 7, 328–342. Breland-Noble, A. M., & AAKOMA Project Adult Advisory Board. (2012). Community and treatment engagement for depressed African American youth: The AAKOMA FLOA pilot. Journal of Clinical Psychology in Medical Settings, 19, 41–48. http://dx.doi.org/10.1007/s10880-011-9281-0 Brody, G. H., Murry, V. M., Kim, S., & Brown, A. C. (2002). Longitudinal pathways to competence and psychological adjustment among African American children living in rural single-parent households. Child Development, 73, 1505–1516. http:// dx.doi.org/10.1111/1467-8624.00486 Brown, C., Schulberg, H. C., Sacco, D., Perel, J. M., & Houck, P. R. (1999). Effectiveness of treatments for major depression in primary medical care practice: A post hoc analysis of outcomes for African American and White patients. Journal of Affective Disorders, 53, 185–192. http://dx.doi.org/10.1016/S01650327(98)00120-7 Cardemil, E. V. (2010). Cultural adaptations to empirically supported treatment: A research agenda. The Scientific Review of Mental Health Practice, 7, 8–21. Cardemil, E. V., Reivich, K. J., & Seligman, M. E. P. (2002). The prevention of depressive symptoms in low-income minority middle school students. Prevention & Treatment, 5, 1–29. http://dx.doi.org/10.1037/1522-3736.5.1.58a Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical Psychology, 66, 7–18. http://dx.doi.org/10.1037/ 0022-006X.66.1.7 Chambless, D. L., Sanderson, W. C., Shoham, V., Bennett Johnson, S., Pope, K. S., Crits-Christoph, P., . . . McCurry, S. (1996). An update on empirically validated therapies. Clinical Psychologist, 29, 5–18. Chambless, D. L., & Williams, K. E. (1995). A preliminary study of African Americans with agoraphobia: Symptom severity and outcome of treatment with in vivo exposure. Behavior Therapy, 26, 501–515. http://dx.doi.org/10.1016/ S0005-7894(05)80097-8 Choi, I., Zou, J., Titov, N., Dear, B. F., Li, S., Johnston, L., . . . Hunt, C. (2012). Culturally attuned Internet treatment for depression amongst Chinese Australians: A randomised controlled trial. Journal of Affective Disorders, 136, 459–468. http://dx.doi.org/10.1016/j.jad.2011.11.003 Cohen, J. A., Deblinger, E., Mannarino, A. P., & Steer, R. A. (2004). A multisite, randomized controlled trial for children with sexual abuse-related PTSD symptoms. Journal of the American Academy of Child & Adolescent Psychiatry, 43, 393–402. http://dx.doi.org/10.1097/00004583-200404000-00005

48       lau et al.

Cooley-Strickland, M. R., Griffin, R. S., Darney, D., Otte, K., & Ko, J. (2011). Urban African American youth exposed to community violence: A schoolbased anxiety preventive intervention efficacy study. Journal of Prevention & Intervention in the Community, 39, 149–166. Costantino, G., Malgady, R. G., & Rogler, L. H. (1986). Cuento therapy: A culturally sensitive modality for Puerto Rican children. Journal of Consulting and Clinical Psychology, 54, 639–645. http://dx.doi.org/10.1037/0022006X.54.5.639 D’Angelo, E. J., Llerena-Quinn, R., Shapiro, R., Colon, F., Rodriguez, P., Gallagher, K., & Beardslee, W. R. (2009). Adaptation of the preventive intervention program for depression for use with predominantly low-income Latino families. Family Process, 48, 269–291. http://dx.doi.org/10.1111/j.1545-5300.2009.01281.x Domenech Rodríguez, M. M., & Bernal, G. (2012). Translating models of research in empirically based practice. In G. Bernal & M. M. Domenech Rodríguez (Eds.), Cultural adaptations: Tools for evidence-based practice with diverse populations (pp. 265–289). Washington, DC: American Psychological Association. Eap, S., & Hall, G. C. N. (2007). Relevance of RCTs to diverse groups. In A. M. Nezu & C. M. Nezu (Eds.), Evidence-based outcome research: A practical guide to conducting randomized controlled trials for psychosocial interventions (pp. 425–444). http://dx.doi.org/10.1093/med:psych/9780195304633.003.0020 Ginsburg, G. S., Becker, K. D., Drazdowski, T. K., & Tein, J.-Y. (2012). Treating anxiety disorders in inner city schools: Results from a pilot randomized controlled trial comparing CBT and usual care. Child & Youth Care Forum, 41, 1–19. http://dx.doi.org/10.1007/s10566-011-9156-4 Ginsburg, G. S., & Drake, K. L. (2002). School-based treatment for anxious AfricanAmerican adolescents: A controlled pilot study. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 768–775. http://dx.doi.org/10.1097/ 00004583-200207000-00007 Gone, J. P., & Alcántara, C. (2007). Identifying effective mental health interventions for American Indians and Alaska Natives: A review of the literature. Cultural Diversity and Ethnic Minority Psychology, 13, 356–363. http://dx.doi. org/10.1037/1099-9809.13.4.356 Gonzales, N. A., Dumka, L. E., Millsap, R. E., Gottschall, A., McClain, D. B., Wong, J. J., . . . Kim, S. Y. (2012). Randomized trial of a broad preventive intervention for Mexican American adolescents. Journal of Consulting and Clinical Psychology, 80, 1–16. http://dx.doi.org/10.1037/a0026063 Griner, D., & Smith, T. B. (2006). Culturally adapted mental health intervention: A meta-analytic review. Psychotherapy: Theory, Research, Practice, Training, 43, 531–548. http://dx.doi.org/10.1037/0033-3204.43.4.531 Hall, G. C. N. (2001). Psychotherapy research with ethnic minorities: Empirical, ethical, and conceptual issues. Journal of Consulting and Clinical Psychology, 69, 502–510. http://dx.doi.org/10.1037/0022-006X.69.3.502 psychotherapy outcome research     

49

Henggeler, S. W., Melton, G. B., & Smith, L. A. (1992). Family preservation using multisystemic therapy: An effective alternative to incarcerating serious juvenile offenders. Journal of Consulting and Clinical Psychology, 60, 953–961. http:// dx.doi.org/10.1037/0022-006X.60.6.953 Hinton, D. E., Chhean, D., Pich, V., Safren, S. A., Hofmann, S. G., & Pollack, M. H. (2005). A randomized controlled trial of cognitive-behavior therapy for Cambodian refugees with treatment-resistant PTSD and panic attacks: A crossover design. Journal of Traumatic Stress, 18, 617–629. http://dx.doi.org/10.1002/ jts.20070 Hinton, D. E., Chhean, D., Pich, V., Um, K., Fama, J. M., & Pollack, M. H. (2006). Neck-focused panic attacks among Cambodian refugees; a logistic and linear regression analysis. Journal of Anxiety Disorders, 20, 119–138. http://dx.doi. org/10.1016/j.janxdis.2005.02.001 Hinton, D. E., Hofmann, S. G., Rivera, E., Otto, M. W., & Pollack, M. H. (2011). Culturally adapted CBT (CA-CBT) for Latino women with treatmentresistant PTSD: A pilot study comparing CA-CBT to applied muscle relaxation. Behaviour Research and Therapy, 49, 275–280. http://dx.doi.org/10.1016/ j.brat.2011.01.005 Hohmann, A. A., & Parron, D. L. (1996). How the new NIH Guidelines on Inclusion of Women and Minorities apply: Efficacy trials, effectiveness trials, and validity. Journal of Consulting and Clinical Psychology, 64, 851–855. Huey, S. J., Jr., & Pan, D. (2006). Culture-responsive one-session treatment for phobic Asian Americans: A pilot study. Psychotherapy: Theory, Research, Practice, Training, 43, 549–554. http://dx.doi.org/10.1037/0033-3204.43.4.549 Huey, S. J., Jr., & Polo, A. J. (2008). Evidence-based psychosocial treatments for ethnic minority youth. Journal of Clinical Child and Adolescent Psychology, 37, 262–301. http://dx.doi.org/10.1080/15374410701820174 Iwamasa, G. Y. (2003, November). Recommendations for the treatment of Asian American/Pacific Islander populations. In Psychological treatment of ethnic minority populations (pp. 8–12) [Brochure]. Washington, DC: Association of Black Psychologists. Kogan, S. M., Brody, G. H., Molgaard, V. K., Grange, C. M., Oliver, D. A. H., Anderson, T. N., . . . Sperr, M. C. (2012). The Strong African American Families–Teen Trial: Rationale, design, engagement processes, and familyspecific effects. Prevention Science, 13, 206–217. http://dx.doi.org/10.1007/ s11121-011-0257-y Kohn, L. P., Oden, T., Muñoz, R. F., Robinson, A., & Leavitt, D. (2002). Adapted cognitive behavioral group therapy for depressed low-income African American women. Community Mental Health Journal, 38, 497–504. http://dx.doi. org/10.1023/A:1020884202677 Kraemer, H. C., Frank, E., & Kupfer, D. J. (2006). Moderators of treatment outcomes: Clinical, research, and policy importance. JAMA, 296, 1286–1289. http:// dx.doi.org/10.1001/jama.296.10.1286

50       lau et al.

Lau, A. S. (2006). Making the case for selective and directed cultural adaptations of evidence-based treatments: Examples from parent training. Clinical Psychology: Science and Practice, 13, 295–310. http://dx.doi.org/10.1111/j.1468-2850.2006.00042.x Liddle, H. A., Rowe, C. L., Dakof, G. A., Ungaro, R. A., & Henderson, C. E. (2004). Early intervention for adolescent substance abuse: Pretreatment to posttreatment outcomes of a randomized clinical trial comparing multidimensional family therapy and peer group treatment. Journal of Psychoactive Drugs, 36, 49–63. http://dx.doi.org/10.1080/02791072.2004.10399723 Lochman, J. E., & Wells, K. C. (2004). The coping power program for preadolescent aggressive boys and their parents: Outcome effects at the 1-year follow-up. Journal of Consulting and Clinical Psychology, 72, 571–578. Mak, W. W. S., Law, R. W., Alvidrez, J., & Pérez-Stable, E. J. (2007). Gender and ethnic diversity in NIMH-funded clinical trials: Review of a decade of published research. Administration and Policy in Mental Health and Mental Health Services Research, 34, 497–503. http://dx.doi.org/10.1007/s10488-007-0133-z Martinez, C. R., Jr., & Eddy, J. M. (2005). Effects of culturally adapted parent management training on Latino youth behavioral health outcomes. Journal of Consulting and Clinical Psychology, 73, 841–851. http://dx.doi.org/10.1037/ 0022-006X.73.5.841 Matos, M., Bauermeister, J. J., & Bernal, G. (2009). Parent–child interaction therapy for Puerto Rican preschool children with ADHD and behavior problems: A pilot efficacy study. Family Process, 48, 232–252. http://dx.doi.org/10.1111/ j.1545-5300.2009.01279.x Matos, M., Torres, R., Santiago, R., Jurado, M., & Rodríguez, I. (2006). Adaptation of parent–child interaction therapy for Puerto Rican families: A preliminary study. Family Process, 45, 205–222. http://dx.doi.org/10.1111/j.1545-5300.2006.00091.x McCabe, K. M., & Yeh, M. (2009). Parent–child interaction therapy for Mexican Americans: A randomized clinical trial. Journal of Clinical Child and Adolescent Psychology, 38, 753–759. http://dx.doi.org/10.1080/15374410903103544 McCabe, K. M., Yeh, M., Garland, A. F., Lau, A. S., & Chavez, G. (2005). The GANA Program: A tailoring approach to adapting parent child interaction therapy for Mexican Americans. Education & Treatment of Children, 28, 111–129. Miranda, J., Bernal, G., Lau, A., Kohn, L., Hwang, W. C., & LaFromboise, T. (2005). State of the science on psychosocial interventions for ethnic minorities. Annual Review of Clinical Psychology, 1, 113–142. http://dx.doi.org/10.1146/annurev. clinpsy.1.102803.143822 Miranda, J., Chung, J. Y., Green, B. L., Krupnick, J., Siddique, J., Revicki, D. A., & Belin, T. (2003). Treating depression in predominantly low-income young minority women: A randomized controlled trial. JAMA, 290, 57–65. http:// dx.doi.org/10.1001/jama.290.1.57 Mufson, L., Weissman, M. M., Moreau, D., & Garfinkel, R. (1999). Efficacy of interpersonal psychotherapy for depressed adolescents. Archives of General Psychiatry, 56, 573–579. http://dx.doi.org/10.1001/archpsyc.56.6.573 psychotherapy outcome research     

51

Muñoz, R. F., & Mendelson, T. (2005). Toward evidence-based interventions for diverse populations: The San Francisco General Hospital prevention and treatment manuals. Journal of Consulting and Clinical Psychology, 73, 790–799. http:// dx.doi.org/10.1037/0022-006X.73.5.790 Nathan, P. E., & Gorman, J. M. (1998). Treatments that work—And what convinces us that they do. In P. E. Nathan & J. M. Gorman (Eds.), A guide to treatments that work (pp. 3–25). New York, NY: Oxford. National Advisory Mental Health Council Workgroup. (2010). From discovery to cure: Accelerating the development of new and personalized interventions for mental illness. Retrieved from http://www.nimh.nih.gov/about/advisory-boards-andgroups/namhc/reports/fromdiscoverytocure_103739.pdf Pan, D., Huey, S. J., Jr., & Hernandez, D. (2011). Culturally adapted versus standard exposure treatment for phobic Asian Americans: Treatment efficacy, moderators, and predictors. Cultural Diversity and Ethnic Minority Psychology, 17, 11–22. http://dx.doi.org/10.1037/a0022534 Parra Cardona, J. R., Domenech-Rodriguez, M., Forgatch, M., Sullivan, C., Bybee, D., Holtrop, K., . . . Bernal, G. (2012). Culturally adapting an evidence-based parenting intervention for Latino immigrants: The need to integrate fidelity and cultural relevance. Family Process, 51, 56–72. http://dx.doi.org/10.1111/ j.1545-5300.2012.01386.x Perini, S., Titov, N., & Andrews, G. (2009). Clinician-assisted Internet-based treatment is effective for depression: Randomized controlled trial. Australian & New Zealand Journal of Psychiatry, 43, 571–578. http://dx.doi.org/10.1080/00048670902873722 Reid, M. J., Webster-Stratton, C., & Beauchaine, T. P. (2001). Parent training in Head Start: A comparison of program response among African American, Asian American, Caucasian, and Hispanic mothers. Prevention Science, 2, 209–227. http://dx.doi.org/10.1023/A:1013618309070 Robbins, M. S., Mayorga, C. C., Mitrani, V. B., Szapocznik, J., Turner, C. W., & Alexander, J. F. (2008). Adolescent and parent alliances with therapists in Brief Strategic Family Therapy with drug-using Hispanic adolescents. Journal of Marital and Family Therapy, 34, 316–328. http://dx.doi.org/10.1111/ j.1752-0606.2008.00075.x Rosselló, J., & Bernal, G. (1999). The efficacy of cognitive–behavioral and interpersonal treatments for depression in Puerto Rican adolescents. Journal of Consulting and Clinical Psychology, 67, 734–745. http://dx.doi.org/10.1037/ 0022-006X.67.5.734 Rosselló, J., Bernal, G., & Rivera-Medina, C. (2008). Individual and group CBT and IPT for Puerto Rican adolescents with depressive symptoms. Cultural Diversity and Ethnic Minority Psychology, 14, 234–245. http://dx.doi.org/10.1037/ 1099-9809.14.3.234 Santisteban, D. A., Coatsworth, J. D., Perez-Vidal, A., Kurtines, W. M., Schwartz, S. J., LaPerriere, A., & Szapocznik, J. (2003). Efficacy of brief strategic family therapy in modifying Hispanic adolescent behavior problems and substance

52       lau et al.

use. Journal of Family Psychology, 17, 121–133. http://dx.doi.org/10.1037/ 0893-3200.17.1.121 Shetgiri, R., Kataoka, S., Lin, H., & Flores, G. (2011). A randomized, controlled trial of a school-based intervention to reduce violence and substance use in predominantly Latino high school students. Journal of the National Medical Association, 103, 932–940. Silverman, W. K., Kurtines, W. M., Ginsburg, G. S., Weems, C. F., Lumpkin, P. W., & Carmichael, D. H. (1999). Treating anxiety disorders in children with group cognitive–behavioral therapy: A randomized clinical trial. Journal of Consulting and Clinical Psychology, 67, 995–1003. http://dx.doi.org/10.1037/ 0022-006X.67.6.995 Smith, T. B., Rodríguez, M. D., & Bernal, G. (2011). Culture. Journal of Clinical Psychology, 67, 166–175. http://dx.doi.org/10.1002/jclp.20757 Stein, B. D., Jaycox, L. H., Kataoka, S. H., Wong, M., Tu, W., Elliott, M. N., & Fink, A. (2003). A mental health intervention for schoolchildren exposed to violence: A randomized controlled trial. JAMA, 290, 603–611. http://dx.doi. org/10.1001/jama.290.5.603 Sue, S., & Zane, N. (2006). How well do both evidence-based practices and treatment as usual satisfactorily address the various dimensions of diversity? In J. C. Norcross, L. E. Beutler, & R. F. Levant (Eds.), Evidence-based practices in mental health: Debate and dialogue on the fundamental questions (pp. 329–374). Washington, DC: American Psychological Association. Task Force on Promotion and Dissemination of Psychological Procedures. (1995). Training in and dissemination of empirically-validated psychological treatments: Report and recommendations. Clinical Psychologist, 48, 3–23. van Loon, A., van Schaik, A., Dekker, J., & Beekman, A. (2013). Bridging the gap for ethnic minority adult outpatients with depression and anxiety disorders by culturally adapted treatments. Journal of Affective Disorders, 147(1–3), 9–16. Wells, K., Miranda, J., Bruce, M. L., Alegría, M., & Wallerstein, N. (2004). Bridging community intervention and mental health services research. The American Journal of Psychiatry, 161, 955–963. http://dx.doi.org/10.1176/appi.ajp.161.6.955 Zoellner, L. A., Feeny, N. C., Fitzgibbons, L. A., & Foa, E. B. (1999). Response of African American and Caucasian women to cognitive behavioral therapy for PTSD. Behavior Therapy, 30, 581–595. http://dx.doi.org/10.1016/ S0005-7894(99)80026-4

psychotherapy outcome research     

53

II Measurement and Statistical Issues

3 THREATS TO CULTURAL VALIDITY IN CLINICAL DIAGNOSIS AND ASSESSMENT: ILLUSTRATED WITH THE CASE OF ASIAN AMERICANS FREDERICK T. L. LEONG AND ZORNITSA KALIBATSEVA

Clinical diagnosis and assessment are key aspects of psychotherapy. This chapter presents a conceptual model for evaluating threats to cultural validity in clinical diagnosis and assessment among racial and ethnic minorities and Asian Americans, in particular. Culturally informed evidence-based practice requires accurate diagnosis and assessment of racial and ethnic minority clients. However, our review shows that there are challenges and threats to arriving at such culturally sensitive and appropriate diagnosis and assessment of culturally diverse patient populations. The delineation of these threats to cultural validity in clinical diagnosis and assessment has both clinical and research implications with regard to evidence-based practice in psychology. Virtually every school of psychotherapy considers assessment and diagnosis important initial steps in the therapy process. Indeed, it has been argued that clinical diagnosis is of great importance because appropriate treatment may depend on a correct diagnosis (Garfield, 1984). The diagnosis of http://dx.doi.org/10.1037/14940-004 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

57

mental disorders has four major goals (Guindon & Sobhany, 2001). First, by giving someone a diagnosis, the mental health professional attempts to identify the problems. Second, the professional tries to recognize the factors contributing to and maintaining the identified problems. Third, when a diagnosis is present, the professional can choose and carry out the most appropriate course of treatment. Fourth, the professional can change the treatment to meet the needs of the client if necessary. Thus, giving an accurate diagnosis is of utmost importance for the correct identification of the problems and the best choice of intervention. However, a number of significant problems have been associated with the clinical diagnosis of psychopathology and the value of the diagnostic process has remained controversial (Garfield, 1984; Sadler, 2005). CONCEPT OF CULTURAL VALIDITY An essential problem in assessment and clinical diagnosis has been the neglect of the role of cultural differences on psychopathology. One way to conceptualize this problem is to use Tyler, Sussewell, and Williams-McCoy’s (1985) ethnic validity model in psychotherapy. According to Tyler et al., the ethnic validity model refers to the “recognition, acceptance, and respect for the presence of communalities and differences in psychosocial development and experiences among people with different ethnic or cultural heritages” (p. 312). We propose that instead of using the concept of ethnic validity, it may be more useful to broaden the concept to that of cultural validity. The issue of cultural validity in assessment and clinical diagnosis would then include an exploration of inaccuracies in evaluation and diagnosis due to differences in race, ethnicity, nationality, or culture. Similar to the concept of external validity (Bracht & Glass, 1968), cultural validity is arguably an important corollary to psychometric validities (e.g., face, construct, predictive, and concurrent). The concept of cultural validity refers to the effectiveness of a measure or the accuracy of a clinical diagnosis to address the existence and importance of essential cultural factors. Such cultural factors may include values, beliefs, experiences, communication patterns, and epistemologies inherent to the clients’ cultural backgrounds (Solano-Flores & Nelson-Barber, 2001). The problem of cultural validity is not merely an academic issue because the lack of cultural validity in clinical diagnosis could result in two major sets of problems. Clinically, the lack of cultural validity may result in an incorrect diagnosis and ineffective treatment of culturally different populations. Socially, such individuals may be unnecessarily stigmatized and institutionalized as a result of diagnostic errors. 58       leong and kalibatseva

THREATS TO CULTURAL VALIDITY The interpretation of assessment data, the accuracy of clinical diagnosis, and the outcome of psychotherapy with culturally diverse populations can be influenced by many factors. However, a universalist perspective has consistently been used in assessment and diagnosis, which assumes that all people, regardless of race, ethnicity, or culture, develop along uniform psychological dimensions (Canino & Alegría, 2008; Malgady, 1996). The assumption of uniformity across racial and ethnic groups has prevailed, and one of the arguments in this chapter is that this cultural uniformity assumption prevents clinicians from recognizing cultural differences that may affect the assessment and diagnosis of culturally diverse clients. Several major factors may contribute to the lack of cultural validity in clinical diagnosis. Borrowing from Campbell and Stanley’s (1966) concept of threats to validity, the lack of cultural validity in clinical assessment and diagnosis can be conceptualized in terms of multiple threats to validity. These threats to cultural validity in clinical assessment are largely due to a failure to recognize or a tendency to minimize cultural factors in clinical assessment and diagnosis. Our review of the literature suggests that several factors may serve as the sources of threats to cultural validity. These factors include but are not limited to (a) pathoplasticity of psychological disorders, (b) cultural factors influencing symptom expression, (c) therapist bias in clinical judgment, (d) language capability of the client, and (e) inappropriate use of clinical and personality tests. A series of articles and chapters has reviewed cross-cultural problems in assessment and clinical diagnosis with various cultural groups (e.g., Adebimpe, 2004; Cheung, Leong, & Ben-Porath, 2003; Eap, Gobin, Ng, & Hall, 2010; Malgady, 1996; Westermeyer, 1985). This chapter illustrates each of the listed sources of threats to cultural validity in clinical assessment and diagnosis with an emphasis on the literature on Asian Americans. PATHOPLASTICITY OF PSYCHOLOGICAL DISORDERS The concept of pathoplasticity of psychological disorders refers to the variability in symptoms, course, outcome, and distribution of mental disorders among various cultural groups (Westermeyer, 1985). Westermeyer (1985) provided three examples of pathoplasticity. First, features associated with schizophrenia may vary widely from one culture to another and even among ethnic groups in a single country. Such differences may involve the content, severity, or relative frequency of symptoms, such as withdrawal, volubility, agitation, compliance, and paranoia. Second, the pathoplasticity of nonpsychotic disorders may be observed in the different rates of mood, threats to cultural validity     

59

anxiety, and somatoform disorders, which differ from one culture to another. A third example of pathoplasticity is the better outcome of schizophrenia in developing countries than in developed countries, which has been attributed to sociocultural factors (e.g., more stable social networks, integration in the community). Fourth, it is possible that certain psychopathological states may be represented by different diagnoses in different cultures. For instance, Kleinman (1982) found that 87% of Chinese outpatients diagnosed with neurasthenia met criteria for major depressive disorder and suggested that somatic symptoms may mask depression. The pathoplasticity of psychological disorders, therefore, serves as a major threat to cultural validity in clinical diagnosis resulting from a failure to recognize the cultural plasticity often associated with various forms of psychopathology. Different studies have demonstrated this pathoplasticity of psychological disorders among Asian Americans. These studies have generally fallen into one of two categories: those involving severity of disorders and those pertaining to distribution of mental disorders. Studies of mental disorders among Chinese and Japanese Americans have consistently found that Asian Americans reported a greater severity of disturbance when compared with Whites (Chu & Sue, 2011; Leong, 1986). On the basis of Diagnostic and Statistical Manual of Mental Disorders (2nd ed.; DSM–II; American Psychiatric Association, 1968) diagnoses, S. Sue and McKinney (1975) found that almost a quarter of the Asian American clients (22%) in community mental health centers received diagnoses of psychoses compared with 13% of White clients. This difference remained even after statistically controlling for age and education. Other studies using objective psychological tests have also found that Asian Americans exhibit more severe symptoms. In a study using records from a university clinic, S. Sue and Sue (1974) compared the Minnesota Multiphasic Personality Inventory (MMPI; Hathaway & McKinley, 1940) profiles of Asian American patients with those of other patient groups. They found that Chinese and Japanese American patients scored significantly higher on several of the MMPI scales. In addition, the investigators compared the two sets of profiles using the Goldberg formula and found that the Asian patients had more psychotic profiles. Asian males scored significantly higher than did the controls on the following clinical scales: Hypochondriasis, Depression, Psychopathic Deviate, Paranoia, Psychasthenia, and Schizophrenia. Asian females, on the other hand, scored higher on the Paranoia, Psychasthenia, Schizophrenia, and Social Introversion scales. Durvasula and Sue (1996) examined a large sample of Asian American and White clients who utilized mental health services over a 5-year period and compared the two groups on severity of diagnosis, ratings of functioning, and presence of psychotic features. The Asian American clients had a higher 60       leong and kalibatseva

proportion of severe diagnoses, lower functioning scores, and a higher proportion of psychotic features among those with mood disorders. Thus, Asian Americans may show a higher severity of disturbance not only in serious mental disorders (e.g., schizophrenia) but also in substance abuse and mood disorders. Although the pattern of greater severity is consistent, there are two limitations to these studies. The majority of these studies included participants who sought mental health services and received a clinical diagnosis; however, there may have been cultural bias and stereotyping in the diagnostic process. In addition, quite a few of these studies were conducted in Chinatowns or urban ethnic enclaves, which are more likely to be composed of lower class and less educated members. Another example of pathoplasticity among Asian Americans is the distribution of mental disorders identified in treated case studies. For example, S. Sue (1977) compared the rates of various disorders among Asian Americans, Whites, and other racial/ethnic groups. The author found that Asian Americans had the highest percentage of psychosis (22.4%), as compared with Whites (12.7%), Blacks (13.8%), Chicanos (14.5%), and Native Americans (17.6%). Moreover, there is evidence that the distribution of mental disorders varies within ethnic subgroups. Jackson et al. (2011) examined the prevalence rates of major depressive episode (MDE) among the various racial/ ethnic groups in the Collaborative Psychiatric Epidemiological Surveys. The National Latino and Asian American Study sampled three Asian ethnic groups—Filipinos, Vietnamese, and Chinese—and a fourth group included “other Asian.” According to Jackson et al., the Asian ethnic groups reported the lowest rates of lifetime MDE compared with all others (non-Latino Whites, Hispanics, Caribbean Blacks, and African Americans) and Filipinos were the ethnic group with the lowest rate (7.2%). In addition, Jackson et al. compared the prevalence rates of MDE for U.S. born and non-U.S. born participants and consistently found that among the Asian ethnic groups, nonU.S. born participants reported lower prevalence rates. Specifically, the MDE prevalence rate for U.S.-born Chinese Americans was 21.5% as opposed to 7.7% for the non-U.S.-born Chinese Americans. Therefore, it is important to examine the interactions of culture, race, ethnicity, and immigration in the assessment of individuals from diverse populations. In summary, two major examples of the pathoplasticity of psychiatric disorders among Asian Americans are the severity of psychiatric symptoms and the distribution and rates of mental disorders. Severity of symptoms among Asian Americans remains an unresolved issue because studies have not yet differentiated between various causal factors. Three possible causes for the observed higher severity are higher rates of mental health problems among Asian Americans, underutilization of mental health services, or misdiagnosis as a result of miscommunication or lack of cultural knowledge. threats to cultural validity     

61

CULTURAL FACTORS INFLUENCING SYMPTOM EXPRESSION Another threat to cultural validity in clinical diagnosis is the influence of the clients’ cultural background on their symptom expression. To see how ethnic origin may influence symptom expression, Enright and Jaeckle (1963) examined the behavioral patterns of Japanese and Filipino patients with schizophrenia at Hawaii State Hospital and found that Japanese patients expressed more depression, withdrawal, and disturbance in thinking. Filipino patients exhibited more overt disturbance of behavior and had more delusions of persecution. In another study of how culture may influence the manifestation of symptoms among Asian Americans, Katz, Sanborn, and Gudeman (1969) provided an example of intergroup differences. They studied the descriptions of Japanese and White acute schizophrenia patients before hospital admission and during their stay in the hospital. Before hospitalization, the Japanese were described as more socially obstreperous, nervous, and hyperactive than the Whites. Once they were in the hospital, their behaviors changed. The Whites were rated as more excited and disturbed than the Japanese. Therefore, in making diagnoses of Asian Americans, there may be a need to sample their behaviors broadly rather than base the diagnosis on observations and information obtained only in the diagnostic interview because cultural background may mediate when and how symptoms are expressed. A more recent case study illustrates the continual problem of culture-related mis­ diagnosis (Alberque & Eytan, 2001). One of the most common claims in cross-cultural psychopathology has been that Asian Americans tend to somatize distress (Parker, Cheah, & Roy, 2001; Yang & WonPat-Borja, 2007). Pang’s (1990) study on Hwabyung provided an interesting example of the tendency to somatize psychological symptoms among Asian Americans. Because of the Korean culture’s esteem of restraint, suppression of verbal aggression, and avoidance of confrontation, Hwabyung is a uniquely Korean culture-bound syndrome in which suppressed emotions reflecting anger, disappointment, sadness, misery, hostility, grudges, and unfulfilled dreams or expectations manifest themselves physically. Symptoms include chronic indigestion, poor appetite, constipation, heart palpitations, pains in knees or legs, cold hands or feet, vomiting blood, altered sensory perception, nightmares, decreased urine output, and hypothyroidism. Because of the inappropriateness of and stigma attached to expressing psychological symptoms, Hwabyung may allow Koreans to deal with life problems by linking together emotional and bodily distresses in a model congruent with their cultural context. More recently, Ryder et al. (2008) found differences in symptom expression where depressed Canadian outpatients reported more psychological symptoms than did depressed Chinese outpatients. Ryder and colleagues explored depressive symptom presentations among Chinese and Euro 62       leong and kalibatseva

Canadian outpatients and concluded that the type of assessment (spontaneous problem report, symptom self-report questionnaire, or structured clinical interview) influenced the type and frequency of the symptoms that the patients reported. Chinese outpatients were found to report more depressive somatic symptoms in spontaneous report and structured interviews, whereas Euro Canadian outpatients reported significantly more depressive affective symptoms (e.g., depressed mood, anhedonia, worthlessness, guilt) in all three assessment modalities. On the basis of their findings, Ryder et al. suggested that researchers may have spent too much time discussing Chinese somatization of depression. Instead, they argued that it is more likely that Westerners overemphasize the affective or psychological aspects of depression compared with other cultures. This phenomenon is referred to as the psychologization of depression (Ryder et al., 2008). The DSM classification system has received criticisms for its ethnocentricity and lack of cultural objectivity (Chang & Fabian, 2012; Chang & Kwon, 2014; Fabrega, 1996). Krener and Sabin (1985) suggested that cultural values of American society regarding child development underlie Diagnostic and Statistical Manual of Mental Disorders (3rd ed.; DSM–III; American Psychiatric Association, 1980) categorization of normal and abnormal behaviors. On the basis of their diagnostic and treatment work with Indochinese immigrant children, they found DSM–III diagnostic categories to have little applicability cross-culturally. Although it was possible to assign a diagnosis, diagnoses were often incomplete or inadequate for understanding the true problems. In particular, a misdiagnosis of attention deficit disorder may be made if there are differences in the age at which the culture socializes a child to be attentive and quiet. In addition, separation anxiety may be the norm in a Vietnamese child who sleeps with his or her parents and worries that they may be harmed by the gods (Krener & Sabin, 1985). In both cases, behaviors that may be considered indicative of psychological disorders are actually the norm in a different culture. THERAPIST BIAS IN CLINICAL JUDGMENT Therapist bias is the third source of threat to cultural validity in clinical diagnosis. In this case, therapist bias may be also conceptualized as culturebased countertransference because the clients’ racial, ethnic, and cultural characteristics are likely to elicit therapist reactions, which may in turn affect the services provided (Constantine, 2007; Rosenberger & Hayes, 2002). For example, Katz, Cole, and Lowery (1969) asked experienced British (n = 40) and American (n = 40) psychiatrists to view a filmed diagnostic interview of an American patient and then provide a diagnosis. They found a clear preference in each group for certain diagnostic categories over others, with threats to cultural validity     

63

two thirds of the Americans diagnosing the case as schizophrenic, whereas no British psychiatrist did so. Over 50% of the British group diagnosed the client with an affective disorder. When the level of psychopathology of the cases was controlled for, the British psychiatrists saw significantly more depression than did the American psychiatrists. Overall, racial or ethnic differences may affect therapist clinical judgment and assessment such that a therapist may overpathologize a culturally different client (Whaley, 1997). Problems of misdiagnosis due to therapist bias have been observed with African Americans, Hispanics, and Asian Americans (Garb, 1997). An analog study by Li-Repac (1980) showed that the degree of familiarity with the patient’s cultural background may influence the diagnostic process. Li-Repac had five White and five Chinese American therapists rate the same Chinese and White clients on a videotaped interview on several dimensions. The author found that the White therapists were more accurate in predicting self-descriptive responses of White clients than they were of Chinese clients. More important, Li-Repac found that the White therapists rated the Chinese clients higher on the depression/inhibition dimension and lower on the social poise/interpersonal capacity dimension than did the Chinese American therapists. Moreover, the Chinese American therapists judged the White clients to be more severely disturbed than did the White therapists. These findings point to the subjective nature of assessment data interpretation and clinical diagnosis. Russell, Fujino, Sue, Cheung, and Snowden (1996) examined the effects of therapist–client ethnic match in the assessment of mental health functioning. Therapists evaluated clients’ mental health functioning based on the Global Assessment Scale. The sample consisted of thousands of outpatient clients in the Los Angeles County mental health system, including African Americans, Mexican Americans, Asian Americans, and White Americans. The results indicated that therapists who were ethnically matched with clients provided a higher mental health functioning score than did therapists in mismatched therapy dyads. This effect was still evident for African Americans and Asian Americans after controlling for age, gender, marital status, and referral source. The study showed that factors associated with therapist–client ethnic match may be important in the assessment of mental health functioning in therapy. LANGUAGE CAPABILITY OF THE CLIENT Language capability of the client is another source of threat to cultural validity in assessment and clinical diagnosis. As Leong (1986) pointed out, language may serve as a barrier to effective therapeutic communication in 64       leong and kalibatseva

several ways. Asian Americans who speak little or no English may be misunderstood by their therapists (D. W. Sue, 1981; S. Sue & Morishima, 1982). As Shuy (1983) indicated, the use of dialects or nonstandard English may interfere with the effective exchange of information or worse, stimulate bias in the therapist performing the evaluation. Another language problem in clinical diagnosis is the use of interpreters and its effects on diagnostic evaluations with Asian Americans. Problems inherent in interpreter-mediated interviews are of particular relevance to immigrant clients from Asian countries because many of them may not speak or understand English (Goh, Dunnigan, & Schuchman, 2004; Lee, 1980). A few studies have shown that the use of interpreters may result in distortions that may negatively influence diagnostic evaluations. Sabin (1975) reviewed the clinical records of two suicide cases among Spanish-speaking patients who were evaluated by English-speaking psychiatrists using interpreters. It was found that the degree of patients’ emotional suffering and despair may have been underestimated because of distortions by the interpretation process. The clinician effect (i.e., error due to a particular clinician) was ruled out because both clinicians conducting the evaluations made the same errors. Sabin concluded that the diagnostic errors were due to the interpreter effect and not the clinician effect. Language capability is also relevant to diagnosis if the therapist and client do not attach the same meanings and connotations to words used in diagnosing the patient’s problems. Imada’s (1989) study compared the Japanese equivalents of anxiety (Fu-an), fear (Kyo-fu), and depression (Yu-utsu) for their connotative meanings in an effort to find out whether direct translation of these terms from English to Japanese brings all relevant connotations. The Japanese experience of Fu-an seemed more psychological than physical and overlapped very little in connotative meaning with anxiety. Both anxiety and depression were reported as more physical experiences than were Fu-an and Yu-utsu. By asking subjects to think of a specific instance in which they had experienced each emotion, it was determined that American descriptions of the experience of anxiety often referred to not being able to attain goals they were actively pursuing. Japanese descriptions of the experience of Fu-an often referred to an uneasy expectancy of losing peace and comfort they had already attained. Thus, the connotations of Fu-an and anxiety in a lay context seemed different. Recently, Yakushko (2010) examined the personal and professional factors related to positive and successful clinical care with individuals with limited English proficiency. The results suggested that key characteristics of both therapists and interpreters are being flexible, open to learning, and attentive to nuances. In addition, settings that supported less traditional and structured provision of mental health services provided best clinical threats to cultural validity     

65

care, which often consisted of longer or more frequent sessions. In addition, culture-specific knowledge and skills were essential for the successful practice of both therapists and interpreters. Thus, clinicians and interpreters who may lack flexibility, openness, and knowledge of the client’s culture may not be able to provide best clinical care to clients with limited English proficiency. INAPPROPRIATE USE OF CLINICAL AND PERSONALITY TESTS A fifth source of threat to cultural validity is the inappropriate use of clinical and personality tests. In general, clinical and personality test results have tended to show that Asian Americans have more severe symptoms and profiles than do Whites (Leong, 1986). Although a convergence of data from many studies indicates that Asian Americans have more neurotic and disturbed personality profiles on objective self-report measures, the results of these tests still need to be interpreted with a great deal of caution for several reasons. First, most of these measures were developed and normed on White samples. They were designed to be predictive for Whites and their predictive validity for other ethnic and cultural groups needs to be independently established. Second, although promising, only a few of the clinical diagnostic instruments have been translated into Asian languages. Takeuchi, Kuo, Kim, and Leaf (1989) challenged the assumption that instruments standardized on Whites used to assess need for mental health services can be used on racial and ethnic minorities. They analyzed data from the Symptom Checklist (SCL) for four ethnic groups in Hawaii: Whites, Filipinos, Japanese, and Native Hawaiians. The SCL factor dimensions are anxiety, depression, interpersonal sensitivity, obsessive–compulsive, and somatization. As expected, Whites had the highest number of items that fell into hypothesized factor dimensions (27), although this number was still only half of the total number of items on the SCL (54). However, the number of items that loaded on the six factors was substantially lower for the ethnic minority groups. Only 19 of the 54 items loaded for the Japanese participants, followed by Filipinos (14) and Native Hawaiians (6). Leong, Okazaki, and Tak (2003) reviewed the assessment of depression and anxiety in Asia based on self-report measures. The authors found that only a limited number of instruments were translated. According to the authors, the instruments that are currently used seem to be valid and reliable; however, these imported measures may not be able to capture culture-specific expressions of these constructs. For example, Cheung, van de Vijver, and Leong (2011) discussed the importance of the indigenously 66       leong and kalibatseva

derived Interpersonal Relatedness factor in Chinese samples. The authors reviewed the study of culture and personality measurement and suggested that although Western models of personality structure may provide a valid picture of intraindividual aspects of personality, they need to be complemented with indigenous models of personality that focus on the social and relational aspects of personality. In addition, Leong, Leung, and Cheung (2010) suggested that one of the fundamental problems in cross-cultural research relates to the measurement equivalence of the tests and measures researchers use. Similarly, the measurement equivalence of personality and diagnostic tests is problematic in assessment and diagnosis. In particular, these tests may be inappropriately used if their linguistic, functional, conceptual, and metric equivalence has not been established. Finally, in using personality and diagnostic tests with Asian Americans, one needs to recognize that there may be important cross-cultural differences in definitions of mental illness and mental health, and that there are indeed “many ways of being human” (Tyler et al., 1985, p. 312). Clinicians who use the existing clinical and personality tests to diagnose Asian American patients without being aware of these issues and limitations may formulate culturally invalid diagnoses. CONCLUSION AND RECOMMENDATIONS The problem of cultural validity in clinical diagnosis and assessment of Asian Americans has been shown to result from at least five factors: pathoplasticity of psychological disorders, cultural factors influencing symptom expression, therapist bias in clinical judgment, language capability of the client, and inappropriate use of clinical and personality tests. Although existing literature provides examples of these sources of threats to cultural validity, more research is needed in each of these areas that targets specific settings or populations so that firmer conclusions may be drawn using larger amounts of congruent data. If one conclusion is clear, however, it is that cultural factors are important to clinical diagnosis and assessment of psychopathology. Although this chapter has illustrated several examples of threats to cultural validity, therapists may still wonder how to eliminate these threats. Following are some guidelines and suggestions for reducing or eliminating these sources of threats to cultural validity. The first recommendation is for therapists to conduct culturally relevant interventions. Crystal (1989), who suggested this idea to improve service delivery, explained that, for Asian Americans, the therapist may be asked threats to cultural validity     

67

and expected to respond to personal questions about his or her family, age, and so on, which is unlike most Americans’ expectations of professionals to be formal, distant, and rather impersonal. Also, Asian Americans may view the therapist as an authority figure who is more directive and in command, rather than one who waits for the patient to solve his or her own problems via open-ended discussions. The second recommendation is for therapists to be aware of communication and language differences. Crystal (1989) pointed out that many Asians communicate nonverbally and by use of silence. The therapist has to be able to recognize communication style differences and not confuse them with patients’ resistance or problems with expressing oneself. Also, as detailed before, interpreter problems must be taken into consideration when counseling clients with limited English proficiency. A third recommendation is that counselors must be careful not to misinterpret cultural effects. In the case of Asian Americans, Crystal (1989) suggested that dependency is closely related to family loyalty and a sense of obligation. Thus, the therapist who insists on combating dependency tendencies by making the client more self-assertive may not be approaching the problem correctly; that is, he or she may misinterpret filial piety as dependency. D. Sue and Sue (1987) pointed out that subtleties to behaviors are often missed when counseling ethnic clients. For example, Asian Americans may tend to be nonassertive, but American society requires people to be assertive. Thus, Asian Americans do show deficits on assertiveness. However, the more important consideration is how the nonassertiveness came into being and is maintained. The fourth suggestion (Crystal, 1989) is that informal support sources, such as churches, ethnic clubs, family associations, and community leaders, be called upon for assistance in understanding cultural traditions and systems when working with Asian American clients. This solution is consistent with the fact that many Asian Americans subscribe to a more collectivistic view, rather than the Western individualistic outlook. A fifth recommendation includes ensuring the measurement equivalence of the assessment instruments therapists use (Leong et al., 2010). Because the norms of the majority of personality and diagnostic tests were established with predominantly White samples, it is important to remember that certain clinical elevations or low scores may be associated with cultural differences rather than psychopathology. Thus, instruments without established measurement equivalence for other cultural groups need to be interpreted with caution. A sixth recommendation is to be aware of within-group differences as well as between-group differences. D. Sue and Sue (1987) emphasized that although many Asian Americans share certain cultural styles and traditions, 68       leong and kalibatseva

there are many differences between groups of Asian Americans as well. With the increase of immigration from different Asian countries, this disparity will increase, as will generational differences, with time. Even members of the same ethnic group (e.g., Japanese Americans) can vary widely in language ability, acculturation level, socioeconomic status, religious practices, and generation in the United States. Another factor that may be overlooked is the effect that the society in which the client lives has on his or her perceptions and how these may vary by region. For example, Chinese Americans living in Chinatowns in large cities may be facing different problems than do those living in rural areas. Resolving problems with standard means of diagnosis is the seventh guideline to avoid threats to cultural validity. For example, Velásquez, Johnson, and Brown-Cheatham (1993) identified four problems when using the Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev.; DSM– III–R; American Psychiatric Association, 1987) with ethnic minority clients and developed a framework for teaching counseling students how to avoid making culturally invalid diagnoses. The four steps were (a) training students to learn the fundamentals of diagnosing via the DSM, (b) teaching students to be critical of the DSM and recognize differences between culturally sensitive and insensitive diagnoses, (c) challenging students to examine their fundamental beliefs or assumptions concerning abnormal or normal behavior, and (d) requiring students to apply their diagnostic skills with ethnic minority clients in actual clinical practice. In short, to establish treatment alliance, mental health professionals need to understand and comply, when appropriate, with the implicit model that clients from different ethnic backgrounds may have about the service provider. The therapists also need to be aware of potential language differences and culturally determined behavioral differences among culturally diverse clients and outreach to their communities when necessary. Measurement equivalence, clinical judgment, and cultural sensitivity in applying diagnostic categories also need to be ensured when working with culturally diverse individuals. Yet, it is important to remember that there is a wide variation among ethnically diverse groups. The crux of these recommendations is that it is imperative for therapists to forgo the emic perspectives on assessing and diagnosing culturally different clients. Although it may be convenient to use broad categorizations for diagnosing all people seen in mental health settings, the ramifications of doing so with culturally different clients can be severe. By demonstrating these threats to cultural validity in the Asian American population, we hope that this problem will be given more attention in research and clinical practice and similar threats will be examined among other racial and ethnic groups. threats to cultural validity     

69

REFERENCES Adebimpe, V. R. (2004). A second opinion on the White norms in psychiatric diagnosis of Black patients. Psychiatric Annals, 34, 542–551. http://dx.doi.org/10.3928/ 0048-5713-20040701-15 Alberque, C., & Eytan, A. (2001). Chronic pain presenting as major depression in a cross-cultural setting. International Journal of Psychiatry in Medicine, 31, 73–76. http://dx.doi.org/10.2190/3T68-99R0-QBVL-FNN0 American Psychiatric Association. (1968). Diagnostic and statistical manual of mental disorders (2nd ed.). Washington, DC: Author. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Arlington, VA: Author. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Arlington, VA: Author. Bracht, G. H., & Glass, G. V. (1968). The external validity of experiments. American Educational Research Journal, 5, 437–474. http://dx.doi.org/10.3102/ 00028312005004437 Campbell, D. T., & Stanley, J. C. (1966). Experimental and quasi-experimental designs for research. Chicago, IL: Rand McNally. Canino, G., & Alegría, M. (2008). Psychiatric diagnosis—Is it universal or relative to culture? Journal of Child Psychology and Psychiatry, 49, 237–250. http://dx.doi. org/10.1111/j.1469-7610.2007.01854.x Chang, E. C., & Fabian, C. G. (2012). Introduction: From nomothetic to idiographic approaches. In E. C. Chang (Ed.), Handbook of adult psychopathology in Asians: Diagnosis, etiology, and treatment (pp. 1–11). http://dx.doi.org/10.1093/ med:psych/9780195179064.003.0001 Chang, E. C., & Kwon, P. (2014). Special issue on psychopathology in Asians and the DSM–5: Culture matters. Asian Journal of Psychiatry, 7, 66–67. http://dx.doi. org/10.1016/j.ajp.2013.12.001 Cheung, F. M., Leong, F. T. L., & Ben-Porath, Y. S. (2003). Psychological assessment in Asia: Introduction to the special section. Psychological Assessment, 15, 243–247. http://dx.doi.org/10.1037/1040-3590.15.3.243 Cheung, F. M., van de Vijver, F. J. R., & Leong, F. T. L. (2011). Toward a new approach to the study of personality in culture. American Psychologist, 66, 593–603. http://dx.doi.org/10.1037/a0022389 Chu, J. P., & Sue, S. (2011). Asian American mental health: What we know and what we don’t know. Online Readings in Psychology and Culture, 3(1). http:// dx.doi.org/10.9707/2307-0919.1026 Constantine, M. G. (2007). Racial microaggressions against African American clients in cross-racial counseling relationships. Journal of Counseling Psychology, 54, 1–16. http://dx.doi.org/10.1037/0022-0167.54.1.1

70       leong and kalibatseva

Crystal, D. (1989). Asian Americans and the myth of the model minority. Social Casework: The Journal of Contemporary Social Work, 70, 405–413. Durvasula, R., & Sue, S. (1996). Severity of disturbance among Asian American outpatients. Cultural Diversity and Mental Health, 2, 43–51. http://dx.doi.org/ 10.1037/1099-9809.2.1.43 Eap, S., Gobin, R. L., Ng, J., & Hall, G. C. N. (2010). Sociocultural issues in the diagnosis and assessment of psychological disorders. In J. E. Maddux & J. P. Tangney (Eds.), Social psychological foundations of clinical psychology (pp. 312–328). New York, NY: Guilford Press. Enright, J. B., & Jaeckle, W. R. (1963). Psychiatric symptoms and diagnosis in two subcultures. The International Journal of Social Psychiatry, 9, 12–17. http://dx.doi. org/10.1177/002076406300900102 Fabrega, H. (1996). Cultural and historical foundations of psychiatric diagnosis. In J. E. Mezzich, A. Kleinman, H. Fabrega, & D. L. Parron (Eds.), Culture and psychiatric diagnosis: A DSM–IV perspective (pp. 3–14). Washington, DC: American Psychiatric Press. Garb, H. N. (1997). Race bias, social class bias, and gender bias in clinical judgment. Clinical Psychology: Science and Practice, 4, 99–120. http://dx.doi. org/10.1111/j.1468-2850.1997.tb00104.x Garfield, S. (1984). Methodological problems in clinical diagnosis. In H. E. Adams & P. B. Sutker (Eds.), Comprehensive handbook of psychopathology (pp. 27–45). http://dx.doi.org/10.1007/978-1-4615-6681-6_2 Goh, M., Dunnigan, T., & Schuchman, K. M. (2004). Bias in counseling Hmong clients with limited English proficiency. In J. L. Chin (Ed.), The psychology of prejudice and discrimination: Ethnicity and multiracial identity (Vol. 2, pp. 109–136). Westport, CT: Praeger/Greenwood. Guindon, M. H., & Sobhany, M. S. (2001). Toward cultural competency in diagnosis. International Journal for the Advancement of Counselling, 23, 269–282. http:// dx.doi.org/10.1023/A:1014443901294 Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic personality schedule (Minnesota): 1. Construction of the schedule. Journal of Psychology, 10, 249–254. Imada, H. (1989). Cross-language comparisons of emotional terms with special reference to the concept of anxiety. Japanese Psychological Research, 31, 10–19. Jackson, J. S., Abelson, J. M., Berglund, P. A., Mezuk, B., Torres, M., & Zhang, R. (2011). Ethnicity, immigration, and cultural influences on the nature and distribution of mental disorders: An examination of major depression. In D. Regier, W. Narrow, E. Kuhl, & D. Kupfer (Eds.), Conceptual evolution of DSM–5 (pp. 267–285). Arlington, VA: American Psychiatric Publishing. Katz, M. M., Cole, J. O., & Lowery, H. A. (1969). Studies of the diagnostic process: The influence of symptom perception, past experiences, and ethnic background on diagnostic decisions. The American Journal of Psychiatry, 125, 937–947. http:// dx.doi.org/10.1176/ajp.125.7.937 threats to cultural validity     

71

Katz, M. M., Sanborn, K. O., & Gudeman, H. (1969). Characterizing differences in psychopathology among ethnic groups: A preliminary report on HawaiianJapanese and Mainland American schizophrenics. In W. Caudill & T. Y. Lin (Eds.), Mental health research in Asia and the Pacific (pp. 148–163). Honolulu, HI: East-West Center Press. Kleinman, A. (1982). Neurasthenia and depression: A study of somatization and culture in China. Culture, Medicine and Psychiatry, 6, 117–190. http://dx.doi. org/10.1007/BF00051427 Krener, P. G., & Sabin, C. (1985). Indochinese immigrant children: Problems in psychiatric diagnosis. Journal of the American Academy of Child Psychiatry, 24, 453–458. http://dx.doi.org/10.1016/S0002-7138(09)60564-1 Lee, E. (1980). Mental health services for the Asian-Americans: Problems and alternatives. In U.S. Commission of Civil Rights, civil rights issues of Asian and PacificAmericans: Myths & realities (pp. 676–680). Washington, DC: U.S. Government Printing Office. Leong, F. T. L. (1986). Counseling and psychotherapy with Asian Americans: Review of the literature. Journal of Counseling Psychology, 33, 196–206. http:// dx.doi.org/10.1037/0022-0167.33.2.196 Leong, F. T. L., Leung, K., & Cheung, F. M. (2010). Integrating cross-cultural psychology research methods into ethnic minority psychology. Cultural Diversity and Ethnic Minority Psychology, 16, 590–597. http://dx.doi.org/10.1037/a0020127 Leong, F. T. L., Okazaki, S., & Tak, J. (2003). Assessment of depression and anxiety in East Asia. Psychological Assessment, 15, 290–305. Li-Repac, D. (1980). Cultural influences on clinical perception: A comparison between Caucasian and Chinese-American therapists. Journal of Cross-Cultural Psychology, 11, 327–342. http://dx.doi.org/10.1177/0022022180113006 Malgady, R. G. (1996). The question of cultural bias in assessment and diagnosis of ethnic minority clients: Let’s reject the null hypothesis. Professional Psychology: Research and Practice, 27, 73–77. http://dx.doi.org/10.1037/0735-7028.27.1.73 Pang, K. Y. (1990). Hwabyung: The construction of a Korean popular illness among Korean elderly immigrant women in the United States. Culture, Medicine and Psychiatry, 14, 495–512. Parker, G., Cheah, Y.-C., & Roy, K. (2001). Do the Chinese somatize depression? A cross-cultural study. Social Psychiatry and Psychiatric Epidemiology, 36, 287–293. http://dx.doi.org/10.1007/s001270170046 Rosenberger, E., & Hayes, J. (2002). Therapist as subject: A review of the empirical countertransference literature. Journal of Counseling & Development, 80, 264–270. http://dx.doi.org/10.1002/j.1556-6678.2002.tb00190.x Russell, G. L., Fujino, D. C., Sue, S., Cheung, M.-K., & Snowden, L. R. (1996). The effect of therapist–client ethnic match in the assessment of mental health functioning. Journal of Cross-Cultural Psychology, 27, 598–615. http://dx.doi. org/10.1177/0022022196275007

72       leong and kalibatseva

Ryder, A. G., Yang, J., Zhu, X., Yao, S., Yi, J., Heine, S. J., & Bagby, R. M. (2008). The cultural shaping of depression: Somatic symptoms in China, psychological symptoms in North America? Journal of Abnormal Psychology, 117, 300–313. http://dx.doi.org/10.1037/0021-843X.117.2.300 Sabin, J. E. (1975). Translating despair. The American Journal of Psychiatry, 132, 197–199. http://dx.doi.org/10.1176/ajp.132.2.197 Sadler, J. Z. (2005). Values and psychiatric diagnosis. Oxford, England: Oxford University Press. Shuy, R. W. (1983). Three types of interference to an effective exchange of information in the medical interview. In S. Fisher & A. D. Todd (Eds.), The social organization of doctor–patient communication (pp. 189–202). Washington, DC: Center for Applied Linguistics. Solano-Flores, G., & Nelson-Barber, S. (2001). On the cultural validity of science assessments. Journal of Research in Science Teaching, 38, 553–573. http://dx.doi. org/10.1002/tea.1018 Sue, D., & Sue, S. (1987). Cultural factors in the clinical assessment of Asian Americans. Journal of Consulting and Clinical Psychology, 55, 479–487. http://dx.doi. org/10.1037/0022-006X.55.4.479 Sue, D. W. (1981). Counseling the culturally different: Theory and practice. New York, NY: Wiley. Sue, S. (1977). Community mental health services to minority groups. Some optimism, some pessimism. American Psychologist, 32, 616–624. http://dx.doi. org/10.1037/0003-066X.32.8.616 Sue, S., & McKinney, H. (1975). Asian Americans in the community mental health care system. American Journal of Orthopsychiatry, 45, 111–118. http://dx.doi. org/10.1111/j.1939-0025.1975.tb01172.x Sue, S., & Morishima, J. K. (1982). The mental health of Asian Americans. San Francisco, CA: Jossey-Bass. Sue, S., & Sue, D. W. (1974). MMPI comparisons between Asian American and non-Asian students utilizing a student health psychiatric clinic. Journal of Counseling Psychology, 21, 423–427. http://dx.doi.org/10.1037/h0037074 Takeuchi, D. T., Kuo, H., Kim, K., & Leaf, P. J. (1989). Psychiatric symptom dimensions among Asian Americans and Native Hawaiians: An analysis of the symptom checklist. Journal of Community Psychology, 17, 319–329. http:// dx.doi.org/10.1002/1520-6629(198910)17:4 3.0.CO;2-Q Tyler, F. B., Sussewell, D. R., & Williams-McCoy, J. (1985). Ethnic validity in psychotherapy. Psychotherapy, 22, 311–320. Velásquez, R. J., Johnson, R., & Brown-Cheatham, M. (1993). Teaching counselors to use the DSM–III–R with ethnic minority clients: A paradigm. Counselor Education and Supervision, 32, 323–331. http://dx.doi.org/10.1002/j.1556-6978.1993. tb00259.x threats to cultural validity     

73

Westermeyer, J. (1985). Psychiatric diagnosis across cultural boundaries. The American Journal of Psychiatry, 142, 798–805. http://dx.doi.org/10.1176/ajp.142.7.798 Whaley, A. L. (1997). Ethnicity/race, paranoia, and psychiatric diagnoses: Clinical bias versus sociocultural differences. Journal of Psychopathology and Behavioral Assessment, 19, 1–20. http://dx.doi.org/10.1007/BF02263226 Yakushko, O. (2010). Clinical work with limited English proficiency clients: A phenomenological exploration. Professional Psychology: Research and Practice, 41, 449–455. http://dx.doi.org/10.1037/a0020996 Yang, L. H., & WonPat-Borja, A. J. (2007). Psychopathology among Asian Americans. In F. T. L. Leong, A. Ebreo, L. Kinoshita, A. G. Inman, & L. H. Yang (Eds.), Handbook of Asian American psychology (2nd ed., pp. 379–405). Thousand Oaks, CA: Sage.

74       leong and kalibatseva

4 STATISTICAL AND METHODOLOGICAL ISSUES IN PLANNING RANDOMIZED CLINICAL TRIALS WITH ETHNIC MINORITIES CARMEN L. RIVERA-MEDINA AND JOSé N. CARABALLO

Even though ethnic minority groups have been consistently grow­ ing, comprising over a third of the population in the United States, ethno­ cultural groups (ECGs) represent less than 10% of participants in clinical trials (Sangi-Haghpeykar, Meddaugh, Liu, & Grino, 2009). After a review of 40 years of publications in the PsycINFO database, Bernal, Trimble, Burlew, and Leong (2003) confirmed that the scientific information available regard­ ing diverse ethnic, racial, or language groups is not commensurate with their population rates. The scarcity of available mental health interventions for ECGs not only prevents them from receiving evidence-based treatments (EBTs) according to their particular needs but also prevents the proper assessment EBTs for ECGs that could be used as the gold standard for future The authors wish to thank Dr. Rafael Ramirez and Dr. Patrick Shrout for the topics suggested for the chapter as well as research assistants Raúl I. Camacho (for helping with the references and organization) and Ronald Brown (for the English editing). This work was supported by the Institute for Psychological Research at the University of Puerto Rico, Río Piedras. http://dx.doi.org/10.1037/14940-005 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

75

interventions and evaluation methods such as benchmarking (Weersing, 2005) or meta-analysis. In this chapter, we consider several issues pertaining to the design and analysis of randomized controlled trials (RCTs). Although these issues are not exclusive to RCTs with minority populations, they are critical to ensure the quality necessary if they will serve as the gold standard in the evaluation of the effectiveness of the interventions, having immediate impact on the internal and external validity of the studies. An overview of some metaanalyses demonstrates how a significant percentage of potential studies to be evaluated are excluded because of quality issues related to design, sample size, inadequate outcome selection, and mishandling of missing data, among others (Lawlor & Hopker, 2001; Linde et al., 1996; Troeung, Egan, & Gasson, 2013; Wang et al., 2008). When planning an RCT, the researchers must take into account sev­ eral statistical and methodological considerations to obtain valid conclusions regarding the effectiveness of the treatment. Among these are (a) type of variables to be used as outcomes, which determine the analytic strategy to be used; (b) sample size and its impact on the power to detect the effective­ ness of the intervention; (c) handling of missing data; (d) consequences of adding secondary outcomes or of multiple testing; and (e) specific questions to answer with the data obtained that determine the information needed for the study and how it would be analyzed. Ignoring these issues or treating them too lightly at initial phases of an RCT could evolve into later statistical challenges when working with ECGs that could substantially undermine the internal validity of the study. In this chapter, we discuss some methodological concerns that directly affect the results or statistical strategy to be used in an RCT with ethnic minorities, and we provide some strategies for dealing with those concerns. Researchers considering obtaining external funds from an agency such as the National Institutes of Health (NIH) need to be aware of policy issues that may impact the quality of their RCTs with minority groups. A leading issue is how primary outcomes should be chosen and how they should be opera­ tionalized (dimensional vs. dichotomous) according to the research question. The intent to treat (ITT) analysis is a strategy, once thought of as the rule of thumb for RCTs, that has brought about several concerns in the field. We also present the multiplicity of testing issues when considering primary and secondary outcomes in RCTs, therefore increasing the possibility of a Type II error in the results. In addition, we present some issues that may affect the power of analyses and results. Finally, we describe how to handle missing data. Although the topics discussed in this chapter are directly related to statisti­ cal concerns, they are covered more in a conceptual than in a technical or mathematical way. 76       rivera-medina and caraballo

NATIONAL INSTITUTES OF HEALTH POLICY ISSUES A substantial percentage of RCTs could be submitted for NIH fund­ ing. More than a decade ago, NIH established a policy for the inclusion of women and members of minority groups (NIH, 2001) in all NIH-funded clinical research. Although this policy is an effort to increase the representa­ tion of women, minorities, and children in NIH treatment research, there are serious methodological and statistical concerns if the researcher imple­ ments the policy as suggested. First, the NIH policy did not necessarily help reduce the gap in the evidence-based treatments available for ethnic minori­ ties (Miranda, Nakamura, & Bernal, 2003). Second, it established guidelines that are confusing and contradictory, bringing in additional methodological and statistical challenges for investigators who attempt to put together a wellthought-out RCT to produce results that reflect the cultural, linguistic, and socioeconomic diversity of the population. For example, NIH’s definition for detecting significant differences1 in the 2001 policy also has implications for design and sampling decisions that directly impact the analysis and results obtained from the RCT, therefore affecting its internal validity. NIH’s ambiguous definition of significant dif­ ferences gives the impression that a representative or large enough sample would not be necessary because it would not produce data of clinical impor­ tance. Their statement resembles the discussions related to the interpreta­ tion of results based on statistical significance tests, effect size, or clinical significance that have been a source of debate in the literature for more than 30 years (Atkins, Bedics, McGlinchey, & Beauchaine, 2005; Bauer, Lambert, & Nielsen, 2004; Hayes & Haas, 1988; Jacobson, 1988; Jacobson & Truax, 1991; Shaver, 1993; Thompson, 1996, 2002). The lack of consensus has led researchers to report effect size, p values, and clinical significance in RCTs. However, in an RCT with diverse ECGs, the researcher may be interested not only in the overall clinical significance, which implies the compari­ son between groups after therapy with well-functioning peers (Jacobson & Revenstorf, 1988), but also in the comparison of therapy effects across and within diverse ethnic groups. Obtaining a reasonable sample size to attain these objectives may be an almost insurmountable challenge. The question

“For purposes of this policy, a ‘significant difference’ is a difference that is of clinical or public health importance, based on substantial scientific data. This definition differs from the commonly used ‘statistical significant difference,’ which refers to the event that, for a given set of data, the statistical test for a difference between the effects in two groups achieves statistical significance. Statistical significance depends upon the amount of information in the data set. With a very large amount of information, one could find a statistically significant, but clinically small difference that is of very little clinical importance. Conversely, with less information one could find a large difference of potential importance that is not statistically significant” (NIH, 2001). 1

statistical and methodological issues     

77

is, can we reliably detect an EBT effect for an ethnic minority group if this group is represented in the sample in proportion to its size in the U.S. popula­ tion? Add to this the diversity of ECGs (African Americans, Latinos, Asian Americans, and Native Americans) in the United States. The problem is further complicated when the heterogeneity within the same ECG is con­ sidered as the case with Latinos or Asian Americans (Bernal, Cumba-Avilés, & Rodríguez-Quintana, 2014; Rogler, Malgady, Costantino, & Blumenthal, 1987). Miranda and colleagues (2003) pointed out that trying to include rep­ resentative numbers of minorities in samples to comply with the policy may produce unreliable findings because the number of participants from ethnic minority groups would generally be too small. In trying to comply with the NIH policy for conducting a valid analysis, another concern surfaces. One of the principal requirements in the policy for ensuring valid analysis is the unbiased evaluation of the outcome(s) of the study participants. However, to ensure a valid analysis, one should start by employing valid instruments to collect unbiased data that ensures measure­ ment equivalence across diverse groups and cultures. A more in-depth treat­ ment of this topic is presented in Chapter 6, this volume, so we only highlight the seriousness of this issue because biased instruments would result in biased results for ECGs. To establish measurement equivalence, so that an unbiased measure is used, researchers should conduct a meticulous evaluation of the psychometric properties of the measures to be used, which goes beyond just the computation of a reliability coefficient. However, the plans to evaluate the validity of the instruments within an ECG in a researcher’s proposal are constantly questioned by the NIH or other funding agencies in the name of generalizability of the results. CHOOSING PRIMARY OUTCOMES OF THE TREATMENT RESEARCH Statistical consultants commonly confront the situation in which they are asked to give advice on testing the effectiveness of an intervention once the intervention is already in course or (worse) when the data are already collected. Too frequently the information obtained to assess effectiveness does not necessarily respond to the research question of the study because of how it was operationalized, thus restricting the statistical analysis that can be performed. Specifically, in RCTs the most common expected outcome is the remission of the mental health condition diagnosed at the beginning of the study; therefore, investigators may think that a yes or no answer is suf­ ficient (a dichotomous primary outcome). However, the decision of whether to use dichotomous (presence/absence of the disorder or condition by the 78       rivera-medina and caraballo

end of treatment) or dimensional (presence of symptoms that characterize the disorder) outcomes in RCT has been given much attention in the litera­ ture (Demjaha et al., 2009; Dutta et al., 2007; Helzer, Bucholz, & Gossop, 2007; Kraemer, 2007; Kraemer, Noda, & O’Hara, 2004; Kraemer, Shrout, & Rubio-Stipec, 2007; MacCallum, Zhang, Preacher, & Rucker, 2002; Widiger & Samuel, 2005). For years, investigators and clinicians have argued over the issue of determining how to measure progress as the result of an intervention. Significant discussions are still to come given the withdrawal of the National Institute of Mental Health’s support (2013) of all research that established its diagnostic criteria based on the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM–5; American Psychiatric Association, 2013). The movement toward the use of research domain criteria is an attempt by the NIH to use a dimensional approach to diagnosis but unfortunately with no clear and specific criteria on how to do it. From the statistical point of view, whether an investigator uses one criterion (e.g., DSM–5, International Classification of Diseases [10th ed.; ICD–10]) or another is not as important as whether the criterion to be employed (dimensional or categorical) will answer the research question in an RCT. However, the decision of whether to use a categorical over a dimensional outcome imposes restrictions that directly affect the precision of the RCT analysis and results. The advantages of dimensional over categorical outcomes have been discussed in detail over the past 10 years (Dutta et al., 2007; Helzer et al., 2007; Kraemer, 2007; Kraemer et al., 2004, 2007; MacCallum et al., 2002; Royston, Altman, & Sauerbrei, 2006; Streiner, 2002; Widiger & Samuel, 2005), so we offer a review of some of the most important issues. In real­ ity this is not a matter of deciding which outcomes are better than others but rather under which circumstances one should use the outcomes that are appropriate given the research questions (Helzer, Kraemer, & Krueger, 2006; Kraemer et al., 2004). Failure to operationalize the outcome appro­ priately for the question of the study could substantially impair the clinical and research decision-making process. That being said, we do support the inclusion of categorical and dimensional information in RCTs because both can complement the information obtained. The issue is which one should be selected as the primary outcome measure based on the aims of the study. In a study of psychosis, Demjaha et al. (2009) concluded that the categori­ cal diagnostic constructs are enriched and further enhanced by the use of dimensional information. We could not agree more with such a statement. However, the concern is still in choosing the primary outcomes of an RCT to evaluate the effectiveness or efficacy of an intervention. In the endeavor to select an appropriate primary outcome for an RCT, it is important to understand the distinction between disorder and diagno­ sis to clarify the categorical or dimensional approach. A disorder represents statistical and methodological issues     

79

something of clinical concern in a patient (Kraemer et al., 2007), physical or mental (the latter is relevant to mental health researchers). Once pres­ ent, a disorder is really both categorical (present or not) and dimensional (given the characteristics or symptoms cluster that describe the disorder). Diagnosis is the procedure used to identify whether the disorder is present or not (Kraemer et al., 2007). However, regarding outcomes, the problem is not related to the disorder itself but to how to diagnose the disorder, if the diagnosis should follow a categorical or dimensional approach. Diagnosing a disorder categorically implies identifying the participant as having or not having the disorder. Using a dimensional approach for diagnosis results in assigning each participant an ordinal score—the higher the score, the stronger or more severe the disorder (Kraemer et al., 2004). It is also well known that the more the outcome allows the dimension to increase its score, to obtain higher or lower points, the better the likelihood of being able to account for variability among participants and their individuality to treatment response over time. Such an approach allows investigators to evaluate changes in severity from before, during, and after treatment and the range of those changes. If one uses the dichotomous approach, then it is possible to evalu­ ate only when the disorder is or is no longer present; the disorder cannot be characterized. The investigator can always transform a dimensional approach into categorical by defining a threshold or cutoff point but cannot convert a categorical approach into a dimensional one. Furthermore, too much informa­ tion may be lost if the initial decision is to employ only categorical variables. MacCallum et al. (2002) discussed in detail the consequences of dichot­ omizing for measurement and statistical analyses showing that the arguments in favor of such practice have no statistical or conceptual base. After examin­ ing each of the more consistent arguments in favor of dichotomizing, these authors summarized the following consequences: loss of information about individual differences, loss of effect size and power in general and in bivariate relationships, the possibility of a spurious statistical significance and over­ estimation of effect size when considering two independent variables, risk of overlooking nonlinear relationship, and loss of measurement reliability. Although with predictor variables, not with outcome measures, Royston and colleagues (2006) also demonstrated how the use of dichotomization of con­ tinuous variables in a randomized trial should not be applied when using multiple regression, indicating consequences similar to the ones mentioned above and suggesting other alternative methods. It has been proposed that an outcome should be dichotomized only when a continuous variable is highly skewed (MacCallum et al., 2002; Streiner, 2002) with most of the partici­ pants accumulated at the extremes of the variable giving clear evidence of distinct groups. Even then, this dichotomization does not minimize the con­ sequences of such action. 80       rivera-medina and caraballo

Specifically in RCTs, because the hypothesis implied in the study is to test the efficacy of an intervention and the decrease in power in the statistical analyses when a categorical approach is used, the dimensional approach or both should be used. Investigators should also be aware that, when dichoto­ mizing a dimensional outcome, variations on the cutoff point used to dichot­ omize a variable may lead to different conclusions concerning the efficacy of the treatment. Kraemer et al. (2004) showed how results of statistical test­ ing of treatment efficacy varies with the cutoff point. Even with the same instrument, using different cutoff points may lead to erroneous or inaccurate conclusions if they were to be used in diverse ECGs to establish treatment efficacy. As demonstrated by a study with Puerto Rican adolescents (Rivera, Bernal, & Rosselló, 2005), when evaluating the predictive validity of the Children’s Depression Inventory (CDI) for major depression disorder and its suggested cutoff points (Kovacs, 1992), the authors found that a higher cutoff point than the one suggested with the Euro American population was needed. The study also showed that, even using the cutoff point recommended for the instrument by Kovacs (1992), the sensitivity and specificity scores to identify the disorder differed from one population to another. The results provide evidence that investigators from minority groups should not assume metric equivalency across cultures or even within groups as is the case for Latinos. Investigators also need to be aware of two other consequences of dichot­ omizing a dimensional diagnosis: first, the reliability of the study decreases and second, the sample size that would be necessary to obtain precision of the estimates for the reliability of the study increases (Kraemer et al., 2007). The impact of dichotomization could be attenuated when the dimensional mea­ sure is dichotomized at its median in the population, increasing the impact as the cutoff point moves away from the median. However, Kraemer and col­ leagues warned that such a procedure, though it may increase reliability, can reduce validity. We may add to this warning that the median cutoff points for the population may not be adequate for ECGs, thus such procedures are not necessarily reliable. This point is a major issue in RCTs with ECGs because nonreliable outcome measures imply nonreliable diagnosis. As asserted in the literature, even when significant findings are obtained, lack of reliability attenuates treatment effect sizes, adding a negative impact on the study’s clinical significance (Kraemer et al., 2007). The categorical approach to diagnosis could still be important in RCTs when it comes to the decision-making process. Establishing the inclusion and exclusion criteria for the study (based on who is more likely to respond to treat­ ment, who needs to continue in treatment, or who is already in remission) is better answered by a categorical approach (Kraemer et al., 2004). However, investigators should be clear that the method used to establish the categories for the decision-making process has not been determined arbitrarily and has statistical and methodological issues     

81

been validated with the ECG of interest because it may have an impact over the hypothesis testing with dimensional outcomes. Finally, when working with ECGs, investigators should consider that whether they select categorical or dimensional primary outcomes, most of the measures used were developed and validated with the Euro-American population. Most of the disorders and the criteria used to establish a diagnosis for a disorder are also based on criteria developed with Euro-Americans. In this regard, it is important to not lose perspective that there may be serious concerns with the validity of these instruments if applied blindly to ECGs. Consequently, the internal validity as well as the external validity of the outcomes may also be compromised. INTENT-TO-TREAT AND ALTERNATIVES FOR ETHNOCULTURAL GROUPS RCTs have been recognized for decades as the best approach to evalu­ ate effectiveness between different interventions (Altman, 1996; Sussman & Hayward, 2010). Only RCTs allow valid inferences of cause and effect; they are thus considered the gold standard for studying the efficacy of treatment interventions (Polit & Gillespie, 2010; Sussman & Hayward, 2010). Although RCTs have been reported for more than 50 years, it was not until 1996 that the CONSORT statement was published, suggesting to the field all the items that should be included in reporting of RCTs. These guidelines suggest that investigators report the rationale and methods for the statistical analysis and whether they were completed based on an ITT approach (Altman, 1996). For years, there has been a debate about the bias related to RCTs and the implication of those biases in the reliability of results, many of them related to the ITT assumption. ITT is the method of analysis for RCTs in which all patients are analyzed as they were initially randomized regardless of behavior or treatment received (Wright & Sim, 2003). This approach implies that a classic ITT requires either 0% participant attrition or the use of statistical techniques that account for missing outcome data. However, it is precisely the 0% attrition requirement that has brought several criticisms regarding ITT. If the participants of the study do not receive the treatment as it was intended once they are randomized, this may lead to biased results, or what Sussman and Hayward (2010) called treatment contamination. Such treatment contamination can occur because of not receiving the treatment for any reason (participant decision or treatment consequences) or treatment crossover (receive the intervention that was intended for the other group). There is evidence that treatment contamination can be as high as 30% in large trials (Sussman & Hayward, 2010). 82       rivera-medina and caraballo

In an effort to meet ITT requirements and CONSORT guidelines, most investigators end up with a modified ITT in which no participant is removed from the analysis and there are deliberate efforts to follow those who were noncompliant with the treatment or who dropped out (Polit & Gillespie, 2010). Nevertheless, Polit and Gillespie (2010) noted that although the CONSORT guidelines suggest that ITT should be used, these guidelines do not offer suggestions about how to implement or provide a consistent defini­ tion of the ITT approach. These authors consider that the lack of a consistent definition for ITT could be a result of a variety of reasons as to why missing­ ness occurs. It is interesting that, out of the seven reasons provided, four of them are due to decisions made by the investigator as to when and how to remove participants from the study. Two main approaches have been proposed to deal with bias in ITT: (a) as treated and (b) per protocol (Mazumdar, Liu, Houck, & Reynolds, 1999). These, however, have been equally critiqued and consensus on this issue is lacking. As treated implies that the participants are analyzed on the basis of the treatment ultimately received, regardless of the treatment to which they were assigned (Sussman & Hayward, 2010). Per protocol (or com­ pliers only) implies that participants are included in the analysis only if they followed the assigned protocol and are removed from the analysis entirely if they do not follow the protocol. Advocates of this approach assert that it is not sensible to include in the intervention group participants who did not actually receive the intervention, as they will decrease the estimates of true treatment efficacy (Polit & Gillespie, 2010). Likewise, it can be argued that if data are analyzed according to par­ ticipants’ behaviors, assignment is no longer random and therefore the ben­ efit of randomization is lost (Sussman & Hayward, 2010). As sustained by those advocating for ITT, removal of noncompliant participants might lead to biased estimates by undermining the randomization’s ability to balance the known and unknown confounders between the groups being studied, otherwise compromising the integrity of the method (Polit & Gillespie, 2010). Conducting a sensitivity analysis on a hypothetical trial, Wright and Sim (2003) concluded that biased estimates of effect may occur when devia­ tion is nonrandom or when a large percentage of participants switch treat­ ments or are lost to follow-up, among other reasons. The suggestion here is for researchers to use sensitivity analyses on their data and compare the characteristics of participants who do with the characteristics of those who do not deviate from the trial protocol. Another suggestion is to use alterna­ tive methods in all RCTs, such as inverse probability weighting, g-estimation, and instrument variable estimation to reduce the bias produced by nonadher­ ence and loss to follow-up in as treated or per protocol analyses (Hernán & Hernández-Díaz, 2012). statistical and methodological issues     

83

Sussman and Hayward (2010) suggested incorporating the use of instru­ mental variables (IVs) to manage the treatment contamination as a result of an ITT approach that they called the contamination-adjusted ITT (CA ITT) method. The CA ITT is the method of analysis for RCTs in which all par­ ticipants are analyzed as they were randomized and then the results adjusted for treatment contamination by using an instrumental variable. The IV is an analytical technique that uses a variable associated with the factor under study but not directly associated with the outcome variable or any potential confounders. These authors noted that using an IV approach allows for an assessment of how well the instrument predicts the exposure to outcome then uses the information to understand how the exposure predicts the outcome. In the CA ITT, the RCT is treated as an IV, with treatment assignment as the instrument, then the effect of treatment assignment on outcome observed (ITT) is adjusted by the percentage of assigned participants who actually received the treatment (contamination adjustment). CA ITT improves the accuracy in estimating the size of treatment benefit for a patient who actually receives the treatment. However, Sussman and Hayward recognized that the technique is not frequently used and that it has some limitations when treat­ ment contamination is not collected or properly measured. The technique also assumes that if nonadherents had received the treatment, it would have had the same effect as it did in adherents, which may not necessarily be the case. Therefore, the technique does not account for heterogeneity of treatment effects. CA ITT is not intended for those RCTs in which par­ ticipants are lost due to attrition in follow-up evaluations, which requires a different approach that we discuss below. Sheiner (2002) proposed that what the ITT estimator really generates is an answer for use effectiveness (the causal effect on outcomes of prescrib­ ing the treatment) instead of what should be more important, method effec­ tiveness (the causal effect on outcome of actually receiving the treatment). Therefore Sheiner presented a conceptual framework for thinking about causal estimands and estimators2 for both use and method effectiveness, focusing on the problems posed by deviations from protocol, notably non­ compliance and dropout. In what may look similar to the instrumental vari­ able discussed above, Sheiner proposed that method effectiveness—average outcome difference due to taking treatment versus a control—is an important causal estimand on which to base therapeutic decisions. The ITT estimator of treatment effect validly estimates use effectiveness but only if any miss­ ing data are ignorable, which could be an estimand valid for public policy decisions not method effectiveness. While incorporating information that Causal estimands (population quantities describing causal effects of treatments) are defined on potential data, while estimators are limited to actual (observed) data. 2

84       rivera-medina and caraballo

generates variables that respond to noncompliance behavior as well as respon­ dent behavior into the analyses, the researcher would be able to account for use and method effectiveness, with method effectiveness probably being the most significant for the researcher. In light of the concerns generated by the ITT models, another statistical approach gaining popularity in RCTs is linear mixed models, which may be used when there are multiple measurements of an outcome, as in longitudinal design (Polit & Gillespie, 2010). This approach has been attractive because it allows for accommodating missing values, does not require the same number of measurement for all participants, and has been found to yield relatively few statistical errors when missing outcomes are missing at random. A comprehen­ sive discussion about this and other advanced methods is presented by Byrne (2014) in the Handbook of Multicultural Psychology. THE PROBLEM OF MULTIPLICITY OF TESTING RCTs, although generally oriented to affect one primary outcome, as could be the case for a treatment for depression symptoms, usually also have an effect on other secondary areas as well that could be of interest for the researcher such as anxiety, self-concept, suicidal ideations, and internal or external behaviors. Also, a researcher may want to evaluate the effective­ ness of a treatment on various primary outcomes, as could be the case for neuropsychological research in which a cognitive–behavioral treatment may have a simultaneous effect over several cognitive areas. The above examples introduce the problem of multiplicity of testing because each primary or sec­ ondary outcome will require a separate analysis on the same sample. Thus, multiplicity occurs when performing more than one statistical test with a dataset from a given study. When this problem is ignored, the probability of declaring a test significant, when in fact it is due to a chance occurrence, increases; that is, an investigator may conclude that a treatment is better than another in one or several outcomes when in fact there is no statistical difference between them. Multiplicity is also a concern when the analysis of subgroups is of interest. The analysis of subgroups is particularly important when dealing with ECG as is the usual approach in studies considering diverse ethnic groups or heterogeneity within groups such as Hispanics or Asians. In addi­ tion, multiplicity is a worry when testing ad hoc hypotheses (data snoop­ ing) and in longitudinal studies when comparing the means between groups at different time points, as is frequently the case in RCTs that need to demonstrate treatment efficacy between groups at the end of intervention and follow-up assessments. statistical and methodological issues     

85

Primary Objective With More Than One Outcome In an experiment where there is more than one primary outcome, it is important to control the overall Type I error (the probability of rejecting the null hypothesis being true) to some reasonable value aE (the familywise error rate). A general rule of thumb is that the aE would be approximately 0.05, or 5%; however, when more than one primary or secondary outcome is analyzed, aE may increase above the nominal level desired. For example, when conducting n independent statistical test, each with a nominal Type I error (a), the overall error rate is given by 1 - (1 - a)n. Thus, if five tests are performed, each at the 0.05 level, the overall error rate is 1 - (1 - 0.05)5 = 0.23; that is, there is a chance of approximately 1 in 4 that one of the five hypotheses is rejected by chance alone and a significant treatment effect is declared erroneously. One approach to control the overall error rate in these cases is to use the Bonferroni correction to adjust the significance level of each hypothesis tested to a lower value a = aE/n (Pocock, 1997). For example, for a aE of 0.05, if two tests are performed, the significance level for each test needs to be adjusted to a = 0.05/2 = 0.025. For five tests, a = 0.05/5 = 0.01, and so on. This is a conservative approach, however, as the Bonferroni correction assumes that the tests are independent and it is usually the case that tests per­ formed with the same sample are not independent. In the attempt to control for Type I error, the probability for Type II error (the probability of declaring a treatment noneffective when in fact the opposite is true) increases. Thus, higher Type II errors are expected. A second approach to deal with the problem of multiplicity, when there is more than one primary outcome, is to create a composite variable; thus, only one statistical test needs to be performed. In this approach, the outcome variables are standardized (mean = 0, standard deviation = 1), with the compos­ ite variable being a linear combination of them, that is, each standardized vari­ able is multiplied by an appropriate weight before adding the variables together (Pocock, 1997). Consider, for example, the situation in which a researcher is interested in evaluating participants’ severity at the end of the intervention. The researcher has two measures: the Symptoms Checklist 36 (SCL-36) to assess general psychopathology and the Beck Depression Inventory (BDI) to assess depression symptoms. In this situation one could create a composite of severity using both measures. For those two outcomes, equal weights of [2(1 + r)]–1/2, where r is the correlation between the two outcome variables (SCL-36 and BDI), can be used. Analysis of the composite variable should be followed by individual analyses of its components, treating them as secondary outcomes.

86       rivera-medina and caraballo

A caveat with composite outcomes, however, is that their interpretation may be obscured if the components measure conceptually different outcomes such as mood disorders and family dysfunction. Furthermore, if the magni­ tude of the treatment effects upon the individual outcomes is dissimilar, nonsignificant effects in some of the components may mask a significant effect on another component (Freemantle, Calvert, Wood, Eastaugh, & Griffin, 2003). Secondary Outcomes The procedures for the analysis of secondary outcomes should be made part of the research protocols, along with that of the primary outcomes, and reported as such. In that case the number of statistical tests can be known in advance, and control of the Type I error is possible using a procedure similar to the one described above. When one is testing for secondary outcomes, the same significance level need not be assigned to all the tests; a greater portion can be assigned to one secondary outcome and the remaining to the others. Thus, in a study with three secondary outcomes, such as anxiety, self-concept, and suicidal ideations, suicidal ideations can be tested at a significance level of 0.03, while anxiety and self-concept can be tested at the 0.01 level. In this way an overall familywise error of no more than 0.05 can be achieved. Subgroups As part of the analysis from an RCT with an ECG, it would be neces­ sary to perform analyses considering ethnicity or other specific character­ istics of the subjects, such as gender or comorbidity. If planned in advance, these analyses should be made part of the research protocol and treated as indicated above for secondary outcomes. However, analysis of subgroups is particularly problematic, not only because of the possible multiplicity prob­ lem but also because of the fact that the lower sample size may result in a test with low power. In fact, the probability of retaining the null hypothesis when it is false increases. Thus, if subgroup analyses are planned in advance, determination of an adequate sample size for the subgroup needs to be per­ formed. We discuss this issue in more detail in the Power Considerations section below. Data Snooping Frequently in RCTs, there may be variables or outcomes that could be of interest, besides the disorder it is intended to impact, such as family, cognitive, and interpersonal factors, or process outcomes that do not necessarily become

statistical and methodological issues     

87

part of the primary or secondary aims. Data snooping, also called data dredging or fishing expeditions, refers to the process of performing unplanned statistical tests on data with the idea of exploring possible relationships between variables. As such, this procedure serves only to generate new hypotheses that will need confirmation on subsequent experiments; therefore, no Bonferroni adjustments are warranted. There is no intrinsic problem with data snooping as long as the investigator is aware that (a) the chance of finding false positives increases with the number of tests being performed, so that some significant results may be spurious and thus (b) the results obtained cannot be given the same status as the results of the analyses for the primary or secondary outcomes. Tests at Multiple Time Points Longitudinal RCTs typically look at the effect of treatment as it is administered during a given time frame and may also look at long-term treat­ ment efficacy, characterized by measuring one or more outcome variables at different points in time. In fact, the more the time points considered in the design, the better the alternatives for statistical strategies to help explain the expected participant’s improvement across time. Nonetheless, in many studies situations may arise in which the investigator decides to test for treat­ ment effects at midpoints not initially planned. For example, it may be that, due to dropouts, the number of data points at end time is less than optimal, so the decision is to test for significant differences between treatments at earlier times. Similarly, a researcher may be interested in testing for signifi­ cant treatment effects at midpoints where the differences in the group means are larger than at the end point. If the analyses at some intermediate time are not preplanned, then they will fit the category of data snooping and any significant difference should be treated as a tentative result. If all the tests are planned in advance, correction due to multiple testing can be implemented, as discussed above. POWER CONSIDERATIONS In planning a treatment study, one of the biggest concerns is the deter­ mination of the sample size. As previously mentioned, the NIH 2001 policy brings additional concerns when working with ECGs. If too few cases are studied, as could be the case when including ethnocultural participants rep­ resented in the sample in proportion to its size in the U.S. population, there is the risk of not finding a treatment effect when in fact there is one. On the other hand, too many cases increase the cost of the study and may also put an unnecessary burden on the additional participants who are not needed for the 88       rivera-medina and caraballo

study, unless there is a theoretical question related to differential treatment effects related to culture, race, and/or ethnicity. Retaining the null hypothesis of no treatment effect when in reality there is one is a Type II error (b). Power (P) is defined as 1 - b and thus rep­ resents the probability of making the correct decision when rejecting the null hypothesis, which is what is expected in an RCT. Power (and b) depends on the magnitude of the treatment effect size (d), the significance level or Type I error (a) as previously discussed, and the sample size. For a cross-sectional RCT with two groups and a continuous outcome, the relationship between these quantities is given by the following formula, which is used to determine the sample size (n) required for the study: n≥

2 z1−α 2 + z1−β  d2

2

,

where z1–a/2 and z1–b are the quantiles for the normal distribution at 1 - a/2 and 1 - b, respectively, and d is a measure of the effect size. The effect size is the ratio of the difference between the means of the two groups on the out­ come variables, divided by its standard deviation. It is customary to use the standard deviation of the control group, if there is one, or a pooled3 estimate of the standard deviation of the two groups. Effect sizes around 0.2 are consid­ ered small; around 0.5, medium; and around 0.8 or more, high (Cohen, 1988, 1992). If previous estimates of the standard deviation of the outcome measure are on hand, the effect size can be computed by dividing the expected differ­ ence in treatment means by this estimate. In calculating sample size values, a and b are set usually at 0.05 and 0.20 (or 0.10). For example, to test a non­ directional null hypothesis in a cross-sectional study with a = 0.05 (z1–a/2 = 1.96) and P = 0.80 (b = 0.20, z1–b = 0.84), for effect sizes of 0.2, 0.5, and 0.8 we get required sample sizes of 393, 63, and 25 per group, respectively. This is significant in RCTs with ECGs because the smaller the effect size, the larger the sample size required per group conditions. If one is interested in estimat­ ing differences considering ethnicity, sample size matters, contrary to what is suggested by the NIH 2001 policy. For longitudinal studies, Fitzmaurice, Laird, and Ware (2004) presented an approximate formula for calculating the sample size that is similar to that for a two-group cross-sectional study. If one assumes a linear relationship

3Rosenthal (1994) maintained that although Hedges and Olkin (1985) suggested that pooled standard deviations tend to provide better estimates of effect sizes, when the standard deviations from the two different conditions differ greatly from each other, it would be reasonable to use the control group standard deviation because of the higher probability that the variability in the experimental condition could be too large or small compared with the control group.

statistical and methodological issues     

89

between time and treatment effect and that measures are conducted at sev­ eral times that are not necessarily equally spaced, the treatment effect can be assessed by the difference between the slopes of the treatment and control groups divided by an error term that considers the variance of the measures, the sum of squares of the time periods, and the variance of the standard devia­ tion of the regression coefficient for the time variable. Clearly, a challenge with using this formula is to have preliminary data to estimate the standard deviations. A more general approach for calculating sample size is through Monte Carlo simulations (Muthén & Muthén, 2002). In this approach, once a statis­ tical model is specified, by identifying the independent variables that impact the outcomes and how those independent variables are related, population values for each parameter are selected. As in the previous methods discussed, the parameter values are obtained from either theory or previous research. Through a systematic search, sample sizes are selected until the desired power is attained for the parameters of interest and prespecified criteria for biases of parameters and standard errors and coverage are met. An advantage of this method is that sample size for a desired power level can be explored under conditions of missingness and nonnormality. In RCTs with ECGs, given that the power, as the effect size, is also affected by the variability in the outcomes, if diversity in ECGs is not consid­ ered in the analysis, it may not be possible to detect treatment effects when the ECGs considered in the study are too heterogeneous. For example, it is a common practice to put African Americans, Latinos, Asians Americans, and Native Americans into one group to establish a comparison between minority and nonminority groups. Another instance is when Mexicans, Puerto Ricans, and Dominicans are collapsed into a Hispanic category. Nonetheless, if the means for the observed variable are significantly different between groups, collapsing the subgroups into one category will have the effect of inflating the variance of the observed variable for this category. As the magnitude of a treatment effect is inversely proportional to the variance, the inflated vari­ ance may shrink the treatment effect to the point of not being statistically significant. Thus, a conclusion of no treatment effect for the category (say, Hispanics or minorities) may be reached, when in fact the treatment is effec­ tive for some or all of the subgroups. Lloyd Rogler called attention to cultural insensitivity in the field and suggested greater sensitivity to diversity in ECGs (1999; Rogler et al., 1987). This tendency to collapse ECGs into fewer cat­ egories to increase the sample size typically is the result of an unsuccessful sampling strategy. The literature amply discusses some strategies to improve not only participant recruitment but also participant retention throughout the study with ECGs (Bernal et al., 2014; Janson, Alioto, Boushey, & Asthma Clinical Trials Network, 2001; Knight, Roosa, Calderón-Tena, & González, 90       rivera-medina and caraballo

2009; Sangi-Haghpeykar et al., 2009) as well as some institutional barriers that should be taken into consideration (Joseph & Dohan, 2009). We suggest a careful review of this literature to successfully accomplish the study goals at each step of the project. IDENTIFYING MISSING DATA MECHANISMS AND HOW TO HANDLE MISSING DATA CASES As noted above, in designing a treatment research, careful attention is given to determining the number of participants to have adequate power in the study. In practice, once the study is being carried out, some participants may not be able to attend one or more treatment sessions and/or to provide outcome data. It may even be the case that for some participants information collected is misplaced or not entered because of administrative problems not related to treatment administration. In these situations we encounter a miss­ ing data problem. The presence of missingness affects the power of the study, as the actual sample size may be less than the value agreed upon in the design phase of the study for some or all of the analyses to be performed. Moreover, missing data can be a source of bias in a clinical trial, especially in situa­ tions in which the missingness is related to the treatment effect. Missing data present challenges to the analyst who needs to understand its sources and provide reasonable options to data analysis. A failure to clearly understand the problem may lead to parameter estimates that are biased and possibly to unwarranted conclusions about treatment effectiveness. A common misconception is to think that a small percentage of miss­ ing data is acceptable and does not appreciably affect the treatment effect. In reality, the impact of missing data is determined by the particular question asked, the information in the observed data, and the reason or mechanism for the missing data (Carpenter & Kenward, 2007). For example, in studies of rare events, such as suicide, a small number of missing data may significantly alter the estimates of the event rates. Treatment effects can be biased if sub­ jects withdraw more from a one-treatment condition than another. Similarly, standard error estimates may be biased if the remaining subjects are more homogeneous on the outcome measures. Missing data can occur in treatment research in innumerable ways but can be categorized into three types (Little & Rubin, 2002): (a) missing com­ pletely at random (MCAR), (b) missing at random (MAR), and (c) missing not at random (MNAR). This categorization provides general guidelines on how to account for missing data when performing statistical analyses. A brief description of these categories follows and although it is recognized that an accu­ rate description of all potential causes of missingness is not possible (Schafer & statistical and methodological issues     

91

Graham, 2002), knowing its possible distribution is helpful to determine how to handle it. Missing Completely at Random Data are MCAR if the missingness is a result of a process that is inde­ pendent of the observed values, therefore independent of any of the indepen­ dent (X) or dependent (Y) variables. For a longitudinal RCT for adolescents with symptoms of depression, the missingness of a value for depression symp­ toms at one time point is not associated with the treatment conditions or any other independent variable such as sex, comorbidity, or parent’s marital status or with the dependent variable itself such as severity of depression symptoms. Some authors (Allison, 2012) even believe it could be MCAR if the missingness in depression symptoms (Y) is related to missingness in any other variable. According to Graham, Cumsille, and Shevock (2013), data could also be considered MCAR if the cause of missingness is not correlated with the variable that contains the missing values. As these authors explain, the significance of this mechanism of missingness is that once the cause of missingness is omitted from the missing data model it has no estimation bias. Missing at Random Data are MAR when the missingness in a variable Y is a function of some other observed variable but independent of Y, allowing the proba­ bilities of missingness to depend on observed data but not on missing data (Schafer & Graham, 2002). Therefore, the cause of missingness is correlated with variables that have the missing data, but these variables have been measured and thus are available for consideration in the missing data model, which corrects for all biases associated with them (Graham et al., 2013). Suppose that in the above example the treatment condition and sex of the respondents are correlated with the missingness of depression symptoms at one point during the follow-up. Considering treatment condition and sex in the missing data model for imputation of depression symptoms at the follow-up would be necessary to correct bias estimation. Several authors have warned that the concept of randomness brings confusion among psy­ chologists in the use of the term MAR. Schafer and Graham (2002) advised how the term random has a different meaning for statisticians (a probabilis­ tic process) and for psychologists (unpredictable or extraneous). Therefore, because this concept comes from statisticians, the investigator should be aware of its proper use because it is really not random or an ignorable non­ response, as some describe it. The term random applies once the cause of missingness is identified. 92       rivera-medina and caraballo

Missing Not at Random For MNAR, missingness in Y is a function of Y itself; the cause of miss­ ingness is correlated with the variable containing the missing data, but (unlike MAR) the cause has not been measured or is not available for consideration in the missing data model (Graham et al., 2013). A more concrete example would be a randomized experiment for the treatment of depression in which patients with high levels of depression have a high probability of dropping from the study, so posttreatment values of depression are not available for them. These three missing data mechanisms may cause the two general kinds of missing data: item nonresponse or wave nonresponse (Graham et al., 2013). Testing for Missing Completely at Random Testing for the mechanism of missingness is important to determine the best procedure to deal with it. However, sometimes researchers have no con­ trol over the missingness, so testing for MAR is not feasible to obtain precise information. MAR becomes more of an assumption, though in psychologi­ cal studies departures from MAR are probably not that serious (Schafer & Graham, 2002). Nevertheless, one actually can test for MCAR, which is the missing mechanism best suited for older data editing procedures such as case deletion or single imputation. If researchers include in the data set for a RCT not only data for the independent variable of interest (Y) but also data from some auxiliary variable (Xi), such as demographic information (e.g., age, gen­ der, sex), it may be possible to test for MCAR. A simple test for determining whether missingness is MCAR is to perform a t test or chi-square test using missingness in Y (0 = not missing, and 1 = missing) as a factor and the values of Xi as dependent variables. A significant test statistic will be evidence that missingness is not MCAR and the investigator may proceed to handle miss­ ing data methods that assume MAR. How to Deal With Missingness Discussing in detail the methods to handle missing data is beyond the scope of this chapter. Instead, the investigator should consult the primary sources mentioned throughout. Nevertheless, we would like to provide an overview of some of the methods to deal with missingness that have been proposed in the literature. Among them are analysis of complete cases, last observation carried forward (LOCF) and its variations, single imputation, multiple imputations, and likelihood-based analyses. It should be stated right from the start that none of these approaches are satisfactory when data are MNAR or nonignorable. statistical and methodological issues     

93

An analysis of complete cases means ignoring cases with missing values and performing the statistical analyses with those cases that have no missing values. This method, among the oldest for missing data, has several draw­ backs. First, it is generally valid only under MCAR, so if data are MNAR or MAR, parameter estimates generally will be biased (Schafer & Graham, 2002). Second, even under the MCAR missingness mechanism, this method is still problematic, mostly in multivariate analyses involving several items, where the researcher would see the sample size decreasing at a very fast rate. It may be useful when a very small proportion of the sample is missing; how­ ever, researchers should verify that the discarded cases do not influence the results (Schafer & Graham, 2002). The smaller sample size will result in a loss of statistical power. A biased estimate means that the obtained value for the statistic will be either too big or too small compared with the population value. This bias will be particularly problematic in longitudinal or repeated measures designs commonly used in RCTs (Ibrahim & Molenberghs, 2009). LOCF is a strategy used in repeated measures designs, when a subject provides data up to a certain measurement occasion and then drops from the study, and thus no more data are available for the subject. The idea is to impute, or fill, the missing data with the last observed value for the subject and proceed with the statistical analyses. Parameters, such as means or regres­ sion coefficients, estimated using this method will be biased, in general. In addition, making the imputed value equal to a previous value has the effect of reducing variance estimates for the parameters so that their standard errors will be biased—this time, downward. Other forms of imputation include mean substitution (substituting a missing value by the average of the variable across cases), regression imputa­ tion (predicting the value of a missing data using a regression model), and hot-deck imputation (missing values replaced by values from similar cases in the sample). Mean substitution, available in some statistical software, may look reasonable to psychologists given that most of the outcome variables are constructs created by several items. In item nonresponse missingness, there may be the idea that the items are exchangeable and there were equally reli­ able measures of the unidimensional construct or trait; however, this method is not theoretically driven and introduces bias under MCAR (Schafer & Graham, 2002). Regression imputation or single imputation could be more efficient than case deletion or mean substitution because it allows for the retention of the full sample and prevents the power loss that could result from a decrease in sample size. Nonetheless, the several single imputation methods may reduce variances or distort covariances and intercorrelations between variables, producing biased estimates for many parameters under any missing­ ness mechanisms. Any of the above methods will fail if the added variance due to nonresponse is not properly accounted for (Little & Rubin, 2000). 94       rivera-medina and caraballo

Graham and colleagues (2013) suggested to never use these procedures even for small rates of missingness. Literature has consistently shown that multiple imputation (MI) and maximum likelihood (ML) are acceptable procedures to handle missing data (Allison, 2012; Graham et al., 2013; Schafer & Graham, 2002; Wayman, 2003). Specifically in MI, missing values are predicted using a statistical model that incorporates existing values from other variables, resulting in a full data set. This process is repeated a number of times, usually from three to five, so that a collection of complete datasets is obtained (Wayman, 2003). Standard statistical analyses are then performed on each data set replicate. The parameter estimates obtained with each replicate are then combined to produce one overall analysis. The principal feature of MI is that the miss­ ing values are predicted from observed values, with random noise added to preserve a correct amount of variability in the imputed data sets (Schafer & Graham, 2002). MI has become a popular approach to imputing missing val­ ues, so many software packages have the procedures to perform it, although NORM, available free for Windows (Schafer, 1997), seems to be the better choice (Graham et al., 2013) to handle diverse functions. A different approach to missing values consists of computing ML esti­ mates for the parameters. In this method, an overall likelihood function is constructed that considers the likelihood functions for observations with com­ plete data and observations with incomplete data. Once this overall function is constructed, the logarithm of the overall likelihood function is maximized and parameter estimates are obtained. In a way, ML estimates are obtained by skipping the missing values (Allison, 2012). As compared with MI, the ML approach is more efficient, easier to implement, and thus more attractive for the practicing researcher. An important limitation of this method is that the estimates obtained can be heavily biased for small samples. This procedure has become popular for those researchers prompted to integrate mediator or mod­ erator variables in their intervention and longitudinal models. The statistical software Mplus (Muthén & Muthén, 1998) is one of the most useful programs to implement ML. Ten years ago, Schafer and Graham (2002) concluded that researchers should, when possible, apply the likelihood procedures or paramet­ ric MI, which would be appropriate under MAR assumptions. At present, this recommendation still holds true (Graham et al., 2013). CONCLUSION First, it is important to acknowledge that it is not possible to provide an exhaustive discussion of all of the issues of concern that arise while plan­ ning an RCT. We focused our discussion on some of the more fundamental statistical and methodological issues     

95

statistical and methodological issues that investigators and clinicians should consider in this process. A central concern is the goal of the RCT and the primary hypothesis of the study in choosing the primary outcomes in terms of efficacy. Several strategies have been suggested to manage the ITT issue. As explained in the ITT and missing data sections, the ITT strategy does not appear to be sustainable on either theoretical or statistical grounds. We sug­ gest considering instead an appropriate approach to handle missing data, such as ML or MI, according to the nature of the missingness, sample size (espe­ cially for ML), and the statistical software available. Secondary outcomes, subgroups, and data snooping are important issues to consider in an RCT and they may be incorporated as long as one takes the appropriate methods to handle them. Finally, contrary to what is suggested by the NIH 2001 policy, sample size does matter in RCTs, especially when considering ECGs. If a small sample size is considered, as suggested by this policy, it will directly impact the variability in the study, hence affecting effect size and power. It is also important to not overlook the impact that collapsing subgroups have (as a result of having too small sample sizes within subgroups) that may directly pro­ duce biased parameters, thus making it impossible to detect treatment effects when in fact the ECGs considered in the study are too heterogeneous. In the discussion of categorical versus dimensional outcomes, specifically for diagno­ sis, Kraemer (2007) suggested the interaction that ideally should take place between statisticians and nosologists when discussing DSM–5 disorders. We recommend that the same kind of interaction should occur between clinicians and statisticians. Ideally, greater interaction between these two professions should contribute to designing well-thought-out RCTs and should take place from the beginning when the study is being conceptualized. REFERENCES Allison, P. D. (2012). Handling missing data by maximum likelihood (Paper 3122012). Haverford, PA: Statistical Horizons. Retrieved from http://www. statisticalhorizons.com/wp-content/uploads/MissingDataByML.pdf Altman, D. G. (1996). Better reporting of randomised controlled trials: The CONSORT statement. British Medical Journal, 313, 570–571. http://dx.doi.org/ 10.1136/bmj.313.7057.570 American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Atkins, D. C., Bedics, J. D., McGlinchey, J. B., & Beauchaine, T. P. (2005). Assessing clinical significance: Does it matter which method we use? Journal of Consulting and Clinical Psychology, 73, 982–989. http://dx.doi.org/10.1037/ 0022-006X.73.5.982

96       rivera-medina and caraballo

Bauer, S., Lambert, M. J., & Nielsen, S. L. (2004). Clinical significance methods: A comparison of statistical techniques. Journal of Personality Assessment, 82, 60–70. http://dx.doi.org/10.1207/s15327752jpa8201_11 Bernal, G., Cumba-Avilés, E., & Rodríguez-Quintana, N. (2014). Methodological challenges in research with ethnic, racial, and ethnological groups. In F. T. L. Leong (Ed.), APA handbook of multicultural psychology: Vol. 1. Theory and research (pp. 105–123). http://dx.doi.org/10.1037/14189-006 Bernal, G., Trimble, J. E., Burlew, A. K., & Leong, F. T. L. (2003). Introduction: The psychological study of racial and ethnic minority psychology. In G. Bernal, J. E. Trimble, A. K. Burlew, & F. T. L. Leong (Eds.), Handbook of racial and ethnic minority psychology (pp. 1–12). Thousand Oaks, CA: Sage. Byrne, B. M. (2014). Recent statistical advances. In F. T. L. Leong (Ed.), APA handbook of multicultural psychology: Vol. 1. Theory and research (pp. 125–140). Washington, DC: American Psychological Association. Carpenter, J. R., & Kenward, M. G. (2007). Missing data in randomized controlled trials—A practical guide. Retrieved from http://missingdata.lshtm.ac.uk/ downloads/rm04_jh17_mk.pdf Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. http://dx.doi. org/10.1037/0033-2909.112.1.155 Demjaha, A., Morgan, K., Morgan, C., Landau, S., Dean, K., Reichenberg, A., . . . Dazzan, P. (2009). Combining dimensional and categorical representation of psychosis: The way forward for DSM–V and ICD–11? Psychological Medicine, 39, 1943–1955. http://dx.doi.org/10.1017/S0033291709990651 Dutta, R., Greene, T., Addington, J., McKenzie, K., Phillips, M., & Murray, R. M. (2007). Biological, life course, and cross-cultural studies all point toward the value of dimensional and developmental ratings in the classification of psychosis. Schizophrenia Bulletin, 33, 868–876. http://dx.doi.org/10.1093/schbul/sbm059 Fitzmaurice, G. M., Laird, N. M., & Ware, J. H. (2004). Applied longitudinal analysis. Hoboken, NJ: Wiley. Freemantle, N., Calvert, M., Wood, J., Eastaugh, J., & Griffin, C. (2003). Composite outcomes in randomized trials: Greater precision but with greater uncertainty? Journal of the American Medical Association, 289, 2554–2559. http://dx.doi.org/ 10.1001/jama.289.19.2554 Graham, J. W., Cumsille, P. E., & Shevock, A. E. (2013). Methods for handling miss­ ing data. In J. A. Schinka & W. F. Velicer (Eds.), Handbook of psychology: Vol. 2. Research methods in psychology (pp. 109–141). New York, NY: Wiley. Hayes, S. C., & Haas, J. R. (1988). A reevaluation of the concept of clinical signifi­ cance: Goals, methods, and methodology. Behavioral Assessment, 10, 189–196. Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic Press. statistical and methodological issues     

97

Helzer, J. E., Bucholz, K. K., & Gossop, M. (2007). A dimensional option for the diagnosis of substance dependence in DSM–V. International Journal of Methods in Psychiatric Research, 16(1, Suppl. 1), S24–S33. http://dx.doi.org/10.1002/mpr.210 Helzer, J. E., Kraemer, H. C., & Krueger, R. F. (2006). The feasibility and need for dimensional psychiatric diagnoses. Psychological Medicine, 36, 1671–1680. http://dx.doi.org/10.1017/S003329170600821X Hernán, M. A., & Hernández-Díaz, S. (2012). Beyond the intention-to-treat in comparative effectiveness research. Clinical Trials, 9, 48–55. http://dx.doi. org/10.1177/1740774511420743 Ibrahim, J. G., & Molenberghs, G. (2009). Missing data methods in longitudinal studies: A review. Test, 18, 1–43. http://dx.doi.org/10.1007/s11749-009-0138-x Jacobson, N. S. (1988). Defining clinically significant change: An introduction. Behavioral Assessment, 10, 131–132. Jacobson, N. S., & Revenstorf, D. (1988). Statistics for assessing the clinical signifi­ cance of psychotherapy techniques: Issues, problems, and new developments. Behavioral Assessment, 10, 133–145. Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12–19. http://dx.doi.org/10.1037/0022-006X.59.1.12 Janson, S. L., Alioto, M. E., Boushey, H. A., & Asthma Clinical Trials Network. (2001). Attrition and retention of ethnically diverse subjects in a multicenter randomized controlled research trial. Controlled Clinical Trials, 22(6, Suppl. 1), S236–S243. http://dx.doi.org/10.1016/S0197-2456(01)00171-4 Joseph, G., & Dohan, D. (2009). Recruiting minorities where they receive care: Institutional barriers to cancer clinical trials recruitment in a safety-net hos­ pital. Contemporary Clinical Trials, 30, 552–559. http://dx.doi.org/10.1016/ j.cct.2009.06.009 Knight, G. P., Roosa, M. W., Calderón-Tena, C. O., & González, N. A. (2009). Methodological issues in research on Latino populations. In F. A. Villarruel, G. Carlo, J. M. C. Grau, M. Azmitia, N. J. Cabrera, & T. J. Chahin (Eds.), Handbook of U.S. Latino psychology: Developmental and community-based perspectives (pp. 45–62). Thousand Oaks, CA: Sage. Kovacs, M. (1992). Children’s Depression Inventory (CDI). New York, NY: MultiHealth Systems. Kraemer, H. C. (2007). DSM categories and dimensions in clinical and research contexts. International Journal of Methods in Psychiatric Research, 16(1, Suppl. 1), S8–S15. http://dx.doi.org/10.1002/mpr.211 Kraemer, H. C., Noda, A., & O’Hara, R. (2004). Categorical versus dimensional approaches to diagnosis: Methodological challenges. Journal of Psychiatric Research, 38, 17–25. http://dx.doi.org/10.1016/S0022-3956(03)00097-9 Kraemer, H. C., Shrout, P. E., & Rubio-Stipec, M. (2007). Developing the Diagnostic and Statistical Manual V: What will “statistical” mean in DSM–V? Social

98       rivera-medina and caraballo

Psychiatry and Psychiatric Epidemiology, 42, 259–267. http://dx.doi.org/10.1007/ s00127-007-0163-6 Lawlor, D. A., & Hopker, S. W. (2001). The effectiveness of exercise as an interven­ tion in the management of depression: Systematic review and meta-regression analysis of randomised controlled trials. British Medical Journal, 322, 763–767. http://dx.doi.org/10.1136/bmj.322.7289.763 Linde, K., Ramirez, G., Mulrow, C. D., Pauls, A., Weidenhammer, W., & Melchart, D. (1996). St John’s wort for depression—An overview and meta-analysis of randomised clinical trials. British Medical Journal, 313, 253–258. http://dx.doi. org/10.1136/bmj.313.7052.253 Little, R.  J., & Rubin, D.  B. (2000). Causal effects in clinical and epidemiological studies via potential outcomes: Concepts and analytical approaches. Annual Review of Public Health, 21, 121–145. http://dx.doi.org/10.1146/annurev.publhealth. 21.1.121 Little, R. J., & Rubin, D. B. (2002). Statistical analysis with missing data. http://dx.doi. org/10.1002/9781119013563 MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D. (2002). On the prac­ tice of dichotomization of quantitative variables. Psychological Methods, 7(1), 19–40. http://dx.doi.org/10.1037/1082-989X.7.1.19 Mazumdar, S., Liu, K. S., Houck, P. R., & Reynolds, C. F., III. (1999). Intent-totreat analysis for longitudinal clinical trials: Coping with the challenge of miss­ ing values. Journal of Psychiatric Research, 33, 87–95. http://dx.doi.org/10.1016/ S0022-3956(98)00058-2 Miranda, J., Nakamura, R., & Bernal, G. (2003). Including ethnic minorities in mental health intervention research: A practical approach to a long-standing problem. Culture, Medicine and Psychiatry, 27, 467–486. http://dx.doi.org/10.1023/ B:MEDI.0000005484.26741.79 Muthén, L. K., & Muthén, B. O. (1998). Mplus users guide. Los Angeles, CA: Authors. Muthén, L. K., & Muthén, B. O. (2002). How to use a Monte Carlo study to decide on sample size and determine power. Structural Equation Modeling, 9, 599–620. http://dx.doi.org/10.1207/S15328007SEM0904_8 National Institutes of Health. (2001). NIH policy and guidelines on the inclusion of women and minorities as subjects in clinical research. Retrieved from http://grants. nih.gov/grants/funding/women_min/guidelines_amended_10_2001.htm Pocock, S. J. (1997). Clinical trials with multiple outcomes: A statistical perspec­ tive on their design, analysis, and interpretation. Controlled Clinical Trials, 18, 530–545. http://dx.doi.org/10.1016/S0197-2456(97)00008-1 Polit, D. F., & Gillespie, B. M. (2010). Intention-to-treat in randomized controlled trials: Recommendations for a total trial strategy. Research in Nursing & Health, 33, 355–368. http://dx.doi.org/10.1002/nur.20386 Rivera, C. L., Bernal, G., & Rosselló, J. (2005). The Children’s Depression Inventory (CDI) and the Beck Depression Inventory (BDI): Their validity as screening statistical and methodological issues     

99

measures for major depression in a group of Puerto Rican adolescents. International Journal of Clinical and Health Psychology, 5, 485–498. Rogler, L. H. (1999). Methodological sources of cultural insensitivity in mental health research. American Psychologist, 54, 424–433. http://dx.doi.org/10.1037/ 0003-066X.54.6.424 Rogler, L. H., Malgady, R. G., Costantino, G., & Blumenthal, R. (1987). What do culturally sensitive mental health services mean? The case of Hispanics. American Psychologist, 42, 565–570. http://dx.doi.org/10.1037/0003-066X.42.6.565 Rosenthal, R. (1994). Parametric measures of effect size. In H. M. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis (Vol. 236, pp. 232–243). New York, NY: Russell Sage Foundation. Royston, P., Altman, D. G., & Sauerbrei, W. (2006). Dichotomizing continuous pre­ dictors in multiple regression: A bad idea. Statistics in Medicine, 25, 127–141. http://dx.doi.org/10.1002/sim.2331 Sangi-Haghpeykar, H., Meddaugh, H. M., Liu, H., & Grino, P. (2009). Attrition and retention in clinical trials by ethnic origin. Contemporary Clinical Trials, 30, 499–503. http://dx.doi.org/10.1016/j.cct.2009.06.004 Schafer, J. L. (1997). Analysis of incomplete multivariate data. http://dx.doi.org/ 10.1201/9781439821862 Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7, 147–177. http://dx.doi.org/10.1037/1082-989X.7.2.147 Shaver, J. P. (1993). What statistical significance testing is, and what it is not. Journal of Experimental Education, 61, 293–316. http://dx.doi.org/10.1080/00220973. 1993.10806592 Sheiner, L. B. (2002). Is intent-to-treat analysis always (ever) enough? British Journal of Clinical Pharmacology, 54, 203–211. Streiner, D. L. (2002). Breaking up is hard to do: The heartbreak of dichotomizing continuous data. The Canadian Journal of Psychiatry, 47, 262–266. Sussman, J. B., & Hayward, R. A. (2010). An IV for the RCT: Using instrumental variables to adjust for treatment contamination in randomised controlled trials. British Medical Journal, 340, c2073. http://dx.doi.org/10.1136/bmj.c2073 Thompson, B. (1996). AERA editorial policies regarding statistical significance testing: Three suggested reforms. Educational Researcher, 25(2), 26–30. Thompson, B. (2002). Statistical, practical, and clinical: How many kinds of signif­ icance do counselors need to consider? Journal of Counseling & Development, 80, 64–71. http://dx.doi.org/10.1002/j.1556-6678.2002.tb00167.x Troeung, L., Egan, S. J., & Gasson, N. (2013). A meta-analysis of randomised placebocontrolled treatment trials for depression and anxiety in Parkinson’s disease. PLoS ONE, 8(11), e79510. http://dx.doi.org/10.1371/journal.pone.0079510 Wang, H., Qi, H., Wang, B. S., Cui, Y. Y., Zhu, L., Rong, Z. X., & Chen, H. Z. (2008). Is acupuncture beneficial in depression: A meta-analysis of 8 randomized

100       rivera-medina and caraballo

controlled trials? Journal of Affective Disorders, 111(2–3), 125–134. http://dx.doi. org/10.1016/j.jad.2008.04.020 Wayman, J. C. (2003, April). Multiple imputation for missing data: What is it and how can I use it? Paper presented at the Annual Meeting of the American Educa­ tional Research Association, Chicago, IL. Retrieved from http://www.csos.jhu. edu/contact/staff/jwayman_pub/wayman_multimp_aera2003.pdf Weersing, V. R. (2005). Benchmarking the effectiveness of psychotherapy: Program evaluation as a component of evidence-based practice. Journal of the American Academy of Child and Adolescent Psychiatry, 44, 1058–1062. Widiger, T. A., & Samuel, D. B. (2005). Diagnostic categories or dimensions? A question for the Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition. Journal of Abnormal Psychology, 114, 494–504. http://dx.doi.org/10.1037/ 0021-843X.114.4.494 Wright, C. C., & Sim, J. (2003). Intention-to-treat approach to data from random­ ized controlled trials: A sensitivity analysis. Journal of Clinical Epidemiology, 56, 833–842. http://dx.doi.org/10.1016/S0895-4356(03)00155-0

statistical and methodological issues     

101

5 STATISTICAL METHODS FOR VALIDATING TEST ADAPTATIONS USED IN CROSS-CULTURAL RESEARCH JOSEPH A. RIOS AND RONALD K. HAMBLETON

With increased globalization, psychologists, educational researchers, and policymakers are becoming more interested in cross-cultural studies to understand similarities and differences in human behavior, to assist in contributing to social goals, and to resolve societal issues (Hambleton, 2005; Oakland, 2005). However, in the conduct of cross-cultural research, numerous methodological issues, if not resolved, can lead to invalid inferences from the data and thus reduce the value of the research (van de Vijver & Matsumoto, 2011). One of the major issues related to cross-cultural research is obtaining tests for cross-lingual populations that produce valid and comparable results. It is thus imperative to employ methodological rigor in establishing measurement equivalence or invariance for all populations of interest (Kankarasˇ & Moors, 2010). Consistent with the theme of this volume, our intent is to support culturally informed evidence-based practices by identifying and describing the types of validity studies that need to be carried out to support http://dx.doi.org/10.1037/14940-006 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

103

cross-cultural research, especially as it relates to the use of test translations and test adaptations. The objective of this chapter is to provide researchers with a basic knowledge of the statistical procedures that can be implemented to evaluate measurement equivalence within a cross-cultural context. Historically, researchers have incorrectly assumed that finding a good translator or two would be sufficient for obtaining equivalent cross-linguistic or cross-cultural tests and surveys. However, Ype Poortinga, one of the foremost cross-cultural psychologists in the world, noted that 80% of the cross-cultural literature before 1990 was probably not worth publishing because of shortcomings with the translation (see Hambleton, Merenda, & Spielberger, 2005). Perhaps the biggest mistake in practice has been the failure to follow up the translation process with a compilation of empirical evidence to support the intended uses of the test scores in these target languages and cultures. In this chapter we have organized many of the statistical methods that have evolved, mainly since 1990, for researchers to use in their work. According to the International Test Commission Guidelines for Translating and Adapting Tests, researchers need to compile both test translation evidence as well as empirical evidence to support their uses of tests in multiple languages and cultures (International Test Commission [ITC], 2010). The chapter is organized around three potential sources of measurement bias in cross-cultural assessment that have been identified by van de Vijver and his colleagues. These sources of bias are construct, method, and item bias (van de Vijver & Hambleton, 1996; van de Vijver & Leung, 1997, 2011; van de Vijver & Poortinga, 1997, 2005; van de Vijver & Tanzer, 1998). Within each section of this chapter, the source of bias is defined and plausible statistical methods for evaluating measurement invariance for each source of bias are presented. When we present statistical methods, we provide a brief description of each procedure, taking into consideration practical issues. Furthermore, references to studies that have employed the specific statistical methods within a cross-cultural context are provided when they are available. For an excellent primer on assessing equivalence of cross-cultural assessments that includes basic concepts, substantive examples, and less statistical jargon, see Chapter 6, this volume. DEFINITIONS OF BIAS, TEST TRANSLATION, AND TEST Adaptation Before we describe the various sources of bias, it is important to first provide a definition of bias within a measurement context. The main concern in defining measurement bias is whether knowledge of an examinee’s group 104       rios and hambleton

membership influences the examinee’s score on the measured variable (e.g., an item, subdomain, or test), given the examinee’s status on the background variables of interest (Millsap, 2011). There are two approaches to evaluating measurement bias within the cross-cultural and cross-linguistic research: judgmental methods and statistical methods. Judgmental methods rely on substantive, cultural, and linguistic experts to provide systematic judgmental evidence related to the accuracy and equivalence of the test adaptation across all language versions (ITC, 2010). However, judgmental evidence is not enough to ensure score equivalence across cultures and languages, as reflected in the ITC Guidelines for Translating and Adapting Tests (ITC, 2010), which state, Test developers/publishers should apply appropriate statistical techniques to (1) establish the equivalence of the different versions of the test or instrument, and (2) identify problematic components or aspects of the test or instrument which may be inadequate to one or more of the intended populations. (pp. 2–3)

One major principle underlying this guideline is that translation equivalence cannot be merely assumed, but instead, it must be empirically examined and documented (American Educational Research Association, American Psychological Association, & National Council on Measurement in Education, 2014). However, as noted by many researchers, such evidence is often lacking in practice (e.g., Rios & Sireci, 2014; van de Vijver & Tanzer, 1998). To evaluate bias from a statistical standpoint, we must investigate the measurement invariance property. Measurement invariance can be conceptualized as the equivalence of measurement properties of the test or survey across populations (e.g., cultural and/or language groups; Millsap, 2011). The meaning seems clear: Measurement invariance follows when an examinee’s group membership adds nothing to the prediction of examinee test and itemlevel performance after conditioning on any relevant background variables. Although this chapter focuses primarily on statistical methods, researchers should be aware that judgmental and statistical procedures should be applied jointly to gain a better perspective on measurement equivalence across populations. For more information on judgmental methods, see Hambleton et al. (2005) as well as Matsumoto and van de Vijver (2011). Perhaps we should note too that we prefer the term test adaptation to the term test translation. The latter term refers to the act carried out by translators, but often making a test accessible and fair in a second language and culture involves much more than a good translation of the words in the test, however well-done the translation is. Perhaps the test directions may need to be revised to make them understandable, sometimes the time limits need to be changed, cultural concepts may need to be changed to make them meaningful, and so on statistical methods for validating test adaptations     

105

(see, e.g., the special issue of Language Testing for many examples of changes that may be needed to make a test fair in target languages and cultures; Hambleton & de Jong, 2003). Test adaptation refers to a whole collection of activities that are carried out to make a test or survey psychologically, linguistically, and semantically equivalent for respondents in multiple languages and cultures (see, e.g., Hambleton, 2005). CONSTRUCT BIAS Construct bias occurs when the studied constructs are nonequivalent across language or cultural groups. This nonequivalence can occur when there is partial overlap in conceptualizing the construct or when the behaviors associated with the construct manifest themselves differentially across cultures (van de Vijver & Hambleton, 1996). For example, the construct of intelligence does not have the same meaning across cultural groups. In Western cultures, the concept of speed has a greater role in the definition and in the assessment of intelligence than it does in Eastern cultures. As a result, the tests implemented do not capture the same underlying dimensions of the construct across groups, which has two implications: The validity of the measurement is lacking, and direct comparisons between samples cannot be made. Before describing the data analytic methods that can be used to evaluate construct bias, we must first define the various levels of construct equivalence. Levels of Measurement Equivalence In evaluating construct bias there are three major levels of equivalence: configural, metric, and scalar equivalence.1 Configural equivalence indicates that the same data configurations or structures of the purported construct are present across cultural or language groups; that is, identical indicators (items) can be used to measure the latent construct of interest across groups. However, this level of equivalence does not imply that the strength of the relationship between the indicators and the latent variable is equal across groups. This form of equivalence is the most basic and necessary condition but does not allow for direct comparisons across groups if met. To make direct comparisons, more restrictive forms of equivalence must be attained.

It should be noted that there are additional levels of equivalence that are more restrictive, such as equivalence of error uniqueness variances (see Dimitrov, 2010). However, for the purposes of directly comparing means across cultural or linguistic groups, the levels of invariance summarized in this chapter are sufficient. 1

106       rios and hambleton

Metric equivalence assumes both configural invariance and equivalent strengths between the indicators and latent variable (factor loadings). More specifically, metric equivalence denotes equal measurement units of the scale designed to measure the latent construct across cultural or language groups. Within a factor analysis framework, this level of equivalence would suggest that the factor loadings for all items are equal across all groups. This form of equivalence allows for indirect comparisons as the score intervals are equal across groups but the measurement units do not share the same origin of the scale. As a result, direct comparisons of group means are not valid. The most restrictive of the three levels is scalar equivalence. This type of equivalence subsumes both configural and metric equivalence as well as assumes that the scales of the latent construct possess the same origin. In a factor analysis framework, scalar equivalence would require equal intercepts, which would signal the absence of differential item functioning and would allow for direct comparisons of group means (Dimitrov, 2010). Most often, cross-cultural researchers are interested in obtaining this last form of equivalence as comparisons of group means are of utmost importance. However, it should be noted that at times attaining equivalence for all items may be a bit too restrictive for the observed data. When this is the case, researchers can evaluate partial invariance, which can be viewed as a compromise between full and complete lack of equivalence, allowing for analysis of the data with a small number of violations (Kankarasˇ & Moors, 2010). An analysis of partial equivalence involves constraining some of the parameters to be equal across groups, while the remaining parameters are allowed to be freely estimated across groups. There is no generally agreed upon rule regarding acceptable levels of partial invariance. Instead, the researcher must use his or her best judgment, taking into consideration contextual factors in deciding the degree of acceptable invariance across populations for his or her particular study. Statistical Procedures for Evaluating Construct Bias Now that we have discussed the various levels of measurement equivalence, statistical methods for evaluating construct bias can be described. In general, these statistical procedures can be classified into two groups: exploratory and confirmatory methods. Exploratory methods include exploratory factor analysis (EFA) and multidimensional scaling (MDS; Fischer & Fontaine, 2011; Sireci, Patsula, & Hambleton, 2005), and the most popular confirmatory procedure is multiple group confirmatory factor analysis (MGCFA).The paragraphs that follow illustrate the appropriateness of these methods for evaluating construct bias. statistical methods for validating test adaptations     

107

Exploratory Factor Analysis EFA was described at one time as being the most frequently used technique to evaluate construct equivalence for multiple cultural or language populations (van de Vijver & Poortinga, 1991, as cited in Sireci et al., 2005). Overall, this statistical procedure has two primary objectives: to determine the number of common factors that account for the common variance between the indicators and to observe the strength of the relationship between each factor and each indicator. Within a cross-cultural context, the researcher will need to first conduct an EFA for the reference group (i.e., the culture or language for which the original measure was created). Upon obtaining a satisfactory solution for the reference group (i.e., the source language group), the same number of factors should be applied to the data for each remaining cultural or language group (i.e., the target language groups), rotating the factor structure toward that of the reference group. Once this has been done, the congruence between factor loadings across groups can be calculated. As noted by Fischer and Fontaine (2011), there are four congruence measures: the linearity, the proportionality (Tucker’s phi), the additivity, and the identity coefficient. These congruent coefficients can range up to 1, with higher values indicating greater congruence of factor loadings between groups. Van de Vijver and Leung (1997) suggested that values greater than .95 indicate acceptable levels of congruence, whereas values lower than .95 may be indicative of incongruence. If incongruence was found, researchers would need to investigate possible explanations for the observed differences. However, it should be noted that no statistical tests are associated with the congruence measures and the coefficients allow only for pairwise comparisons between two groups for one item. Therefore, if one were interested in evaluating congruence of factor loadings for more than two groups, multiple pairwise comparisons for each item would need to be conducted. For a step-by-step guide on how to conduct an EFA for cross-cultural research with calculations of congruence coefficients, see Fischer and Fontaine (2011). Multidimensional Scaling MDS is an additional exploratory method that has been used quite frequently within cross-cultural research (Cleeland et al., 1996; Collazo, 2005; Robin, Sireci, & Hambleton, 2003; Wolff, Schneider-Rahm, & Forret, 2011). The objective of MDS is to provide a visual representation of the observed similarities among a set of objects (Fischer & Fontaine, 2011). A major advantage of MDS over EFA is that multiple group data can be analyzed simultaneously to determine the structural similarities across groups. This is accomplished by using an individual differences MDS analysis and evaluating the group weights 108       rios and hambleton

to modify the common structure for each group (Sireci et al., 2005). An additional advantage of MDS as an exploratory method is that it requires fewer assumptions than does EFA, with the only requirement being use of ordinal data. However, one must note that the major disadvantage of MDS is that it is solely a descriptive technique; that is, it provides no statistical test to evaluate structural differences across groups (Fischer & Fontaine, 2011), requiring the researcher to rely primarily on visual interpretations for assessing configural equivalence. For a more detailed treatment of MDS, study the research by Borg and Groenen (2005). As a whole, exploratory methods for testing invariance are being replaced by CFA techniques (Little & Slegers, 2005). The major disadvantage of exploratory methods is that they can evaluate only configural invariance, whereas more restrictive and often more informative forms of invariance cannot be assessed. This limitation is severe when making cross-cultural or cross-lingual score comparisons across groups. However, this is not to say that exploratory methods are impractical in assessing configural equivalence. For one, exploratory methods may be especially advantageous during test development when the quality of indicators is unknown, which may occur when researchers are employing a decentering method to test develop­ ment. An additional foreseeable use of exploratory methods would occur when the a priori hypothesis of the factor structure is found to be inappropriate for all groups under study. As a result, exploratory methods may be useful in assisting the researcher to reconceptualize the initial factor structure. In both plausible uses of exploratory methods it is advisable that if sample sizes permit, one should conduct an EFA on one subset of the data and cross-validate the EFA results within a CFA framework on another subset of the data. Conducting both EFA and CFA on the same data brings about concerns of capitalizing on chance. Cross-validation provides a means to assess parameter fluctuations caused by sampling error and is recommended when applying exploratory methods to evaluate configural equivalence across groups. Now that we have discussed some of the major limitations associated with exploratory methods, let us describe multiple group CFA. Multiple Group Confirmatory Factor Analysis MGCFA is a theory-driven method used to evaluate formal hypotheses of parameter invariance across groups (Dimitrov, 2010). MGCFA is advantageous to use when establishing construct comparability because (a) it allows for simultaneous model fitting across multiple groups, (b) various levels of measurement invariance can be assessed, (c) the means and covariances of the latent constructs are disattenuated (controls for measurement error), and (d) direct statistical tests are available to evaluate cross-group differences statistical methods for validating test adaptations     

109

of the estimated parameters (Little & Slegers, 2005). Conducting MGCFA requires four hierarchical steps: (a) establishing a baseline model across groups, (b) testing for configural invariance, (c) testing for metric invariance, and (d) testing for scalar invariance (see Table 5.1). This systematic process is known as sequential constraint imposition because model parameters across groups are allowed to be freely estimated, with greater constraints on the parameters being placed as good model fit for less restricted models is obtained. Comparison of hierarchically nested models can be conducted with two approaches: chi-square difference test and change in comparative fit index (DCFI). The latter approach was developed because the chi-square statistic has been suggested to be highly sensitive to sample size, whereas DCFI has been demonstrated in simulation studies to provide stable performance with various conditions, such as sample size, amount of invariance, number of factors, and number of items (Meade, Johnson, & Braddy, 2008). The DCFI is calculated as ∆CFI = CFI M1 − CFI M0 where CFIM1 is equal to the CFI value obtained for model 1, and CFIM0 is equal to the CFI value obtained for model 0. On the basis of simulation analyses, Cheung and Rensvold (2002) recommended that a DCFI ≤ .01 supports the invariance hypothesis. The MGCFA approach to evaluating construct bias for test adaptations within cross-cultural research has been a very popular method in the literature (e.g., Davidov, 2011; Hattrup, Ghorpade, & Lackritz, 2007; Yen & Tu, 2011). For one, it is relatively straightforward, and there are numerous resources to assist researchers with learning it (e.g., Dimitrov, 2010). An additional advantage associated with this method is that it can be extended to more than two groups. For example, Ariely and Davidov (2011) used MGCFA to evaluate the factor structure of an attitudinal survey across 36 different TABLE 5.1 Steps for Conducting Multiple Group Confirmatory Factor Analysis Step

Invariance model

Parameters constrained

1 2 3

Baseline Configural Metric

None Factor loadings

4

Scalar

Factor loadings and intercepts

x 2diff

dfdiff

DCFI

 x 2metric -x 2configural  x 2scalar -x 2metric

 dfmetric -dfconfigural  dfscalar -dfmetric

 CFImetric -CFIconfigural  CFIscalar -CFImetric

Note.  Invariance is met by a nonsignificant x 2diff value and/or DCFI ≤ .01. CFI = comparative fit index.

110       rios and hambleton

countries. MGCFA is also advantageous because it can be applied in numerous software packages, such as LISREL (Jöreskog & Sörbom, 2001), AMOS (Arbuckle, 2010), EQS (Bentler, 1995), and Mplus (Muthén & Muthén, 2007). As a result, MGCFA appears to be the best option for evaluating construct bias within cross-cultural research. For a step-by-step guide on how to conduct this procedure in Mplus, see Dimitrov (2010). Modest sample sizes, however, can limit the validity of CFA. METHOD BIAS An additional source of measurement bias within the test adaptation procedure is method bias, which involves construct-irrelevant aspects associated with method-related factors (van de Vijver & Poortinga, 2005). Far too often, researchers fail to address potential method bias in their empirical analyses. Examples of method bias include shifts in the meaning of categories of a rating scale across cultures, unintended speededness brought on by the use of longer words introduced by the translations process, and the use of an unfamiliar item format (e.g., multiple choice) in one or more of the cultures. Van de Vijver and Tanzer (1998) identified three distinctive areas of method bias: sample, administration, and test bias. Sample bias can be conceptualized as large variations in characteristics of the cross-cultural sample, which may lead to very different results within and across the respective groups. Van de Vijver and Leung (2011) described various sources of administration bias as differences in the physical conditions of test administration (e.g., noise, group size), ambiguous instructions for examinees, distinctions in examiner experience, and differences in familiarity of the test stimulus. Last, test bias involves participants being unfamiliar with the response procedures required on the test (e.g., multiple choice) as well as differential response styles (e.g., extreme responses on a Likert scale may be more likely for one culture than another or examinee responses may be influenced by one culture’s desire to respond in a socially desirable manner; van de Vijver & Poortinga, 1997). Method bias can be dealt with in two ways: a priori and a posteriori. A priori procedures involve careful selection of samples to account for representativeness across cultures and response bias, extensive training of examiners, and development of an administration protocol. For the purposes of this chapter we focus only on a posteriori methods; however, for a review of a priori methods, see van de Vijver and Leung (2011). In particular, generalizability (G) theory and the multitrait–multimethod (MTMM) approach to test validation are discussed as two statistical techniques for evaluating administration and test bias, respectively. statistical methods for validating test adaptations     

111

Generalizability Theory G theory has become a popular statistical framework for distinguishing and quantifying inconsistencies in observed scores obtained from measurement procedures employed by educational and behavioral scientists. This statistical framework can be viewed as an extension of the classical test theory model [observed score (X) = true score (T) + error (E)], whereby multiple sources of measurement error (E) are simultaneously evaluated through certain analysis of variance (ANOVA) procedures (Brennan, 2001). Such an approach would be advantageous in determining whether factors related to the test administration produced a disproportionate amount of construct-irrelevant variance (variance component estimates) for one group when compared with others. This evaluation can be accomplished by computing variance component estimates for each group separately and then comparing estimates across groups via overlapping confidence intervals. Statistical procedures, such as t tests, ANOVAs, and correlations, are not recommended for making direct comparisons across groups because variance components are not normally distributed and so statistical tests would be flawed. (For further recommendations, see Li & Brennan, 2007.) One example of employing multiple group G theory for evaluating administration bias comes from a study conducted by Solano-Flores and Li (2009). The authors evaluated whether individual and cultural communication style differences were present when cognitive interviews were conducted between raters and examinees from various cultural groups. Although they found that raters provided a small amount of score variation across examinee cultural groups, these researchers did present an excellent example of applying this statistical framework for evaluating administration bias within a cross-cultural context. For a more thorough review of G theory, see Brennan (2001) as well as Shavelson and Webb (1991). Multitrait–Multimethod Campbell and Fiske (1959) proposed the MTMM approach to test validation as a means to evaluate variance in observed scores as derived from different traits and methods. Each measure of a construct (e.g., item, subtest, survey) can be conceptualized as a trait–method unit. For example, a multiple-choice item measuring depression is composed of a trait (depression) and a method (multiple choice). Because variance cannot be calculated from a single unit, multiple trait–method units are required. The relationships between

112       rios and hambleton

the multiple trait–method units can be evaluated to examine convergent and discriminant validity. Evidence of convergent validity is indicated by strong relationships between theoretically similar constructs, regardless of the method used (monotrait–heteromethod). In contrast, discriminant validity is reflected in weak relationships between measures of different constructs when the same (heterotrait–monomethod) or different (heterotrait– heteromethod) methods are used. An attractive feature of the MTMM approach is that it also allows for the evaluation of systematic method variance, which is based on the relationship between measures assessing the same method but different traits. In terms of evaluating test bias in a cross-cultural context, studying method variance between groups can demonstrate whether some cultures are unfamiliar with response procedures presented on the test (e.g., multiplechoice, Likert-type, or constructed-response items), which can be evaluated by conducting multiple group MTMM analyses. However, instead of comparing traditional correlations across groups, which is the case in the original matrix proposed by Campbell and Fiske, measurement invariance can be evaluated via MTMM models in a multiple group CFA framework. This latter approach is more advantageous in that it allows for a separation of trait, method, and error components as well as an empirical test of the model assumptions (Eid et al., 2008). Furthermore, parameter estimates are controlled for measurement error. Eid et al. (2008) suggested that more than 16 CFA models have been developed to analyze MTMM data; however, the two most frequently employed MTMM models are the correlated trait–correlated method (CTCM) model first proposed by Jöreskog (1974) and the correlated trait– correlated uniqueness (CTCU) model developed by Kenny (1976). One of the major disadvantages of the CTCM model is that it often produces improper solutions as a result of underidentification (Marsh & Bailey, 1991). It is therefore suggested that researchers interested in applying MTMM CFA should use the CTCU model. When this analysis is conducted, it would be best to apply sequential constraint imposition. With this approach each parameter is constrained to be equal across groups and is evaluated for model fit separately until this process has been completed for all parameters in the model. Obtaining differential method variance across groups is evidence that the response procedure unfairly contributed to the score variance for one group when compared with another. Although MTMM has been applied within cross-cultural research (e.g., see Marsh & Byrne, 1993), to the best of our knowledge this procedure has not been applied specifically in the test adaptation literature for evaluating test bias.

statistical methods for validating test adaptations     

113

ITEM BIAS At the outset, it is important to distinguish among three terms: item impact, differential item functioning (DIF), and item bias. An observed difference in item performance between groups (e.g., one group has a higher proportion of endorsement for an item than does another group) does not necessarily indicate that an item is biased toward one particular group of examinees. Instead, such an item may reflect true group differences, which is referred to as item impact. In contrast, DIF occurs when distinct groups have different probabilities of correctly endorsing an item, after controlling for the trait measured by the test. DIF is based on the logic that examinees with similar overall ability will have the same probability of item endorsement. If this is not the case, there is evidence that group membership influences one’s item score. However, too often within the test adaptation literature researchers have stated that DIF and item bias are synonymous (see van de Vijver & Poortinga 1997, 2005). Item bias is distinct from DIF in that it comprises an evaluative component (Thissen, Steinberg, & Wainer, 1988). In other words, to say that an item is biased requires both statistical evidence of DIF and expert judgments pointing out construct-irrelevant factors present within the item under consideration (Sireci, 2011). Therefore, the first step in evaluating item bias is to determine whether DIF is present. We would recommend too that a 25-item DIF checklist introduced by Hambleton and Zenisky (2011) be applied first to the adapted test of interest before any data are collected. The goal would be to spot any DIF items. Reducing item problems prior to data collection can reduce criticisms from the field and avoid wasting time and money field-testing problematic items. Of course, when in doubt, items can still be field-tested. The Hambleton–Zenisky Checklist consists of 25 questions that are based on common reasons for items showing DIF. For example, Question 1 asks, “Does the item have the same or highly similar meaning in the two languages?” Numerous statistical procedures have been developed to evaluate DIF; however, for the purposes of this chapter we review a limited number. In general, DIF methodologies can be classified into two groups: (a) observed score and (b) item response theory (IRT) procedures (Camilli, 2006). Within the observed score framework, there are methods based on proportion correct (e.g., delta plot: Angoff, 1982; standardization index: Dorans & Kulick, 1986), contingency tables (e.g., Mantel–Haenszel; Holland & Thayer, 1988), and regression models (e.g., logistic regression; Swaminathan & Rogers, 1990). IRT methods include Lord’s chi-square (Lord, 1980), Raju’s area measure (Raju, 1988, 1990), and the IRT likelihood ratio test (Thissen et al., 1988). As noted by Muñiz, Hambleton, and Xing (2001), one of the major contributing 114       rios and hambleton

factors to the lack of empirical analyses conducted to evaluate flawed items in test adaptation is the perception that DIF methodologies require larger examinee samples than are often available. To address this concern, in this section we focus solely on popular small-sample observed score methods for binary items, but we also provide references for polytomous extensions of the procedures. In describing each method, practical considerations, such as sample size requirements, type of data (dichotomous or polytomous), ability to detect different forms of DIF (uniform or nonuniform), and ease of interpretation for practitioners and stakeholders (i.e., statistical tests, effect sizes, and graphical representations) are considered. Refer to Holland and Wainer (1993), Camilli and Shepard (1994), Camilli (2006), Osterlind and Everson (2009), and Millsap (2011) for detailed descriptions of all DIF procedures currently available to researchers. Delta Plot One of the first methods for evaluating DIF involved comparing p values (the proportion correct for dichotomous items and the mean item score on polytomous items) between reference and focal groups (e.g., the test in the source and target languages, respectively) via correlations and/or graphical analysis. However, there are two major issues with this procedure: (a) DIF is confounded with item impact (Sireci et al., 2005) and (b) p values are on an ordinal scale, which means that the relationship between p values across groups is nonlinear, making the detection of outliers more problematic. To deal with the latter issue, Angoff (1982) proposed the delta plot method, which transforms p values onto the delta scale as follows: ∆ ij = 4z ij + 13 where Dij is equal to the delta value for item i in group j, and zij is equal to the z score for the p value on item i in group j, assuming the trait measured by the test is normally distributed. (So, e.g., if p = .16, then the normalized z = 1.0 and the delta value is 17; if p = .50, then the normalized z = 1.0 and the delta value is 13.) This procedure is advantageous in that it can be applied with small sample sizes (50 in each group; Muñiz et al., 2001), both dichotomous and polytomous data can be used, results can be plotted graphically (see Figure 5.1), and no specialized DIF software is needed as the transformation can be made easily in Microsoft Excel. However, before applying this method one must acknowledge that it assumes that total scores are normally distributed, no statistical tests are readily available, power for identifying nonuniform DIF is low compared with other methods (e.g., the standardization index), and statistical methods for validating test adaptations     

115

Figure 5.1.  Example of a delta plot. Data points that exceed the region defined by the dashed lines represent items displaying differential item functioning.

arbitrary rules for flagging potentially flawed items are used (see Holland & Wainer, 1993). For a good example of applying this procedure in the context of identifying potentially flawed items (i.e., poorly translated or adapted items), see Muñiz et al. (2001). Standardization Index Dorans and Kulick (1986) first introduced the standardization index (also referred to as the conditional p value method). This procedure conditions or matches examinees at particular score levels in the two groups and compares the difference in p values (proportion correct for dichotomous items and mean score for polytomous items) between the reference groups (e.g., English-language examinees) and focal groups (e.g., Spanish-language examinees) for item i. The standardization index is calculated as follows:

∑ K s  Pfs − Prs  D std = s =1 S ∑ s =1 K s S

116       rios and hambleton

where S = score level, Ks = weight at score S, Pfs = the proportion correct at score level s for the focal group, and Prs = the proportion correct at score level s for the reference group. One of the major considerations with using the standardization index is deciding how to weight the score levels. The levels can be unweighted or weighted by the size of the reference group, focal group, or both. When one group consistently outperforms the other across the score scale, after the groups are matched on their trait scores, the item is flagged as being problematic. Variations on this statistic are readily found in the DIF literature. To identify items where performance may change between two groups over the test score scale, sometimes the absolute values of the differences are summed (referred as an unsigned method):

∑ s=1 K s  Pfs − Prs  D std = . S ∑ s =1 K s S



The signed version of the standardization index ranges from -1.0 to 1.0 (the unsigned version ranges from 0 to 1); however, a statistical test of DIF is not available for this method. Nevertheless, researchers have suggested an effect size of 0.10 to indicate items that are potentially flawed (Dorans & Kulick, 1986; Muñiz et al., 2001; Sireci, 2011; Sireci et al., 2005). This effectsize criterion is based on the idea that if, for example, five items are deemed to possess DIF for one group, a half point (5 items × 0.10 = 0.5 points), which could be rounded off to result in a full point disadvantage to the focal group, might be consequential. So, items showing a standardized index of 0.10 are considered to be potentially consequential and are reviewed carefully to see if a problem with the item can be identified and corrected. The standardization index procedure has been used extensively in crosscultural and cross-linguistic research (e.g., Harding, 2012) because (a) it is easy to calculate, (b) it can be applied to small sample sizes (50 examinees in each group; Muñiz et al., 2001), (c) both uniform and nonuniform DIF can be identified, and (d) it allows for easy interpretation and explanation to stakeholders (DeMars, 2011). In addition, calculation of the standardization index does not require specialized DIF software and can be calculated in Microsoft Excel, any general statistical software package, or specialized programs, such as Robin’s (2001) STDIF. To provide a visual representation of the results, one can plot the average p value at a given score level separately for the reference and focal groups. Such plots make interpretation and explanation of DIF straightforward for nonstatistical audiences (see Figure 5.2). One of the major disadvantages of this procedure is that p values, like all percentages, suffer from not being calibrated on an equal-interval scale. This limitation makes for misjudgments in comparing across the item difficulty scale statistical methods for validating test adaptations     

117

1 0.9 Proportion Correct

0.8 0.7 0.6 0.5 0.4 0.3

Language A

0.2

Language B

0.1 0 0–30

31–35

36–40 Test Score Group

41–45

46–55

Figure 5.2.  Example of a standardization index plot. This plot of conditional p values compares the performance of two languages at five score levels for one dichotomous item. In this example, the item showed substantial uniform differential item functioning because the p value differences exceeded .30 in four of the five test score groups.

(Wood, 1988). Second, if the sample sizes are too small within a score-level grouping, unstable p value estimates will be obtained, which could erroneously flag items (i.e., Type I errors) or mask the presence of DIF (i.e., Type II errors). Last, the standardization index does not provide a test of statistical significance. Mantel–Haenszel The Mantel–Haenszel (MH) procedure has been used extensively in evaluating DIF for cross-linguistic and cross-cultural research (e.g., Dorans & Kulick, 2006). One of the major advantages of this procedure is that it can be easily computed using general statistical software, such as SAS or SPSS. For detailed syntax and practical examples of performing this procedure in SAS, refer to Stokes, Davis, and Koch (2000) or Camilli and Shepard (1994). Although the MH procedure was originally developed for analysis with dichotomously scored items, Zwick, Donoghue, and Grima (1993) extended it for use with polytomous items. MH effect size guidelines for polytomous items can be found in Zwick, Thayer, and Mazzeo (1997). Furthermore, of the procedures reviewed, only the MH method provides a statistical test of significance (see Holland & Thayer, 1988). However, one major limitation of this procedure is that to obtain stable estimates it requires a minimal sample size of 250 in each group (Camilli & Shepard, 1994; Muñiz et al., 2001; Spray, 118       rios and hambleton

1989), which is significantly larger than the other procedures described. In addition, this procedure is well-known to lack power in identifying nonuniform DIF (Rogers & Swaminathan, 1993). CONCLUSION The need for adapted tests has grown exponentially with both increased globalization and interest in cross-cultural research. Consider, for example, that several popular personality (Minnesota Multiphasic Personality Inventory; Hathaway & McKinley, 1940) and IQ tests (Wechsler Intelligence Scale for Children; Wechsler, 2003) are now available in more than 100 languages. However, numerous methodological challenges are presented with the potential benefits of providing adapted tests and surveys. In particular, the presence of measurement bias due to construct-irrelevant variance reduces the validity of cross-cultural and cross-lingual comparisons. To maximize the validity of these studies, we have provided a general review of methodological procedures to evaluate construct, method, and item (DIF) bias. Such methods, when applied correctly, should assist scientists in modifying their measurement procedures as well as increase the validity of inferences made with adapted tests. REFERENCES American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. Washington, DC: American Educational Research Association. Angoff, W. H. (1982). Use of difficulty and discrimination indices for detecting item bias. In R. A. Berk (Ed.), Handbook of methods for detecting test bias (pp. 96–116). Baltimore, MD: Johns Hopkins University Press. Arbuckle, J. L. (2010). AMOS (Version 18.0) [Computer software]. Chicago, IL: SPSS. Ariely, G., & Davidov, E. (2011). Can we rate public support for democracy in a comparable way? Cross-national equivalence of democratic attitudes in the World Value Survey. Social Indicators Research, 104, 271–286. http://dx.doi.org/10.1007/ s11205-010-9693-5 Bentler, P. M. (1995). EQS structural equations program manual. Encino, CA: Multivariate Software. Borg, I., & Groenen, P. J. F. (2005). Modern multidimensional scaling (2nd ed.). New York, NY: Springer. statistical methods for validating test adaptations     

119

Brennan, R. L. (2001). Generalizability theory. http://dx.doi.org/10.1007/978-14757-3456-0 Camilli, G. (2006). Test fairness. In R. L. Brennan (Ed.), Educational measurement (4th ed., pp. 221–256). Westport, CT: American Council on Education/Praeger. Camilli, G., & Shepard, L. A. (1994). Methods for identifying biased test items. Hollywood, CA: Sage. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait–multimethod matrix. Psychological Bulletin, 56, 81–105. http:// dx.doi.org/10.1037/h0046016 Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9, 233–255. http://dx.doi.org/10.1207/S15328007SEM0902_5 Cleeland, C. S., Nakamura, Y., Mendoza, T. R., Edwards, K. R., Douglas, J., & Serlin, R. C. (1996). Dimensions of the impact of cancer pain in a four country sample: New information from multidimensional scaling. Pain, 67(2–3), 267–273. http://dx.doi.org/10.1016/0304-3959(96)03131-4 Collazo, A. A. (2005). Translation of the Marlowe–Crowne Social Desirability Scale into an equivalent Spanish version. Educational and Psychological Measurement, 65, 780–806. http://dx.doi.org/10.1177/0013164405275660 Davidov, E. (2011). Nationalism and constructive patriotism: A longitudinal test of comparability in 22 countries with the ISSP. International Journal of Public Opinion Research, 23, 88–103. http://dx.doi.org/10.1093/ijpor/edq031 DeMars, C. E. (2011). An analytic comparison of effect sizes for differential item functioning. Applied Measurement in Education, 24, 189–209. http://dx.doi.org/ 10.1080/08957347.2011.580255 Dimitrov, D. M. (2010). Testing for factorial invariance in the context of construct validation. Measurement and Evaluation in Counseling and Development, 43, 121–149. http://dx.doi.org/10.1177/0748175610373459 Dorans, N. J., & Kulick, E. (1986). Demonstrating the utility of the standardization approach to assessing the unexpected differential item functioning on the Scholastic Aptitude Test. Journal of Educational Measurement, 23, 355–368. http:// dx.doi.org/10.1111/j.1745-3984.1986.tb00255.x Dorans, N. J., & Kulick, E. (2006). Differential item functioning on the Mini-Mental State Examination. An application of the Mantel–Haenszel and standardization procedures. Medical Care, 44(11, Suppl. 3), S107–S114. http://dx.doi. org/10.1097/01.mlr.0000245182.36914.4a Eid, M., Nussbeck, F. W., Geiser, C., Cole, D. A., Gollwitzer, M., & Lischetzke, T. (2008). Structural equation modeling of multitrait–multimethod data: Different models for different types of methods. Psychological Methods, 13, 230–253. http://dx.doi.org/10.1037/a0013219 Fischer, R., & Fontaine, J. R. J. (2011). Methods for investigating structural equivalence. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Cross-cultural research

120       rios and hambleton

methods in psychology (pp. 179–215). New York, NY: Cambridge University Press. Hambleton, R. K. (2005). Issues, designs, and technical guidelines for adapting tests into multiple languages and cultures. In R. K. Hambleton, P. Merenda, & C. Spielberger (Eds.), Adapting educational and psychological tests for cross-cultural assessment (pp. 3–38). Mahwah, NJ: Erlbaum. Hambleton, R. K., & de Jong, J. (2003). Advances in translating and adapting educational and psychological tests [Special issue]. Language Testing, 20, 127–134. http://dx.doi.org/10.1191/0265532203lt247xx Hambleton, R. K., Merenda, P., & Spielberger, C. (Eds.). (2005). Adapting educational and psychological tests for cross-cultural assessment. Mahwah, NJ: Erlbaum. Hambleton, R. K., & Zenisky, A. L. (2011). Translating and adapting tests for crosscultural assessments. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Crosscultural research methods in psychology (pp. 46–74). http://dx.doi.org/10.1017/ CBO9780511779381.004 Harding, L. (2012). Accent, listening assessment and the potential for a sharedL1 advantage: A DIF perspective. Language Testing, 29, 163–180. http://dx.doi. org/10.1177/0265532211421161 Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic personality schedule (Minnesota): 1. Construction of the schedule. Journal of Psychology, 10, 249–254. Hattrup, K., Ghorpade, J., & Lackritz, J. R. (2007). Work group collectivism and the centrality of work: A multinational investigation. Cross-Cultural Research, 41, 236–260. Holland, P. W., & Thayer, D. T. (1988). Differential item performance and the Mantel– Haenszel procedure. In H. Wainer & H. Braun (Eds.), Test validity (pp. 129–145). Hillsdale, NJ: Erlbaum. Holland, P. W., & Wainer, H. (Eds.). (1993). Differential item functioning. Hillsdale, NJ: Erlbaum. International Test Commission. (2010). ITC guidelines for translating and adapting tests. Retrieved from http://www.intestcom.org Jöreskog, K. G. (1974). Analyzing psychological data by structural analysis of covariance matrices. In D. H. Krantz, R. C. Atkinson, R. D. Luce, & P. Suppes (Eds.), Contemporary developments in mathematical psychology (2nd ed., pp. 1–56). San Francisco, CA: Freeman. Jöreskog, K. G., & Sörbom, D. (2001). LISREL 8: User’s reference guide. Lincolnwood, IL: Scientific Software International. Kankarasˇ, M., & Moors, G. (2010). Research measurement equivalence in cross-cultural studies. Psihologija [Psychology], 43, 121–136. http://dx.doi. org/10.2298/PSI1002121K Kenny, D. (1976). An empirical application of confirmatory factor analysis to the multitrait–multimethod matrix. Journal of Experimental Social Psychology, 12, 247–252. http://dx.doi.org/10.1016/0022-1031(76)90055-X statistical methods for validating test adaptations     

121

Li, D., & Brennan, R. (2007, April). A multi-group generalizability analysis of a largescale reading comprehension test. Paper presented at the meeting of the National Council on Measurement in Education, Chicago, IL. Little, T. D., & Slegers, D. W. (2005). Factor analysis: Multiple groups with means. In B. Everitt & D. Howell (Eds.), Encyclopedia of statistics in behavioral science (pp. 617–623). http://dx.doi.org/10.1002/0470013192.bsa221 Lord, F. M. (1980). Applications of item response theory to practical testing problems. Hillsdale, NJ: Erlbaum. Marsh, H. W., & Bailey, M. (1991). Confirmatory factor analyses of multitrait– multimethod data: A comparison of alternative models. Applied Psychological Measurement, 15, 47–70. http://dx.doi.org/10.1177/014662169101500106 Marsh, H. W., & Byrne, B. M. (1993). Confirmatory factor analysis of multitrait– multimethod self-concept data: Between-group and within-group invariance constraints. Multivariate Behavioral Research, 28, 313–449. http://dx.doi.org/ 10.1207/s15327906mbr2803_2 Matsumoto, D., & van de Vijver, F. J. R. (Eds.). (2011). Cross-cultural research methods in psychology. New York, NY: Cambridge University Press. Meade, A. W., Johnson, E. C., & Braddy, P. W. (2008). Power and sensitivity of alternative fit indices in tests of measurement invariance. Journal of Applied Psychology, 93, 568–592. http://dx.doi.org/10.1037/0021-9010.93.3.568 Millsap, R. E. (2011). Statistical approaches to measurement invariance. New York, NY: Routledge. Muñiz, J., Hambleton, R. K., & Xing, D. (2001). Small sample studies to detect flaws in item translations. International Journal of Testing, 1, 115–135. http://dx.doi. org/10.1207/S15327574IJT0102_2 Muthén, L. K., & Muthén, B. O. (2007). Mplus user’s guide (5th ed.). Los Angeles, CA: Authors. Oakland, T. (2005). Selected ethical issues relevant to test adaptations. In R. K. Hambleton, P. Merenda, & C. Spielberger (Eds.), Adapting educational and psychological tests for cross-cultural assessment (pp. 65–92). Mahwah, NJ: Erlbaum. Osterlind, S. J., & Everson, H. T. (2009). Differential item functioning (2nd ed.). http://dx.doi.org/10.4135/9781412993913 Raju, N. S. (1988). The area between two item characteristic curves. Psychometrika, 53, 495–502. http://dx.doi.org/10.1007/BF02294403 Raju, N. S. (1990). Determining the significance of estimated signed and unsigned areas between two item response functions. Applied Psychological Measurement, 14, 197–207. http://dx.doi.org/10.1177/014662169001400208 Rios, J. A., & Sireci, S. G. (2014). Guidelines versus practices in cross-lingual assessment: A disconcerting disconnect. International Journal of Testing, 14, 289–312. http://dx.doi.org/10.1080/15305058.2014.924006 Robin, F. (2001). STDIF: Standardization-DIF analysis program. Amherst: University of Massachusetts, Center for Educational Assessment.

122       rios and hambleton

Robin, F., Sireci, S. G., & Hambleton, R. K. (2003). Evaluating the equivalence of different language versions of a credentialing exam. International Journal of Testing, 3, 1–20. http://dx.doi.org/10.1207/S15327574IJT0301_1 Rogers, H. J., & Swaminathan, H. (1993). A comparison of the logistic regression and Mantel–Haenszel procedures for detecting differential item functioning. Applied Psychological Measurement, 17, 105–116. http://dx.doi.org/10.1177/ 014662169301700201 Shavelson, R. J., & Webb, N. M. (1991). Generalizability theory: A primer. Thousand Oaks, CA: Sage. Sireci, S. G. (2011). Evaluating test and survey items for bias across languages and cultures. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Cross-cultural research methods in psychology (pp. 216–240). New York, NY: Cambridge University Press. Sireci, S. G., Patsula, L., & Hambleton, R. K. (2005). Statistical methods for identifying flaws in the test adaptation process. In R. K. Hambleton, P. Merenda, & C. Spielberger (Eds.), Adapting psychological and educational tests for cross-cultural assessment (pp. 93–116). Mahwah, NJ: Erlbaum. Solano-Flores, G., & Li, M. (2009). Generalizability of cognitive interview-based measures across cultural groups. Educational Measurement: Issues and Practice, 28(2), 9–18. http://dx.doi.org/10.1111/j.1745-3992.2009.00143.x Spray, J. A. (1989). Performance of three conditional DIF statistics in detecting differential item functioning on simulated tests (Research Report No. 89-7). Iowa City, IA: American College Testing. Stokes, M. E., Davis, C. S., & Koch, G. G. (2000). Categorical data analysis using the SAS system. Cary, NC: SAS Institute and Wiley. Swaminathan, H., & Rogers, H. J. (1990). Detecting differential item functioning using logistic regression procedures. Journal of Educational Measurement, 27, 361–370. http://dx.doi.org/10.1111/j.1745-3984.1990.tb00754.x Thissen, D., Steinberg, L., & Wainer, H. (1988). Use of item response theory in the study of group differences in trace lines. In H. Wainer & H. I. Braun (Eds.), Test validity (pp. 147–169). Hillsdale, NJ: Erlbaum. van de Vijver, F. J. R., & Hambleton, R. K. (1996). Translating tests. European Psychologist, 1(2), 89–99. http://dx.doi.org/10.1027/1016-9040.1.2.89 van de Vijver, F. J. R., & Leung, K. (1997). Methods and data analysis for cross-cultural research. Newbury Park, CA: Sage. van de Vijver, F. J. R., & Leung, K. (2011). Equivalence and bias: A review of concepts, models, and data analytic procedures. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Cross-cultural research methods in psychology (pp. 17–45). New York, NY: Cambridge University Press. van de Vijver, F. J. R., & Matsumoto, D. (2011). Introduction to the methodological issues associated with cross-cultural research. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Cross-cultural research methods in psychology (pp. 1–16). New York, NY: Cambridge University Press. statistical methods for validating test adaptations     

123

van de Vijver, F. J. R., & Poortinga, Y. H. (1991). Testing across cultures. In R. K. Hambleton & J. N. Zaal (Eds.), Advances in educational and psychological testing: Theory and applications (pp. 277–308). Dordrecht, Netherlands: Springer. van de Vijver, F. J. R., & Poortinga, Y. H. (1997). Towards an integrated analysis of bias in cross-cultural assessment. European Journal of Psychological Assessment, 13(1), 29–37. http://dx.doi.org/10.1027/1015-5759.13.1.29 van de Vijver, F. J. R., & Poortinga, Y. E. (2005). Conceptual and methodological issues in adapting tests. In R. K. Hambleton, P. Merenda, & C. Spielberger (Eds.), Adapting educational and psychological tests for cross-cultural assessment (pp. 39–64). Mahwah, NJ: Erlbaum. van de Vijver, F., & Tanzer, N. K. (1998). Bias and equivalence in cross-cultural assessment: An overview. European Review of Applied Psychology, 47, 263–279. Wechsler, D. (2003). Wechsler Intelligence Scale for Children–Fourth Edition (WISC–IV). San Antonio, TX: The Psychological Corporation. Wolff, H.-G., Schneider-Rahm, C. I., & Forret, M. L. (2011). Adaptation of a German multidimensional networking scale into English. European Journal of Psychological Assessment, 27, 244–250. http://dx.doi.org/10.1027/1015-5759/ a000070 Wood, R. (1988). Item analysis. In J. P. Keeves (Ed.), Educational research, methodology, and measurement: An international handbook (pp. 376–384). New York, NY: Pergamon Press. Yen, C. J., & Tu, C. H. (2011). Multiple-group confirmatory factor analysis of the scores for online social presence: Do they measure the same thing across cultural groups? Journal of Educational Computing Research, 44, 219–242. http://dx.doi. org/10.2190/EC.44.2.e Zwick, R., Donoghue, J. R., & Grima, A. (1993). Assessment of differential item functioning for performance tasks. Journal of Educational Measurement, 30, 233–251. http://dx.doi.org/10.1111/j.1745-3984.1993.tb00425.x Zwick, R., Thayer, D. T., & Mazzeo, J. (1997). Descriptive and inferential procedures for assessing differential item functioning in polytomous items. Applied Measurement in Education, 10, 321–334. http://dx.doi.org/10.1207/s15324818ame1004_2

124       rios and hambleton

6 TESTING INSTRUMENT EQUIVALENCE ACROSS CULTURAL GROUPS: BASIC CONCEPTS, TESTING STRATEGIES, AND COMMON COMPLEXITIES BARBARA M. BYRNE

Although assessment of mean group differences in both achievement and attitudinal scores across culturally diverse groups has been of interest for many decades, investigation into the extent to which their measuring instruments are equivalent (i.e., invariant) across such groups is a relatively recent phenomenon. Of critical concern is the use of instruments that have been developed and normed in one culture and then used in another culture, either in their original linguistic form or as a translated version of the original instrument. Indeed, the testing of both versions of an instrument carries very strong and likely unrealistic assumptions of its equivalence across cultures. Because comparison of group mean (i.e., level) scores represents the primary focus of most multicultural substantive research,1 the extent to which the instrument is measuring the same construct(s) in exactly the same way within Multicultural research is an umbrella term that refers to the comparison of two or more cultural groups. It is equivalently applicable both within an international context (comparisons across countries) and in a national context (comparisons across subcultural groups within a country). 1

http://dx.doi.org/10.1037/14940-007 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

125

each group is clearly critical. Should this assumption not hold, then the issue of bias is of primary concern. The present chapter addresses these measurement issues and provides example applications within the framework of a national multiculturalism perspective that focuses on comparisons across mainstream and subcultural groups within the United States. More specifically, the purposes of this chapter are threefold: (a) to present an overview of the basic concepts associated with the issue of measurement and structural equivalence, (b) to outline procedures used in testing for the equivalence of measuring instruments across cultural groups, and (c) to identify and elaborate upon the many complexities associated with testing for these instrument equivalencies across mainstream and subcultural groups. BASIC CONCEPTS OF MEASUREMENT AND STRUCTURAL EQUIVALENCE All multigroup comparisons of mean (or level) scores assume that both the measuring instrument and the construct being measured are operating in the same way across the populations of interest. In other words, there is presumed equality of (a) factorial structure (i.e., the same number of factors and the same pattern of item loadings onto these factors), (b) perceived item content, (c) factor loadings (i.e., similar size of item estimates), and (d) item intercepts (i.e., item means). In light of their psychometric focus, these characteristics are commonly regarded as representing measurement equivalence (alternatively termed measurement invariance). Likewise, there is presumed equality of the measured construct with respect to (e) its dimensionality (i.e., unidimensional or multidimensional structure) and (f) in the case of multidimensional structure, relations among the construct dimensions. If the focus is on theoretical structure (see Bentler, 1978), the latter characteristics are considered to represent structural equivalence (alternatively termed structural invariance). These assumptions, consistent with all statistical assumptions, need to be tested. Indeed, Vandenberg and Lance (2000) cautioned that failure to establish measurement and structural equivalence is as damaging to substantive interpretations as the inability to demonstrate reliability and validity. When mean group comparisons focus on cross-cultural groups, testing for measurement and structural equivalence falls prey to a host of additional complexities. These complications arise from the customary use of two perspectives in measuring a construct (or constructs) of interest: (a) use of the same measuring instrument, in its original linguistic form, across cultural groups; and (b) use of a translated version of an instrument for populations whose culture differs from the one in which the instrument was originally developed and 126       barbara m. byrne

normed. In both instances, researchers and practitioners have no grounds for assuming either that the instrument operates equivalently or that the norms are equally relevant across groups. Although most research concerned with these methodological issues to date has focused on educational achievement tests, the issues are particularly potent for psychological assessments. ANALYTIC STRATEGIES IN TESTING FOR INSTRUMENT EQUIVALENCE The two primary strategies used in testing for measuring instrument equivalence are the structural equation modeling (SEM) and the item response theory (IRT) approaches, each of which derives from a different disciplinary base, has its own statistical assumptions, and has its own procedural idiom. For example, whereas the SEM approach is rooted in econometrics, the statistical base of the economics discipline, IRT is linked to the discipline of education; whereas SEM is based on a linear model and can address the issue of nonnormal as well as categorical data, IRT is based on a nonlinear model and assumes normally distributed data; whereas the analytic strategy used in testing the multigroup equivalence of items on an assessment scale is termed testing for measurement equivalence in SEM, it is referred to as testing for differential item functioning in IRT. Indeed, one might think of these two analytic strategies as unequal twins. Although both analytic strategies were developed in the late 1960s (SEM: Jöreskog, 1969; IRT: Lord & Novick, 1968), it was not until the early 1980s, largely as a consequence of availability of high-speed computers and the related software, that these methodologies became more widely known and applied in the research community. At that time, the thrust of SEM application focused on issues related to construct validation (of assessment scales; of theoretical systems) and causal relations among theoretically and empirically based networks of interrelated constructs, based on the analysis of covariance structures within the frameworks of confirmatory factor analysis (CFA) and path analytic modeling. Most of this applied research was conducted within the disciplines of economics, psychology, and sociology. In contrast, the focus of IRT application centered on validation of test scores derived from instruments designed to measure academic achievement or performance, albeit largely within the context of large-scale educational assessments (Embretson & Reise, 2000). Indeed, a review of the literature pertinent to IRT applied research suggests that it was not until publication of an IRT textbook designed specifically for use with psychological data (see Embretson & Reise, 2000) that the theory and application of this methodology became of interest to and spawned application by psychologists. In testing instrument equivalence     

127

this chapter, only the SEM approach to tests for instrument equivalence is addressed because it is unique in its capacity to test for both the measurement and structural equivalence of a measuring instrument. Details regarding the IRT strategy are reviewed in Chapter 5 of this volume. The SEM Approach to Testing for Instrument Equivalence Development of a procedure capable of testing for the multigroup equivalence of a measuring instrument derives from the seminal work of Jöreskog (1971). The approach encompasses a series of hierarchical steps that begins with the determination of a well-fitting baseline multigroup model for which sets of parameters are put to the test of equality in a logically ordered and increasingly restrictive fashion. (For a walk-though of these procedures based on different types of models and SEM software, refer to Byrne, 1998, 2006, 2010, 2011, and on specific application to cross-cultural groups, refer to Byrne, 2008.) The first and least restrictive model to be tested is commonly termed the configural model (Horn & McArdle, 1992). With this initial model, only the extent to which the same pattern (or configuration) of fixed and freely estimated parameters holds across groups is of interest, and thus no equality constraints are imposed. The importance of the configural model is that it serves as the baseline against which all subsequent tests for equivalence are compared. In contrast, all remaining tests for equivalence involve the specification of crossgroup equality constraints for particular parameters. Technically speaking, the first three constrained models test for measurement equivalence, whereas the final two tests address the issue of structural equivalence. Measurement equivalence must be established prior to testing for structural equivalence. Testing for Measurement Equivalence This set of three tests focuses on the equality across groups of the factor loadings, the item (i.e., observed variable) intercepts, and the error uniquenesses. Because each of these tests represents an increased level of restrictiveness, Meredith (1993) categorized them as weak, strong, and strict tests of equivalence, respectively. In testing for the equivalence of factor loadings, these parameters are freely estimated for the first group, with those for the remaining groups constrained equal to those of Group 1. When evidence of equivalence is provided, these factor loading parameters remain constrained during simultaneous testing for the equivalence of all additional parameters. On the other hand, presented with findings of nonequivalent factor loadings, one may proceed with subsequent tests for equivalence if the data meet the conditions of partial measurement equivalence (see Byrne, Shavelson, & Muthén, 1989). 128       barbara m. byrne

Tests for equivalent factor pattern and factor loadings are based on the analysis of covariance structures, which assumes that all observed variables (e.g., items, subscores) are measured as deviations from their means (i.e., their means are equal to zero). However, in moving on to the next more restrictive test of measurement equivalence, the equality of item intercepts, analyses are based on mean and covariance structures (i.e., item means are no longer zero); that is, analyses are based on the moment matrix, which includes both the sample means and covariances. Of import in testing for the equivalence of cross-group intercepts is that it subsequently allows for the multigroup comparison of latent construct means (i.e., means on the factors), should this be of interest. Although some researchers have contended that this “strong” test of equivalence should always be conducted (e.g., Little, 1997; Meredith, 1993), others (e.g., Marsh, 1994, 2007; Marsh, Hau, Artelt, Baumert, & Peschar, 2006) have argued that analysis of only covariance structures may be the most appropriate approach to take in addressing the issues and interests of a particular study (e.g., validating structure of a theoretical construct). The final and most stringent test of measurement equivalence, which Meredith (1993) labels strict equivalence, focuses on the equality of error uniqueness2 variances across groups. However, it is now widely accepted that this test of the data is overly restrictive and likely of least interest and importance (Bentler, 2005), unnecessary (Widaman & Reise, 1997), and, as argued more recently, not recommended (Selig, Card, & Little, 2008). Testing for Structural Equivalence In contrast to tests for measurement equivalence, which focus on aspects of the observed variables, tests for structural equivalence center on the unobserved (or latent) variables.3 In the case of testing for the equivalence of a measuring instrument across groups, interest can focus on both the factor variances and their covariances, although the latter are typically of most interest. A review of the SEM literature reveals much inconsistency regarding whether or not researchers test for structural equivalence. In particular,

The term error uniqueness is used here in the factor analytic sense that denotes a composite of random measurement error and specific measurement error pertinent to particular items of the measuring instrument. 3 Byrne and van de Vijver (2014) noted a substantially modified interpretation of the term structural equivalence within the context of cross-cultural psychological research. Specifically, structural equivalence is considered to be evidenced when exploratory factor analytic results identify the same factors across cultural groups, thereby supporting the notion that the underlying construct(s) is identical across groups (see van de Vijver, 2011). In stark contrast, within the SEM tradition, evidence of structural equivalence is based on a CFA framework that allows for the hypothesized equality of both the factorial dimensionality of the underlying construct(s) and the relations among these dimensions to be tested statistically. Claims of equivalence are made only if these hypotheses are tenable. 2

testing instrument equivalence     

129

these tests are of critical import to construct validity researchers whose interests lie in testing either the extent to which the dimensionality of a construct, as defined by theory, holds across groups (see, e.g., Byrne & Shavelson, 1987) or the extent to which an assessment scale, developed within the framework of a particular theory, yields the expected dimensional structure of the measured construct in an equivalent manner across groups (see, e.g., Byrne & Watkins, 2003). In both instances, the parameters of most interest are the factor covariances. I have summarized the basic set of tests for cross-group measurement and structural equivalence based on SEM analyses. Clearly, researchers can determine, a priori, the level of stringency they wish to implement in testing for multigroup equivalence; this decision will depend on the particular data under study (see Widaman & Reise, 1997). Although Meredith (1993) distinguished among three increasingly restrictive tests for equivalence, to date there is no hard and fast rule governing which tests should be conducted once equivalence of the factor loadings has been established. For detailed descriptions of these tests for multigroup equivalence, refer to Horn and McArdle (1992), Little (1997), and Widaman and Reise (1997); for an in-depth and detailed overview of measurement bias as it bears on tests for measurement equivalence based on both SEM and IRT, to Millsap (2011); for an annotated explanation and illustration of diverse models based on the LISREL, AMOS, EQS, and Mplus programs, to Byrne (1998, 2006, 2010, 2011), respectively; and for synopses of, and distinctions between these primary SEM programs, to Byrne (2012). Example Graphical Overview of Testing for Instrument Equivalence Adhering to the old cliché that a picture is worth a thousand words, I now present a brief graphical overview of a hypothetical, albeit typical, SEM model for which measurement and structural equivalence can be tested across two or more comparison groups. Models depicting a single instrument, for which the validity and equivalence of its factorial structure are under investigation, classically represent CFA schema. (For a comprehensive overview of diverse CFA models, refer to Brown, 2006.) The model shown in Figure 6.1 pictorially describes the factorial structure of the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1981, 1986), an assessment scale designed to measure three dimensions of burnout considered by the developers to represent emotional exhaustion (EE), depersonalization (DP), and reduced personal accomplishment (PA). The term burnout denotes the inability to function effectively in one’s job as a consequence of prolonged and extensive jobrelated stress, emotional exhaustion represents feelings of fatigue that develop as one’s energies become drained, depersonalization represents the development 130       barbara m. byrne

ITEM 1 ITEM 2 ITEM 3 EMOTIONAL EXHAUSTION F1

ITEM 6 ITEM 8 ITEM 13 ITEM 14 ITEM 16 ITEM 20

ITEM 5 ITEM 10 DEPERSONALIZATION F2

ITEM 11 ITEM 15 ITEM 22

ITEM 4 ITEM 7 PERSONAL ACCOMPLISHMENT F3

ITEM 9 ITEM 12 ITEM 17 ITEM 18 ITEM 19 ITEM 21

Figure 6.1.  Model of hypothesized structure for the Maslach Burnout Inventory (Maslach & Jackson, 1981, 1986). From Structural Equation Modeling with Mplus: Basic Concepts, Applications, and Programming (p. 97), by B. M. Byrne, 2011, New York, NY: Routledge/Taylor & Francis. Copyright 2011 by Routledge/Taylor & Francis. Reprinted with permission.

of negative and uncaring attitudes toward others, and reduced personal accomplishment represents a deterioration of self-confidence and dissatisfaction in one’s achievements. The MBI is a 22-item instrument structured on a 7-point Likert-type scale that ranges from 0 (feeling has never been experienced) to 6 (feeling experienced daily). Let’s take a look now at Figure 6.1, where a CFA model represents the hypothesized factorial structure of the MBI. For readers who may not be familiar with the symbols associated with SEM models, a brief description is in order. These models are schematically portrayed as path diagrams through the incorporation of four geometric symbols: an ellipse (or circle) representing unobserved latent factors; a rectangle (or square) representing observed variables; a single-headed arrow (→) representing the impact of one variable on another; and a double-headed arrow (↔) representing covariance between pairs of variables. In building SEM models, researchers use these symbols within the framework of different configurations, each of which represents an important component in the analytic process. On the basis of these geometric configurations, decomposition of this CFA model conveys the following information: (a) there are three factors, as indicated by the three ellipses labeled Emotional Exhaustion (F1), Depersonalization (F2), and Personal Accomplishment (F3); (b) the four factors are intercorrelated, as indicated by the three two-headed arrows; (c) there are 22 observed variables, as indicated by the 22 rectangles (ITEM1–ITEM22), each of which represents one item from the MBI; and (d) the observed variables load on the factors in the following pattern: Items 1, 2, 3, 6, 8, 13, 14, 16, and 20 load on Factor 1; Items 5, 10, 11, 15, and 22 load on Factor 2; and Items 4, 7, 9, 12, 17, 18, 19, and 21 load on Factor 3. The single-headed arrows leading to each of these observed variables (i.e., the items) represent random measurement as well as unique error associated with these variables. Prior to the initiation of tests for equivalence, it is essential that the best-fitting, yet concomitantly most parsimonious model first be established separately for each group under study. This best-fitting model is termed a baseline model. In establishing each single-group baseline model, it may be that the analyses identify a few misspecified parameters; for CFA models, two such parameters are possible: (a) that one (or more) item(s) cross-loads onto a nontarget factor and/or (b) as a consequence of overlapping item content, there is evidence of one or more substantially large correlated errors. If and only if these identified misfitting parameters are deemed to be substantively meaningful, while at the same time a watchful eye on scientific parsimony is maintained, these parameters may be justifiably included in the final baseline model for the specific group in question. Once these single-group baseline models are established, one can then proceed in testing for equivalence of the hypothesized factor structure of the instrument across groups. As outlined 132       barbara m. byrne

earlier, this multigroup process begins with the structuring and testing of the configural model, which, in essence, can be considered a multigroup baseline model as it holds specifications for each of the single-group baseline models in one file. (For a more detailed elaboration of these procedures, consult one of the following resources for a more complete step-by-step explanation of these processes: Byrne, 1998, 2006, 2008, 2010, 2011.) COMMON COMPLEXITIES IN TESTING FOR INSTRUMENT EQUIVALENCE ACROSS CULTURAL GROUPS Several researchers have addressed the issue of equivalence in crosscultural research (see, e.g., Johnson, 1998; Leong, Okazaki, & Tak, 2003; Leung & Wong, 2003; van de Vijver & Leung, 1997, 2000) and all agree that it encompasses many complexities that tend to be present, albeit absent in monocultural research. Indeed, it is this same broad range of complexities that leads, typically and ultimately, to findings of nonequivalent measurement and/ or structural equivalence when the sample groups represent different cultural populations. When the assumptions of equivalent instrument measurement and structure fail to hold, the issue of bias is of primary concern. Bias in Multicultural Research: What It Is and How It Happens In general terms, bias refers to the presence of systematic as opposed to random sources of error (e.g., poorly structured items on an assessment scale). When test scores are biased, meaningful interpretation of both their scores and the underlying constructs to which they are linked cannot be compared across groups. In other words, biased scores exhibit differential validity across the groups of interest. Historically, within a North American context, the bias issue has been interpreted within the framework of two somewhat divergent perspectives: (a) the question of test fairness across groups and (b) the question of the extent to which the construct is measuring different aspects of its domain across groups. As a consequence of their diverse orientations, the operational definition of bias has tended to differ for cognitive instruments of measurement compared with affective instruments of measurement. Whereas bias associated with cognitive measures has typically been interpreted as indicating that equally able individuals from different minority/ ethnic groups have unequal opportunities of success, bias associated with affective measures has been interpreted as indicating that test scores based on the same items measure different traits and characteristics for each group. Indeed, given the less concrete nature of psychological constructs and the fact that their structure is so strongly influenced by cultural factors, affective measures testing instrument equivalence     

133

such as attitude scales require very strong evidence that the test items tap the underlying constructs in exactly the same way for all groups. In terms of multicultural research, it is important to note that the issue of bias does not relate to the intrinsic properties of an assessment instrument per se but rather to the characteristics of the respondents from each cultural group (van de Vijver & Tanzer, 2004). Thus, statements regarding bias should always refer to use of an instrument within the context of a particular multicultural comparison. For example, whereas an instrument may reveal evidence of bias in a comparison of African and Caucasian Americans, such evidence may not be present in a comparison of African and Hispanic Americans. In general, problems of bias in multicultural research can be linked to three primary sources: (a) the construct(s) of interest, (b) the methodological procedures, and (c) the item content (van de Vijver & Leung, 1997). I turn now to a brief description and research example representative of these three roots of bias. Construct Bias The first and clearly the most problematic type of bias that can lead to complexities in tests for instrument equivalence is that of construct bias. This form of bias conveys the notion that the construct being measured holds some degree of differential meaningfulness across the cultural groups under study. In essence, the root of the problem in construct bias lies with the behaviors considered by the test developer to most appropriately represent the construct to which they are linked via the item content. In measuring a construct(s) across different cultural groups, inappropriate selection of behaviors can manifest itself in at least three ways, thereby leading to the presence of bias and, ultimately, to evidence of nonequivalent measurements across groups. I turn now to an overview of each of these problematic conditions. The first of these situations leading to construct bias arises when the selected behaviors considered as indicators of a construct are differentially appropriate across cultural groups. That is to say, the item content is irrelevant and, hence, meaningless to one of the groups. A good example here can be drawn from the work of Ho (1996) and others (e.g., Kuang et al., 2010; Liu, Ng, Weatherall, & Loong, 2000) in their studies of filial piety, the concept of being a “good” son or daughter and having a strong sense of duty and obligation to parents and other family members, particularly those who are elderly (Matsumoto, 2000; van de Vijver, 2011). Long regarded as the first and foremost virtue in the Chinese culture, the concept of filial piety is a key element in the socialization of their children. Furthermore, it has been shown to remain an important component of the Chinese value system (Yue & Ng, 1999). In contrast, although the basic principles of this concept may be held by some in Western cultures, the concept of independence, rather than filial 134       barbara m. byrne

piety, tends to form the cornerstone of family socialization patterns. Indeed, interest in these rather divergent constructs has precipitated a substantial number of comparative studies across Western and non-Western cultures. In light of the somewhat contradictory nature of the filial piety and independence constructs, together with their known impact on the socialization of children in mainstream Chinese and American families, respectively, one example of where this first source of construct bias might play out is in the assessment of parental self-concept for Chinese immigrant and mainstream American high school adolescents. Because immigrant Chinese adolescents are still likely to honor the expected obligations toward parents and grandparents, it seems logical to assume that perceptions of self relative to one’s parents would be based on a different set of criteria than would be the case for American adolescents; these criteria, in turn, would likely generate a differential set of behaviors considered to tap the underlying construct of filial piety. A second instance leading to construct bias arises when the sampling of behaviors considered to represent a measured construct is inadequate for one of the groups under study. In other words, the set of behaviors conceived by the test developer as being sufficient to measure a particular construct fails to tap into all possible behaviors for this group. Let’s take, for example, an assessment of social self-concept for elementary school children. One such instrument designed to measure this construct is the Self Description Questionnaire I (Marsh, 1992). More specifically, Marsh (1992) included two separate subscales in measuring this construct; one subscale included items measuring social self-concept related to peers and the other subscale, social self-concept related to family. Consider an example in which two groups of elementary school children are being tested for comparison purposes; one group represents a sample of Mennonite children from the Pennsylvania Dutch area of Pennsylvania and the other, a group of mainstream Caucasian children from Chicago. Whereas extended families are well known to represent the norm for Mennonite communities, they are fast becoming an historical artifact for mainstream urban families in the Western world. Thus, it seems reasonable to assume that for the Mennonite children, perceptions of self within the social context of family (i.e., social self-concept) would be based on a much broader range of social interactive behaviors than would be the case for their elementary school counterparts living in the urban communities of Chicago. Likewise, the same comparison can be made with respect to peer relationships, albeit somewhat in reverse. In light of the somewhat insular nature of Mennonite communities in Pennsylvania, the variety of peers with whom elementary school children will interact on a daily basis is expected to be much narrower than would be the case for the Chicago school children. In this instance, then, the behaviors being tapped by items designed to measure peer social self-concept may be testing instrument equivalence     

135

overrepresented for the Mennonite children and underrepresented for the Chicago children. Finally, a third problematic factor leading to construct bias is the case in which the items fail completely to tap behaviors representative of the measured construct for one of the groups. The concept of street smartness can provide a good example here. Considered a facet of practical intelligence, street smartness tends to be more closely linked to one’s ability to cope in dangerous, unsafe, and/or unfamiliar environments or situations. It conveys the notion of always being aware of one’s surroundings, having an inner sense of dangerous situations, and using one’s common sense in the planning of one’s actions. In broad terms, then, street smartness is considered to be a resourcefulness that is learned not from any formal teaching environment but rather from the real world of everyday living. An example of a situation in which this type of construct bias might come into play is the use of an interview format to assess immediately perceived response to various dangerous scenarios for adolescents attending, say, inner city high schools in New York City versus those attending high schools in the smaller Adirondack communities of upstate New York. (For additional examples of construct bias, albeit pertinent to international samples, see van de Vijver, 2011.) Method Bias A second major source of bias that can lead to findings of instrument nonequivalence is method bias. This type of bias derives from one of three aspects of the methodology used in making mean score comparisons across cultural groups. The first of these is termed sample bias and relates to the incomparability of samples on phenomena other than the target factors under study. A case in point can be made in the measurement of academic self-concepts. Despite the fact that selected groups of children from different cultures might be categorized as belonging to the same grade level, it is nonetheless very easy for their educational experiences to be dramatically different. As a consequence, the criteria upon which they formulate their self-perceptions of academic ability in particular subject areas may be vastly different. Take, for example, the case of verbal self-concept. Historically, it has been customary to link this dimension of academic self-concept to English as a school subject. However, this is one academic area for which the curriculum can vary widely even within the same culture. Without question, then, it seems reasonable to assume that this curriculum will likely differ across culture as well. For example, one curriculum of study might emphasize acquired skills related to literature, grammar, reading ability, and writing ability; in another culture, only reading and writing ability may be considered of primary importance. Thus, in assessing one’s self-perception of verbal ability 136       barbara m. byrne

for, say, Grade 8 adolescents in the large city of Los Angeles where many of the students may be recent Hispanic immigrants, this aspect of method bias, in addition to one’s familiarity with the English language, could very well impact the validity of scores of verbal self-concept and ultimately on evidence of nonequivalent measurements across various groups of Grade 8 students. A second type of method bias derives from problems associated with the assessment measure used and is therefore termed instrument bias. More specifically, it relates to the differential response by comparative groups to the structured format of the assessment instrument. This type of bias is most commonly found with the use of cognitive tests (van de Vijver, 2011). One recognized source of instrument bias is that of stimulus familiarity. An example can be found in the work of Deregowski and Serpell (1971) in which Scottish and Zimbabwean children were asked to sort models of animals and cars, and then asked to do so again based on photographs of these models. Although the authors reported no cultural differences when the actual models were sorted, the Scottish children attained significantly higher scores when the sorting was based on the photographs. Because most self-concept instruments are based on paper-and-pencil tests that are structured around a multiple-choice, Likert scaling format, it is indeed possible that this type of stimulus response may be unfamiliar, for example, to recent immigrant respondents and this unfamiliarity is thereby reflected in a biasing of item scores. A second type of instrument bias can be found with respect to patterns of response. These patterns can reflect evidence of response bias in one of two ways: (a) by consistently selecting one of the two extreme scale points (high or low), with such selection being completely independent of the item content (this type of response bias is termed a response style) and (b) by selecting scale points, either consciously or unconsciously, in such a way as to convey a favorable impression of oneself (e.g., social desirability, acquiescence). This type of response bias is termed a response set. Response bias, whether it be in the form of a response style or a response set, is certainly not uncommon to cross-cultural research. Early work in this area, for example, has shown a clear tendency for Hispanics, as opposed to non-Hispanics, to choose the extreme response option of multicategory Likert scales (see, e.g., Hui & Triandis, 1989; Marín, Gamba, & Marín, 1992). More recently, in a comparison of factor analytic structure related to the Beck Depression Inventory for Canadian, Bulgarian, and Swedish adolescents, Byrne and Campbell (1999) reported a substantially different pattern of response for the latter. Although all three nonclinical adolescent groups typically assigned a large percentage of their responses to the lowest category (no indication of depression), as might be expected, this assignment was dramatically higher and more consistent for Swedish adolescents. This discrepant testing instrument equivalence     

137

responding pattern by the Swedes was attributed to the highly salient and important cultural value of self-disclosure. Two important cross-cultural studies have specifically addressed the issue of differential response bias (Cheung & Rensvold, 2000; Watkins & Cheung, 1995). The first of these (Watkins & Cheung, 1995) focused on patterns of response related to subscale scores from the Self Description Questionnaire for Australian, Chinese, Nepalese, Nigerian, and Filipino children 12 to 14 years of age. Findings from this study revealed no evidence of response set bias but substantial differences in response styles across culture; moreover, strong evidence of a country by gender interaction was reported. The second study (Cheung & Rensvold, 2000) used CFA to test the extent to which extreme and acquiescent response styles related to scores from the Work Orientation subscale of the International Social Survey Program differed across 11 countries. These authors concluded that failure to take into account the differential styles of response patterning relative to each cultural group can have a severe biasing effect on the resulting assessment scores. The final source of method bias is that of administration bias. Although this type of bias can distort all modes of testing, the interview format would appear to be particularly vulnerable. Indeed, van de Vijver and Tanzer (2004) noted that communication problems between interviewers and interviewees can easily occur, particularly when their first languages and cultural backgrounds are different. These authors further posited that, given an interviewee’s insufficient knowledge of the testing language and/or an interviewer’s mode of address being in violation of the cultural norms of the interviewees, the collection of appropriate data can be seriously jeopardized. Item Bias A final category of bias is that of item bias. As its name implies and in contrast to construct and method biases, item bias refers to distortions at the item level. As such, items are said to be biased if they elicit a differential meaning of their content across cultural groups. Differential interpretation of item content by members of culturally different groups derives largely from a diversity of sociocultural contexts that include the family, the school, the peer group, and society at large. For example, Oyserman and Markus (1993) noted that whereas American families urge children to stand up for themselves and not be pushed around, Japanese families stress the value of working in cooperation with others. In contrast to Americans, they do not perceive the yielding of personal autonomy as a depression of one’s self-esteem. Thus, in testing for the equivalence of an instrument designed to measure self-esteem across, say, American Asian and American Caucasian adolescents, it seems evident that 138       barbara m. byrne

differing socialization practices cannot help but lead to different sets of criteria against which to judge one’s perception of self. Other important factors that can contribute to the differential interpretation of item content are the impact of cultural norms (e.g., legality of drug use), ambiguous item content, use of colloquialisms (idiomatic expressions unique to a particular culture), and poor item translation. Because the adequacy of item translation in the adaptation of a measuring instrument from one language to another is covered in depth in Chapter 5, this volume, these issues are not addressed here.

CONCLUSION In this chapter, I have addressed important issues that researchers should consider when their research endeavors include testing for differences across cultural groups. Clearly, these principles apply in testing across international boundaries as well as within national boundaries. Typically, the primary focus of substantive research is to determine the extent to which there is evidence of differences between the means of selected variables across groups. However, as emphasized in this chapter, such inquiry assumes that the measuring instrument is operating equivalently across all groups under study, an assumption that is not always tenable. To the extent that the assessment measure is not operating in exactly the same way across samples, any interpretation of mean difference findings will be dubious, at best. Although the condition of instrument equivalence is always a critically important consideration in multigroup comparisons, it is particularly so when the comparative groups are drawn from different cultural or subcultural bases. As noted in this chapter, many different aspects of the measurement process can weaken the assumed equivalence of the instrument through the introduction of bias, a condition that renders scores to be differentially valid across groups.

REFERENCES Bentler, P. M. (1978). The interdependence of theory, methodology, and empirical data: Causal modeling as an approach to construct validation. In D. B. Kandel (Ed.), Longitudinal research on drug use: Empirical findings and methodological issues (pp. 267–302). New York, NY: Wiley. Bentler, P. M. (2005). EQS 6 structural equations program manual. Encino, CA: Multi­ variate Software. testing instrument equivalence     

139

Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford Press. Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic concepts, applications, and programming. Mahwah, NJ: Erlbaum. Byrne, B. M. (2006). Structural equation modeling with EQS: Basic concepts, applications, and programming (2nd ed.). Mahwah, NJ: Erlbaum. Byrne, B. M. (2008). Testing for multigroup equivalence of a measuring instrument: A walk through the process. Psicothema, 20, 872–882. Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and programming (2nd ed.). New York, NY: Routledge/Taylor & Francis. Byrne, B. M. (2011). Structural equation modeling with Mplus: Basic concepts, applications, and programming. New York, NY: Routledge/Taylor & Francis. Byrne, B. M. (2012). Choosing SEM computer software: Snapshots of LISREL, EQS, AMOS, and Mplus. In R. Hoyle (Ed.), Handbook of structural equation modeling (pp. 307–324). New York, NY: Guilford Press. Byrne, B. M., & Campbell, T. L. (1999). Cross-cultural comparisons and the presumption of equivalent measurement and theoretical structure: A look beneath the surface. Journal of Cross-Cultural Psychology, 30, 555–574. http://dx.doi. org/10.1177/0022022199030005001 Byrne, B. M., & Shavelson, R. J. (1987). Adolescent self-concept: Testing the assumption of equivalent structure across gender. American Educational Research Journal, 24, 365–385. http://dx.doi.org/10.3102/00028312024003365 Byrne, B. M., Shavelson, R. J., & Muthén, B. (1989). Testing for the equivalence of factor covariance and mean structures: The issue of partial measurement invariance. Psychological Bulletin, 105, 456–466. http://dx.doi.org/10.1037/ 0033-2909.105.3.456 Byrne, B. M., & van de Vijver, F. J. R. (2014). Factorial structure of the Family Values Scale from a multilevel-multicultural perspective. International Journal of Testing, 14, 168–192. http://dx.doi.org/10.1080/15305058.2013.870903 Byrne, B. M., & Watkins, D. (2003). The issue of measurement equivalence revisited. Journal of Cross-Cultural Psychology, 34, 155–175. http://dx.doi.org/10.1177/ 0022022102250225 Cheung, G. W., & Rensvold, R. B. (2000). Assessing extreme and acquiescence response sets in cross-cultural research using structural equation modeling. Journal of Cross-cultural Psychology, 31, 187–212. http://dx.doi.org/10.1177/ 0022022100031002003 Deregowski, J. B., & Serpell, R. (1971). Performance on a sorting task: A crosscultural experiment. International Journal of Psychology, 6, 273–281. http:// dx.doi.org/10.1080/00207597108246693 Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Mahwah, NJ: Erlbaum.

140       barbara m. byrne

Ho, D. Y. E. (1996). Filial piety and its psychological consequences. In M. H. Bond (Ed.), Handbook of Chinese psychology (pp. 155–165). Hong Kong, China: Oxford University Press. Horn, J. L., & McArdle, J. J. (1992). A practical and theoretical guide to measurement invariance in aging research. Experimental Aging Research, 18, 117–144. http://dx.doi.org/10.1080/03610739208253916 Hui, C. H., & Triandis, H. C. (1989). Effects of culture and response format on extreme response style. Journal of Cross-Cultural Psychology, 20, 296–309. http:// dx.doi.org/10.1177/0022022189203004 Johnson, T. P. (1998). Approaches to equivalence in cross-cultural and cross-national research. ZUMA-Nachrichten Spezial, 3, 1–40. Jöreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis. Psychometrika, 34, 183–202. http://dx.doi.org/10.1007/ BF02289343 Jöreskog, K. G. (1971). Simultaneous factor analysis in several populations. Psychometrika, 36, 409–426. http://dx.doi.org/10.1007/BF02291366 Kuang, Y., McKitrick, S., Yang, G., Wang, J., Guan, S., & Mankoff, R. (2010). Filial attitudes of American and Chinese undergraduate students. American Association of Behavioral and Social Sciences Journal, 14, 45–65. Leong, F. T. L., Okazaki, S., & Tak, J. (2003). Assessment of depression and anxiety in East Asia. Psychological Assessment, 15, 290–305. http://dx.doi.org/10.1037/ 1040-3590.15.3.290 Leung, P. W. L., & Wong, M. M. T. (2003). Measures of child and adolescent psycho­pathology in Asia. Psychological Assessment, 15, 268–279. http://dx. doi.org/10.1037/1040-3590.15.3.268 Little, T. D. (1997). Mean and covariance structures (MACS) analyses of crosscultural data: Practical and theoretical issues. Multivariate Behavioral Research, 32, 53–76. http://dx.doi.org/10.1207/s15327906mbr3201_3 Liu, J. H., Ng, S. H., Weatherall, A., & Loong, C. (2000). Filial piety, acculturation, and intergenerational communication among New Zealand Chinese. Basic and Applied Social Psychology, 22, 213–223. http://dx.doi.org/10.1207/ S15324834BASP2203_8 Lord, F. N., & Novick, M. R. (1968). Statistical theories of mental test scores. Reading, MA: Addison-Wesley. Marín, G., Gamba, R. J., & Marín, B. V. (1992). Extreme response style and acquiescence among Hispanics. Journal of Cross-Cultural Psychology, 23, 498–509. http://dx.doi.org/10.1177/0022022192234006 Marsh, H. W. (1992). Self Description Questionnaire (SDQ) I: A theoretical and empirical basis for the measurement of multiple dimensions of preadolescent self-concept: A test manual and research monograph. Macarthur, New South Wales, Australia: Faculty of Education, University of Western Sydney.

testing instrument equivalence     

141

Marsh, H. W. (1994). Confirmatory factor analysis models of factorial invariance: A multifaceted approach. Structural Equation Modeling, 1, 5–34. http://dx.doi. org/10.1080/10705519409539960 Marsh, H. W. (2007). Application of confirmatory factor analysis and structural equation modeling in sport and exercise psychology. In G. Tenenbaum & R. C. Eklund (Eds.), Handbook of sport psychology (3rd ed., pp. 774–798). http://dx.doi. org/10.1002/9781118270011.ch35 Marsh, H.  W., Hau, K.-T., Artelt, C., Baumert, J., & Peschar, J.  L. (2006). OECD’s brief self-report measure of educational psychology’s most useful affective constructs: Cross-cultural, psychometric comparisons across 25 countries. International Journal of Testing, 6, 311–360. http://dx.doi.org/10.1207/s15327574ijt0604_1 Maslach, C., & Jackson, S. E. (1981). Maslach Burnout Inventory manual. Palo Alto, CA: Consulting Psychologists Press. Maslach, C., & Jackson, S. E. (1986). Maslach Burnout Inventory manual (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. Matsumoto, D. (2000). Culture and psychology: People around the world (2nd ed.). Belmont, CA: Wadsworth. Meredith, W. (1993). Measurement invariance, factor analysis, and factorial equivalence. Psychometrika, 58, 525–543. http://dx.doi.org/10.1007/BF02294825 Millsap, R. E. (2011). Statistical approaches to measurement invariance. New York, NY: Routledge. Oyserman, D., & Markus, H. R. (1993). The sociocultural self. In J. Suls (Ed.), Psychological perspectives on the self: The self in social perspective (pp. 187–220). Hillsdale, NJ: Erlbaum. Selig, J. P., Card, N. A., & Little, T. D. (2008). Latent variable structural equation modeling in cross-cultural research: Multigroup and multilevel approaches. In F. J. R. van de Vijver, D. A. van Hemert, & Y. H. Poortinga (Eds.), Multilevel analysis of individuals and cultures (pp. 93–119). Mahwah, NJ: Erlbaum. van de Vijver, F. J. R. (2011). Capturing bias in structural equation modeling. In E. Davidov, P. Schmidt, & J. Billiet (Eds.), Cross-cultural analysis: Methods and applications (pp. 3–34). New York, NY: Routledge/Taylor & Francis. van de Vijver, F. J. R., & Leung, K. (1997). Methods and data analysis for cross-cultural research. Thousand Oaks, CA: Sage. van de Vijver, F. J. R., & Leung, K. (2000). Methodological issues in psychological research on culture. Journal of Cross-Cultural Psychology, 31, 33–51. http:// dx.doi.org/10.1177/0022022100031001004 van de Vijver, F., & Tanzer, N. K. (2004). Bias and equivalence in cross-cultural assessment: An overview. European Review of Applied Psychology, 47, 263–279. Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement equivalence literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3, 4–70. http://dx.doi. org/10.1177/109442810031002

142       barbara m. byrne

Watkins, D., & Cheung, S. (1995). Culture, gender, and response bias: An analysis of responses to the Self Description Questionnaire. Journal of Cross-Cultural Psychology, 26, 490–504. http://dx.doi.org/10.1177/0022022195265003 Widaman, K. F., & Reise, S. P. (1997). Exploring the measurement invariance of psychological instruments: Applications in the substance use domain. In K. J. Bryant, M. Windle, & S. G. West (Eds.), The science of prevention: Methodological advances from alcohol and substance abuse research (pp. 281–324). Washington, DC: American Psychological Association. Yue, X., & Ng, S. H. (1999). Filial obligations and expectations in China: Current views from young and old people in Beijing. Asian Journal of Social Psychology, 2, 215–226. http://dx.doi.org/10.1111/1467-839X.00035

testing instrument equivalence     

143

III Methodological Challenges

7 MIXED-METHODS RESEARCH: INTEGRATING QUALITATIVE AND QUANTITATIVE APPROACHES TO THE PSYCHOLOGICAL STUDY OF CULTURE MARINA DOUCERAIN, SYLVANNA VARGAS, AND ANDREW G. RYDER

The study of culture has grown increasingly central to North American psychology over the past several decades as ethnocultural diversity has increased. Confusingly, however, this work is often published within isolated literatures: cross-cultural psychology, several flavors of cultural psychology, multicultural counseling, and ethnic minority psychology. Psychological anthropologists and transcultural psychiatrists also pose broadly similar questions. Although these subdisciplines represent particular intellectual communities with particular histories, a major reason for continuing separation is methodology— in particular, the distinction between quantitative and qualitative methods (Ritsher, Ryder, Karasz, & Castille, 2002). The specific questions that engage these communities help dictate the preferred approach, as do philosophical commitments to various degrees of universalism and relativism. Choice of methods is best made on the basis of questions asked, but such flexibility is rare in practice; rather, institutional http://dx.doi.org/10.1037/14940-008 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

147

pressures privilege certain approaches over others. Psychologists too often begin with a method and search for questions that can be asked using the method—in effect, putting the epistemological cart before the ontological horse (Martin & Sugarman, 1999). Quantitative methods dominate the major journals, academic departments, and training opportunities for research psychologists (Richardson, 1996), an emphasis carried over to many of the subfields concerned with culture in psychology. Unfortunately, sole reliance on quantitative methods happens at the expense of contextualized and historically situated perspectives (Shweder & Sullivan, 1993). For this reason, most other disciplines with a concern for culture emphasize qualitative methods. In any case, the current dominance of quantitative methods in mainstream psychology research obscures its qualitative foundations. For example, the scientist–practitioner model of clinical psychology promotes the mutual influence of scientific research and clinical practice, with neither being sufficient alone. Clinical experience is a crucial element in framing hypotheses and interpreting findings, processes that are not themselves subject to strict scientific rules (Polanyi, 1968; Ritsher et al., 2002). In psychopathology, the traditional division of symptoms into particular syndromes, although subject to modification as a result of quantitative findings, is based on careful observation and phenomenological work carried out during the late 19th and first half of the 20th century. Similarly, social psychology would be inconceivable in the absence of hypotheses grounded in observations of the social world, in effect using the researcher’s own life experience as nonsystematic qualitative investigation. Even brain research has progressed in part through qualitative observations of individual participants. In short, although psychological research is often portrayed as a purely objective form of inquiry, this idealized view bears little resemblance to the actual practice of science (Woolgar, 1988). At the same time, a solely qualitative perspective can obscure quantitative aspects of the research, leading to the same exaggerated loyalty to methods over questions. Qualitative work often contains quantitative elements, data-driven iterative hypothesis testing, and other features usually associated with the quantitative approach. Often, qualitative studies make claims about magnitude, difference, and so on, although rarely using formal quantitative procedures; some go further and incorporate careful reliability checks and counting of important variables. Moreover, just as quantitative studies can be used to study purely local problems, qualitative studies have served as the basis for generalizable claims (Ritsher et al., 2002). Our position is that although individual questions might be better suited to one approach and different research teams might be better trained to emphasize one approach, the field would benefit by a more thorough engagement with both approaches. 148       doucerain, vargas, and ryder

The argument for engaging with both forms of research is rooted in Wilhelm Wundt’s 19th-century formulation of two approaches to psychology (Ritsher et al., 2002). His first psychology emphasized the traditional laboratory experiment for the study of lower-level processes. In contrast, his second psychology—had it been fully pursued—would have involved observational and ethnographic methods to study higher level sociocultural processes (Cole, 1996). Cultural psychology has belatedly emerged as a hybrid of these two approaches, albeit not always knowingly. This subdiscipline has embraced the central idea that culture and mind (or self, or psyche) mutually constitute each other—they “make each other up” (Shweder, 1990, p. 1). More recently, this idea has been extended to include the brain, so that culture–mind–brain might best be understood as a single system with three levels (Ryder, Ban, & Chentsova-Dutton, 2011). If Wundt’s first psychology focuses on mind–brain aspects of this system, his second psychology focuses on mind–culture. Consideration of the whole system requires both approaches (Harré & Moghaddam, 2012). In his argument for peaceful coexistence between these two different approaches to psychology, Taylor (1973) described three levels of psychological research. Level 1 involves infrastructural domains, such as neuro­biology (i.e., brain), and is best apprehended through the classical scientific approach. At the other end, Level 3 involves fully motivated performance in context (i.e., culture) and is best apprehended through a hermeneutic approach. In between, Level 2 involves formalized competencies, such as particular cognitive skills (i.e., mind), and requires a mixed approach. With psychology as a hub science (Cacioppo, 2007) that engages with the entire culture–mind– brain system, most often emphasizing the mind level and its interconnections, a serious consideration of mixed-methods research is warranted. In our view, this claim goes beyond the requirements for what is traditionally identified as cultural research in psychology. Although mixed-methods research approaches may benefit psychology more broadly, it is in the various subfields that engage in various ways with culture that the need is most keenly felt. For that reason, and in keeping with the focus of this volume, we provide in this chapter a brief overview of mixedmethods research approaches for the psychological study of culture. We have therefore selected a group of recent published studies that take some kind of mixed-methods approach and use these studies to illustrate our arguments. Rather than bogging the reader down in the details of these studies, we focus instead on conceptual issues, providing a brief synopsis of each study in an accompanying exhibit. In the first section of this chapter, we begin by defining mixed-methods research before turning to a consideration of its philosophical underpinnings; in the second section, we examine why one might consider mixed-methods research and then review a typology of ways mixed-methods research     

149

in which such research might be done. We conclude with a brief consideration of how the psychological study of culture could be transformed by a serious engagement with mixed-methods approaches. MIXED-METHODS RESEARCH: WHAT IS IT? Defining Mixed-Methods Research A simple definition of mixed-methods research (MMR) is “research that involves collecting, analyzing, and interpreting quantitative and qualitative data in a single study or in a series of studies that investigate the same underlying phenomenon” (Leech & Onwuegbuzie, 2009, p. 267). In other words, MMR involves combining qualitative and quantitative ingredients in the study of a common phenomenon. This seemingly simple beginning is soon challenged, however, by the observation that qualitative and quantitative research might not fall neatly into two distinct categories. Studies in which qualitative data are first thematically coded and then quantitized for statistical analysis exemplify the issue. Karasz, Garcia, and Ferri (2009; see Exhibit 7.1), for example, collected in-depth narratives describing the experience of depression among primary care patients. They then coded this qualitative data by creating dichotomous ratings of dimensions of illness representation EXHIBIT 7.1 Synopsis of Karasz, Garcia, and Ferri (2009): “Conceptual Models of Depression in Primary Care Patients: A Comparative Study” Goals. To compare European Americans to Hispanics and African Americans in conceptions of depression, through (a) quantitative evaluation of whether the biopsychiatric model is more common in European Americans and (b) qualitative examination of the influence of sociocultural context. Sample. The sample included 74 participants in primary care with major depression from three ethnocultural groups (n = 15 African Americans, n = 23 Hispanics, and n = 36 non-Hispanic Whites). Procedure. Participants provided in-depth narratives addressing the five dimensions of the Illness Representation Model (IRM) and answered general life history questions. Data analysis. Quantitative analysis involved the following: (a) quantitization by rating themes of the IRM as present versus absent and (b) chi-square tests of difference in distributions between ethnocultural groups. Qualitative analysis of the data involved thematic analysis of interviews based on a coding scheme used in previous work, with the creation of new codes. Qualitative and quantitative results were presented in parallel, organized by the five dimensions of the IRM. Findings. European Americans were statistically more likely to endorse a bio­ psychiatric model of depression compared with Hispanics and African Americans. Qualitative analyses revealed variations in symptom report, causal attribution, treatment preferences, and impact of depression.

150       doucerain, vargas, and ryder

for each narrative, a step that starts to blur the boundaries between qualitative and quantitative orientations. Finally, they computed chi-square tests of differences between ethnic groups from these ratings, a clearly quantitative procedure. This entire process straddles the demarcation between qualitative and quantitative in a way that defies unambiguous labels for each component. Allwood (2012) argued that, in any case, the distinction between these two approaches is problematic at best, adding that the qualitative approach in particular is characterized by striking heterogeneity. The extent to which qualitative researchers value the generalizability of results is an example— some researchers aim at large-scale generalization, whereas others totally reject the very possibility. As a result, the status of quantitative research also becomes more complicated as it can no longer be assumed to simply be the converse of qualitative research. In addition to the issue of heterogeneity, Allwood (2012) argued that all research components include both qualitative and quantitative aspects. For example, even the most positivist, number-driven study results include qualitative elements such as the formulation of a specific research question and the interpretation of results. Qualitative analyses that emphasize reporting the frequency of themes in the results are a converse example. For instance, Mendenhall and Jacobs (2012; see Exhibit 7.2) collected life history narratives among Mexican immigrant women suffering from Type 2 diabetes. The data were thematically coded for life stressors and a substantial proportion of the results involved commenting on the frequency of each theme in the sample. The authors even chose to include a bar chart displaying the frequency of EXHIBIT 7.2 Synopsis of Mendenhall and Jacobs (2012): “Interpersonal Abuse and Depression Among Mexican Immigrant Women With Type 2 Diabetes” Goals. To investigate the relation between life stressors and depression among diabetic Mexican immigrant women in the United States. Sample. The sample included 121 Mexican women seeking diabetes care in a safety-net clinic in Chicago. Procedure. Participants (a) provided in-depth answers to open-ended questions; (b) completed measures assessing depression, diabetes distress, and acculturation; and (c) provided a blood sample. Data analysis. Quantitative analysis involved two steps: (a) quantitization of various life stressors as present versus absent and (b) logistic regressions predicting depression as a function of life stressors. Qualitative analysis involved a grounded analysis of life stressors in the narrative answers. Findings. Qualitative analysis revealed seven life stressors: interpersonal abuse, health, family, neighborhood violence, immigration status, work, and feeling socially detached. Quantitative results showed that interpersonal abuse was a significant predictor of depression.

mixed-methods research     

151

all life stressors identified. This style of reporting is not uncommon among qualitative studies, supporting the notion that even qualitative results usually comprise quantitative elements. Although this blurred distinction between qualitative and quantitative approaches makes MMR seem like a natural and logical course of action, it also questions the location—and perhaps existence—of its boundaries. Where does MMR start, and where does it stop? What is MMR, and what is not? Is MMR defined solely in terms of its qualitative and quantitative constituents, or does it need its own category? Even if we accept the traditional distinction between qualitative and quantitative research, as molded by historical, political, and discipline-specific conventions, the question of what counts as MMR still remains. The simple definition, “mixing qualitative and quantitative data,” as stated above, may be far too broad to be truly useful. For example, would a survey study in which all data are analyzed quantitatively, but where one or two excerpts from openended responses are mentioned in the discussion, count as a mixed-methods study? Conversely, what about predominantly qualitative mixed-methods studies in which the quantitative component is limited to a few descriptive statistics? For example, Beagan, Etowa, and Bernard (2012; see Exhibit 7.3) conducted a study on racism and spirituality among African-heritage Nova Scotian women. The qualitative thematic analysis of in-depth interviews dominated the study, whereas the quantitative component was limited to reporting mean scores of items from scales assessing experiences of racism, depression, and coping. Although the authors characterized this study as EXHIBIT 7.3 Synopsis of Beagan, Etowa, and Bernard (2012): “‘With God in Our Lives He Gives Us the Strength to Carry on’: African Nova Scotian Women, Spirituality, and Racism-Related Stress” Goal. To examine the impact of racism on mental health and the role of spirituality as a coping mechanism among African Nova Scotian women. Sample. The sample included 50 women self-identifying as African-heritage Nova Scotian. Procedure. Participants (a) completed measures assessing racism, depression, and coping; and (b) provided in-depth answers to open-ended questions about racism, stress, and spirituality. Data analysis. Quantitative analysis involved computing descriptives for questionnaire scores. Qualitative analysis involved a thematic analysis of the narrative answers. Qualitative and quantitative results were presented in parallel, organized by the themes that emerged from the qualitative analysis. Findings. The quantitative results showed that all participants suffered from considerable racism-related stress. In the qualitative results, spirituality emerged as a central coping mechanism, providing women with community and ways of making meaning of their difficult experiences.

152       doucerain, vargas, and ryder

mixed methods, we believe a qualitative label would have been just as appropriate. This study is located in a gray area that begs examination of what counts as MMR. Historically, the MMR movement has emphasized eclecticism and methodological openness, which would not favor a strict threshold of how much mixing is enough to count as MMR. There certainly is value to being comprehensive, but at the same time, if MMR is to stand as an approach that is distinct from existing ones and valuable in itself, clear defining boundaries are needed. Unfortunately, as we show below, there is still considerable debate about the best definition. Philosophical Underpinnings of Mixed-Methods Research The early history of MMR has been marked by the paradigm war between proponents of qualitative and quantitative research traditions. Purists from both camps saw their respective positions as incompatible and incommensurate with that of the other camp—qualitative and quantitative methods cannot and should not be mixed (see Johnson & Onwuegbuzie, 2004, for a description of both positions). In other words, MMR is doomed to fail because of untenable theoretical foundations. Although the majority of the MMR movement has moved beyond this view, the incompatibility debate has not really been adequately resolved (Greene, 2008) and we need to address it here. It may well be that quantitative and qualitative paradigms in their pure form are incompatible. Indeed, from a purely philosophical standpoint, the question remains as to how we can combine a perspective that subscribes to objectivity, unbiased and value-free research, and the separation between the researcher and the researched, with a perspective that emphasizes subjectivity, researcher context, value-laden research, and the inseparability between the researcher and the researched. (Bergman, 2011, p. 272)

However, such concerns might be much less of an issue for the social science researcher wanting to use mixed methods to answer a complex research question. First, ontologies and methods are not synonymous. Although specific methods have traditionally been linked to a specific ontological paradigm, the association between methods and paradigms is “neither sacrosanct nor necessary” (Johnson & Onwuegbuzie, 2004, p. 15). As Bergman (2010) pointed out, qualitative and quantitative analysis techniques do not necessitate a particular view of the nature of reality, privilege a specific research theme and how to research it, or determine the truth value of data or the relationship between researchers and their research subject. (p. 173) mixed-methods research     

153

In this view, it is possible to conduct MMR—which by definition is a mixing of methods—while still retaining a dominant worldview that is more sympathetic to one paradigm or another. The mixed-methods study by Castro and Coe (2007; see Exhibit 7.4) on tradition and perinatal behaviors among Latinas exemplifies this possibility. The authors adopted a predominantly realistic orientation, although not explicitly. They articulated several concerns about the rigorousness of qualitative research and formulated clear research hypotheses. Nevertheless, they successfully integrated quantitative and qualitative components within that hypothesis-testing framework. The task might be more difficult for a purist of either camp because of the traditional association between methods and paradigms, but it is doubtful that a paradigm purist would attempt to conduct MMR in the first place. Second, paradigm assumptions and the like might not play such a central role in actual practice. Greene (2008) asked the question “What actually does influence inquirers’ methodological decisions in practice?” (p. 11) and concluded that “paradigm assumptions were rarely cited as important practical influences” (p. 11). In other words, paradigmatic considerations may play a larger role in textbooks than in actual practice. In our small sample of mixed-methods studies investigating culture/ethnicity and health, only two of eight studies explicitly mention a paradigm: Beagan and colleagues (2012) referred to a naturalistic approach; and Rosen, Miller, Nakash, Halperin, EXHIBIT 7.4 Synopsis of Castro and Coe (2007): “Traditions and Alcohol Use: A Mixed-Methods Analysis” Goals. (1) To test whether traditionalism predicts more self-care beliefs during pregnancy and more abstinence from alcohol in rural Hispanic women. (2) To unpack the meaning of traditionalism. Sample. The sample included 77 young Hispanic women eligible for perinatal health services in rural Arizona. Procedure. Participants (a) provided in-depth answers to open-ended questions about family traditions and rural lifestyles; and (b) completed measures assessing family traditionalism, rural lifestyles, acculturation, folk beliefs, pregnancy self-care beliefs, and alcohol abstinence. Data analysis. Quantitative analysis of the data involved three steps: (a) quantitization of interview data through coding the affective emphasis of qualitative themes, (b) logistic regressions predicting pregnancy self-care beliefs and alcohol abstinence as a function of scale scores, and (c) inclusion in logistic regression models of affectively coded thematic variables. Qualitative analysis involved a semi-automated process of inductive theme generation (word frequency and co-occurrence) and thematic coding. Data were then axially coded: Researchers assigned a degree of intensity or emphasis to each identified category. Correlations among thematic variables were then computed. Findings. Quantitative results revealed a lack of predictive strength for the three scales, but cross-methods regression showed that two qualitatively derived thematic variables predicted pregnancy self-care beliefs and alcohol abstinence.

154       doucerain, vargas, and ryder

and Alegría (2012; see Exhibit 7.5) referred to a pragmatic approach. Both articles mentioned it in passing; neither clarified what theoretical assumptions are entailed by the paradigm chosen, nor did they discuss how paradigmatic considerations influenced the study practically. The reader can infer these influences, especially for the study by Beagan et al., but there is no clear explication. However, the fact that paradigms may not be important practical influences does not mean that they should be ignored or that their assumptions are made more compatible by their marginal role in actual practice. Nor does it mean that one should refrain from embracing both alternatively, attempting to reconcile them, or challenging them. As Creswell (2009) rightly pointed out, “sometimes they may be in tension, and such tension is good. The dialectic between opposing ideas can contribute to new insights and new understandings” (p. 102). As mentioned earlier, the majority of mixed-methods researchers resolved the paradigm war by adopting pragmatism as their philosophical stance, thus giving rise to the third paradigm. Pragmatism is ideally suited for MMR for several reasons. First, its recognition of both the natural–physical world and the emergent social–psychological world pays tribute to both realism and constructivism in a way that does not attribute more importance to one perspective or another (Johnson & Onwuegbuzie, 2004). Moreover, its emphasis on an EXHIBIT 7.5 Synopsis of Rosen, Miller, Nakash, Halperin, and Alegría (2012): “Interpersonal Complementarity in the Mental Health Intake: A Mixed-Methods Study” Goal. To investigate the relation between dimensions of social identity (race/ ethnicity, sex, and age) and complementarity between provider and client during the intake session. Sample. The sample included 44 providers and 114 clients in eight outpatient mental health clinics in the Northeast (ethnocultural distribution: 53% Latino, 36% non-Latino White, 11% African American). Procedure. Intake sessions were videotaped. Participants then provided in-depth answers to open-ended questions about the intake session (e.g., about provider– client rapport or the role of sociocultural factors in care) and completed a demographic questionnaire. Data analysis. Quantitative analysis involved (a) coding videotapes according to existing guidelines grounded in interpersonal theory, (b) computation of a total complementarity score, and (c) multilevel regression using demographics to predict complementarity. Qualitative analysis was based on interviews with client–provider dyads in the top and bottom quartiles on complementarity. Findings. Quantitative results showed that complementarity for African American clients was higher with a White versus Latino provider, and with an age match between client and provider. Qualitative analysis showed that in high-complementarity dyads, client and practitioner gave consistent descriptions of concerns, expectations, and importance of topics discussed.

mixed-methods research     

155

organism–environment transactional view of human behavior provides a way in which realist and constructivist stances can be reconciled (Greene, 2008)— or at least made to peacefully coexist. This potential for conciliation is even furthered by the pragmatic epistemological stance that “knowledge is viewed as being both constructed and based on the reality of the world we experience and live in” (Johnson & Onwuegbuzie, 2004, p. 14). Second, pragmatism advocates eclecticism and pluralism, a stance that fits well with the creative endeavor that MMR can represent. Third, pragmatism focuses on the empirical and practical consequences of ideas or theories. In MMR, this stance has been interpreted as a focus on the best way to answer the research questions at hand (Bazeley, 2009): “Research methods and designs must be judged on the basis of what they can accomplish” (Karasz & Singelis, 2009, p. 910). Although pragmatism should not be reduced to an oversimplified, atheoretical, aphilosophical statement, we believe it offers a desirable third alternative to the purely realist and constructivist positions and is a paradigm ideally suited for MMR. However, although the importance and advisability of pragmatism are discussed in depth in the theoretical literature on MMR, it is worth reiterating that of the eight empirical articles reviewed, only one declared a pragmatic orientation—and did so only in passing. MIXED-METHODS RESEARCH: WHY AND HOW? Rationales for Mixed-Methods Research Over the years, theorists have identified a number of rationales for conducting MMR. One of the most prominent and comprehensive is that developed by Greene, Caracelli, and Graham (1989), who identified five main rationales: (a) triangulation, seeking convergence and corroboration of results obtained from different methods but investigating the same phenomenon; (b) complementarity, seeking elaboration, enhancement, illustration, and clarification of the results from one method by using the results obtained with the other method; (c) development, using the results from one method to inform or develop the other method; (d) initiation, discovering paradoxes and contradictions, or recasting research questions from one method by using that from the other method; and (e) expansion, extending the range and breadth of investigation by using different methods for different inquiry components. In our small sample of articles on culture/ethnicity and health reviewed here, rationales for conducting MMR are formulated without reference to any existing typology or nomenclature, with the exception of triangulation. Nonetheless, although most studies offer idiosyncratic and relatively broad rationales, we can interpret them in light of this framework. 156       doucerain, vargas, and ryder

In their review of the literature, Greene and colleagues (1989) demonstrated that although many studies identify triangulation as a rationale, it is rarely appropriate in the classical sense of triangulation as converging evidence (see definition above). They contended—and we agree with this view— that only independent data sources can achieve triangulation. In our sample, two studies cited triangulation as a rationale for conducting MMR: Deacon, Pendley, Hinson, and Hinson (2011; see Exhibit 7.6); and Tsai, Morisky, Kagawa-Singer, and Ashing-Giwa (2011; see Exhibit 7.7). We believe this rationale is unjustified in the latter case, however. Tsai and colleagues collected quantitative (standardized scales) and qualitative (semi-structured interviews) data in the same sample of first-generation Chinese American women with breast cancer. Note that qualitative and quantitative data are not independent from one another. In contrast, Deacon and colleagues collected quantitative data (survey) in a sample of Chickasaw community members and qualitative data (narratives) in a different sample of employees of the Chickasaw Nation Division of History and Culture. Both sets of data were independent and were brought to bear on the same phenomenon, namely the meaning of healthy families in the Chickasaw tribe, thus justifying a triangulation rationale. In contrast, the other seven studies used the same participants to collect quantitative and qualitative data, a design that precludes triangulation as a possible rationale. The rationale of complementarity was identifiable in most studies in our sample. For example, Tsai and colleagues (2011) aimed to use qualitative methods to “interpret the quantitative findings from the cultural perspective” (p. 3384), a rationale explicitly defined as triangulation but perhaps better EXHIBIT 7.6 Synopsis of Deacon, Pendley, Hinson, and Hinson (2011): “Chokka-Chaffa’ Kilimpi’, Chikashshiyaakni’ Kilimpi’: Strong Family, Strong Nation” Goal. To answer “What is the definition of a strong and healthy Chickasaw family?” Sample. The sample included 230 Chickasaw community members and 7 Chickasaw expert informants. Procedure. Quantitative data collection consisted either of a brief survey (n = 115) or a card sort task (n = 115) designed to assess conceptions of strong families. Qualitative data collection consisted in the seven employees providing narrative definitions of the meaning of strong families. Data analysis. Quantitative analysis involved correlating age of participants with item frequencies. Qualitative analysis involved a thematic content analysis of the narrative definitions. Findings. Qualitative and quantitative results were presented in parallel, organized by themes. Quantitative results showed that greater age predicts tendency to associate family strength with cultural traditions and American Indian ethnicity. Qualitative analysis revealed that strong families are seen as cohesive, extended, grounded in community, proeducation, nonmaterialistic, and valuing tradition.

mixed-methods research     

157

EXHIBIT 7.7 Synopsis of Tsai, Morisky, Kagawa-Singer, and Ashing-Giwa (2011): “Acculturation in the Adaptation of Chinese-American Women to Breast Cancer: A Mixed-Method Approach” Goal. To examine the relation between acculturation and breast cancer survivorship among Chinese immigrant women. Sample. The sample included 107 first-generation Chinese American women diagnosed with breast cancer. Procedure. Participants completed a quantitative survey, comprising scales assessing acculturation, health-related locus of control, social support, and life stress. Some participants (n = 16) took part in a semistructured qualitative interview exploring cancer, cultural practices, and identity. Data analysis. Quantitative analysis consisted of correlations and one-way analyses of variance between level of acculturation and other quantitative variables. Qualitative analysis involved inductive thematic analysis of interviews and participant check to confirm correctness of interpretation. Findings. Quantitative results, presented first, showed that acculturation was related to health beliefs, social support, and life stress. Qualitative analysis revealed that Chinese cultural beliefs significantly influence how Chinese American women make sense of, and adjust to, their illness.

labeled as complementarity. Other examples are Castro and Coe’s (2007) goal to mix methods in order to obtain “a more complete understanding of complex cultural constructs” (p. 271) and Rosen and colleagues’ (2012) statement that the “qualitative portion served primarily to elucidate the quantitative findings through exploration of participant narratives” (p. 189). The rationale of development was very rare in our sample. The only partial example is the case of Rosen et al. (2012), who used their quantitative results to decide which qualitative data to analyze. Namely, they examined only the in-depth interviews of dyads that obtained a complementarity score in the lower and upper quartiles of the distribution. Thus, to some extent, the quantitative results informed one aspect of the qualitative analysis. Karasz, Garcia, and Ferri (2009) provided the only example of initiation as a rationale. One of their explicit goals was to use qualitative results to generate hypotheses for future studies. The other studies did not seem to mix methods with the aim of identifying paradoxes or contradictions between qualitative and quantitative results. None of the eight studies in our sample explicitly referred to the rationale of expansion as such. However, the designs of several studies seem congruent with this goal. For example, Rosen et al. (2012) examined interpersonal complementarity through quantitative analysis of mental health intake videos. The addition of interviews of dyads that scored very low or very high in interpersonal complementarity broadened the scope of investigation by accounting for participants’ experiences and impressions during the intake session. In 158       doucerain, vargas, and ryder

doing so, the authors aimed to “add depth” (p. 187) to the quantitative analysis. Similarly, Castro and Coe (2007) entered qualitatively derived variables into logistic regressions to “add explanatory information” (p. 278). Overall, however, we reiterate that for most studies, rationales were only implied or vaguely defined. With rare exceptions, the authors did not discuss what goal was served by the choice of MMR. In addition, they did not ground their rationales in the existing theoretical literature on MMR—unfortunate, as most studies would benefit from being explicitly anchored in existing MMR theoretical frameworks. The rationales identified by Greene et al. (1989) are important, but there are additional reasons to use mixed methods when conducting cultural research. Karasz and Singelis (2009) identified two problems relevant for psychologists: the problem of culture and the problem of transferability. The first problem refers to the notion that culture has usually been treated as a categorical independent variable, mostly for the sake of establishing comparisons between groups. In such designs, “the contents, processes, and structures that constitute culture are not specified” (Karasz & Singelis, 2009, p. 913). Combining qualitative methods, with their emphasis on constructed meaning with this more postpositivist characterization of culture, can highlight identified psychological differences as more meaningful or challenge their validity altogether. In either case, MMR has the potential to restore complexity and content to culture without ignoring important large-scale cultural group differences. Three studies in our sample were concerned with this issue. In their examination of the experience of Chinese immigrant women living with breast cancer, Tsai and colleagues (2011) showed an explicit concern with the cultural meaning of illness. They discussed the role that the three prominent Chinese philosophical traditions (Taoism, Buddhism, and Confucianism) might play in shaping this meaning and suggested that “coping with breast cancer may draw on the cultural beliefs about life and illness” (p. 3384). Through in-depth interviews of Chinese immigrant women, the authors explored these beliefs and how they might shape not only the meaning of illness but also how one copes with breast cancer. Castro and Coe’s (2007) investigation of the construct of traditionalism was also concerned with cultural meaning. In their view, qualitative approaches are essential in restoring content and complexity to culture: “the measurement of complex cultural constructs, such as traditionalism, in the form of traitlike variables tends to decontextualize these constructs from their full cultural meaning” (p. 270). In keeping with this consideration, their article probes the meaning of traditionalism among rural Latinas, from an insiders’ perspective. Zukoski, Harvey, Oakley, and Branch’s (2011; see Exhibit 7.8) motivations were similar. Specifically, they investigated how the typically Latino cultural concepts of sympatía, confianza, and familismo play out in power and decision mixed-methods research     

159

EXHIBIT 7.8 Synopsis of Zukoski, Harvey, Oakley, and Branch (2011): “Exploring Power and Sexual Decision Making Among Young Latinos Residing in Rural Communities” Goals. To examine (a) definitions of power among young rural Latinos/as and (b) the association of relationship power with decision making about contraceptive use in this population. Sample. The sample included 58 Latinos/as, ages 18 to 25, living in rural U.S. counties (n = 29 men, n = 29 women). Procedure. Participants (a) answered demographics and scales assessing acculturation and relationship control and (b) provided in-depth answers to open-ended questions about power. Data analysis. Quantitative analysis involved computing descriptive statistics for standardized scales. Qualitative analysis involved an inductive content analysis of the narrative answers. Findings. Quantitative and qualitative results were presented in parallel, organized by the themes that emerged from qualitative analysis. Quantitative results showed that, overall, men have more power in relationships. Also, the majority of participants believed that both partners are involved in decisions to take contraceptive measures and use condoms. Qualitative results revealed that relationship power is described mostly in terms of decision-making dominance and relationship control. Some participants also characterized power in terms of joint decision making and equality.

making in sexual relationships among rural Latinos. In their discussion, they also interpreted their results in light of changing cultural norms in this population (immigrant Latina women becoming more assertive and independent and demanding a more egalitarian status in relationships; Zukoski et al., 2011, p. 456). The second problem identified by Karasz and Singelis (2009) is that most psychological theories have been developed by White, Western, middleclass scholars and that they might not be relevant in other cultural contexts. “It is difficult to use exclusively quantitative methods to generate meaningful data about the degree to which cross cultural theories actually make sense across cultural contexts” (p. 913). MMR can solve this transferability issue by including qualitative methods to explore how meaning is constructed in various cultural context and to what extent it is commensurate with “mainstream” (i.e., White, Western, middle class) psychological constructs. This exploration of mainstream meaning construction is a central rationale in Karasz and colleagues’ (2009) study of cultural models of depression, in which they investigated cross-cultural differences in people’s conceptions of depression among three ethnocultural groups: African Americans, Latinos, and non-Hispanic Whites. An important component of their work is to identify contrasts in what it means to be depressed for members of these groups. In that sense their study is concerned with the transferability of depression as a cultural construct. Deacon and colleagues (2011) followed a similar strategy, 160       doucerain, vargas, and ryder

although their study did not involve direct cross-cultural comparisons. In this case, the comparison is implied between their results and mainstream conceptualizations of healthy families. They contended that the existing literature is replete with the latter, and therefore they decided to explore the construct of healthy families in a specific American Indian tribe: the Chickasaw Nation. As such, MMR need not involve samples from different cultural groups to have the issue of cultural transferability as a rationale; the comparison with mainstream ideas can be implied. A third potential reason becomes evident through the recognition of the disciplinary roots of research on culture-related phenomena. One strand of research comes from anthropology, a discipline that has been characterized by its qualitative orientation. A second comes from psychology, which has embraced a quantitative approach, at least over the past several decades. As such, MMR is ideally suited for studies that seek to investigate culture in a deep and holistic fashion in the sense that they combine the strengths of two complementary research traditions and fields. Typologies of Mixed-Methods Research As MMR became more prominent, scholars argued that there is a need to develop a typology of MMR designs (Leech & Onwuegbuzie, 2009). Here, we present one approach that is both comprehensive and relatively prominent. In this typology, Leech and Onwuegbuzie posited three dimensions along which most mixed-methods designs are located, namely (a) level of mixing— whether the research partially versus fully mixes qualitative and quantitative approaches; (b) time orientation—whether the qualitative and quantitative components of the research occur at the same time or concurrently, versus one after the other or sequentially; and (c) emphasis of approaches—whether the qualitative and quantitative components of the research receive the same emphasis or have equal status versus one component having dominant status. They then propose a typology with eight mixed research designs that result from crossing these three dimensions in a 2 × 2 × 2 matrix. We find their framework is actually more useful and less rigid if left in its dimensional form; nevertheless, when planning to conduct MMR, all researchers are faced with choosing where along these three dimensions their research will fall. In terms of level of mixing, adopting a continuum perspective definitely seemed more appropriate than a dichotomous partial versus full mixing characterization given that studies in our sample displayed a variety of strategies. At one end of the continuum are examples of studies that adopted a really partial mixing stance—quantitative and qualitative data are collected separately, at different time points; they are analyzed separately; and results are presented in separate sections. Here, mixing essentially occurs only during the discussion. mixed-methods research     

161

Tsai and colleagues (2011) adopted this strategy in their study on Chinese American women’s experience of breast cancer. In a slightly more mixed version, quantitative and qualitative data collection and analysis take place separately, but quantitative and qualitative results are presented in parallel, in a thematic way. Deacon and colleagues (2011) exemplified this approach: The results section starts with the theme of cultural orientation, where qualitative and quantitative results are presented side by side. It proceeds in a similar fashion with the themes of Chokka-chaffa (family), Chikashsha alhiha (Chickasaw community), and so on. Studies in which qualitative and quantitative data are collected in the same session, then analyzed separately and where results are presented in parallel represent the next level on the mixing continuum observed in our sample. Zukoski and colleagues (2011) as well as Beagan and colleagues (2012) adopted this mixing strategy. Studies that build on the previous characteristics to include cross-method analyses represent the highest level of mixing in our sample. For example, both Mendenhall and Jacobs (2012) and Castro and Coe (2007) quantitized the precoded qualitative data and included it in statistical analyses (although it should be noted that in their case, results were presented separately). In terms of time orientation, six of the eight studies in our sample adopted concurrent collection of qualitative and quantitative data. Typically, doing so involved an extended interview in which participants filled out questionnaires (or answered structured questions) combined with narrative answers to open-ended questions (see, e.g., Beagan et al., 2012; Castro & Coe, 2007; Mendenhall & Jacobs, 2012; Zukoski et al., 2011). Logistically this choice is sensible, as it economizes time, presents fewer scheduling issues, and circumvents dropout problems. All of these reasons might contribute to the popularity of concurrent data collection. This time orientation precludes certain types of studies and rationales, however. For example, it forbids triangulation as a rationale because both sets of data are not independent. Development, in which one method informs the other, is also ruled out as a rationale. In our sample, emphasis of approaches also varied on a continuum, like degrees of mixing—ranging from studies that explicitly declared a quantitative emphasis (Rosen et al., 2012) to studies in which both approaches seemed relatively well-balanced (Karasz et al., 2009; Tsai et al., 2011) to studies that adopted a clearly more qualitative approach (e.g., Beagan et al., 2012; Deacon et al., 2011). In the last case, two studies comprised a minimal and simplistic quantitative component whereby the analysis was limited to descriptive statistics (Beagan et al., 2012; Deacon et al., 2011). This minimalist quantitative component brings to the forefront the potential distinction between emphasis of approaches and shallowness or quality of approaches. Researchers should not confuse questions of quality of a study or analysis 162       doucerain, vargas, and ryder

that are relevant irrespectively of any emphasis with issues of emphasis, which do not address the respective merits of qualitative and quantitative components. With respect to Leech and Onwuegbuzie’s (2009) typology, or to any typology one chooses to adopt, we would like to make two arguments. First, the choice of one design over others should primarily be governed by the research question at hand. For example, the choice of a concurrent time orientation should be driven by a specific research question rather than convenience. Different research questions call for different designs, and the choice of a tool should come in response to a well-formulated goal. It could be argued that in many ways, the formulation of the question constrains the range of sensible methods that could be used. MMR, because it combines qualitative and quantitative tools, expands the range of possible methods. Nevertheless, the requirements of first formulating a research question and then choosing the method best suited to answer that question still apply. A useful guideline in identifying research methods to answer a specific research question is to rely on what has been called the fundamental principle of MMR; that is, researchers should combine methods that have complementary strengths and nonoverlapping weaknesses (Johnson & Onwuegbuzie, 2004). Most or all of the studies in our sample had a clearly formulated research question, but we were struck by the general lack of discussion on how the research question implied or informed the chosen methods and design. The adopted research strategies always seemed sensible in light of the research question, but often one could easily imagine alternative methods or designs that would have served the research question just as well. To take a simple example, Deacon et al. (2011) aimed to explore the concept of family health in the Chickasaw tribe. They chose to collect questionnaire data and conduct in-depth interviews of key informants. This methodological decision was clearly judicious, but other possibilities abound; presumably, analyzing cultural artifacts (e.g., analyzing the content of songs, tales, cultural texts, and rituals) could also have yielded important insights. A justification of the connections between research question and research design and methods would have been desirable. In relation to this last point, note that MMR questions are facing additional requirements compared with their monomethod counterparts. Indeed, as Tashakkori and Creswell (2007) suggested, “mixed methods studies need at least one explicitly formulated mixed methods question or objective about the nature of mixing, linking, or integration” (p. 210). In our sample, only Castro and Coe’s (2007) study met this particular requirement: The authors developed and introduced an integration model, the multistage paradigm for integrative mixed-methods research. This observation echoes our comments with respect to rationales and typologies. Overall, we found that many mixed-methods research     

163

methodological considerations relevant to mixed methods are only implied in the studies we reviewed. Most authors seemed to conduct MMR without explicitly establishing or discussing a methodological framework. This specific requirement brings us to our second point: the importance of integration. Integration includes at minimum a combination of results from qualitative and quantitative strands in the conclusion, but scholars call for more genuine integration (e.g., Bazeley, 2009; Bryman, 2007). In this view, which is also our position, quantitative and qualitative components can be considered ‘integrated’ to the extent that these components are explicitly related to each other within a single study and in such a way as to be mutually illuminating, thereby producing findings that are greater than the sum of parts. (Woolley, 2009, p. 7)

We believe that in our sample, Castro and Coe (2007) achieved the highest level of integration. For example, the same data were analyzed both qualitatively and quantitatively. Open-ended interview responses were first analyzed thematically and these themes were interpreted inductively. Second, the same data were quantitized and included side by side with other quantitative data in the same logistic regression models. The various levels of processing of the qualitative data (e.g., partialing out the influence of source attribution by creating another qualitative code before conducting cross-method analyses) showed a high concern for integration. Perhaps not surprisingly, the highest level of integration was achieved in the only article that explicitly presented an integration strategy. Achieving such a level of integration is neither easy nor straightforward. Bryman (2007) identified several barriers to genuine integration, such as publication issues or the realization that when writing up their results, mixed-methods researchers sometimes end up addressing their qualitative and quantitative findings to different audiences. Nevertheless, integrating qualitative and quantitative strands so that they are mutually illuminating is a worthy goal. We believe that, ideally, integration should not be confined to the concluding paragraphs of a study and that genuine integration, as defined above, should be a prominent goal for researchers conducting MMR. Refer to Bazeley (2009) for a list of integration strategies. CONCLUSION We began this chapter by arguing that the psychological study of culture emphasizes mind–culture links within an overall concern for culture– mind–brain (see also Ryder et al., 2011). To pursue this ambitious goal, 164       doucerain, vargas, and ryder

researchers will need to engage seriously with quantitative, qualitative, and mixed-methods approaches. Sole reliance on one or the other will give an incomplete, even flawed picture. The MMR examples provided here demonstrate the incompleteness of conclusions that would have been made had the researchers relied solely on the quantitative or qualitative components of their investigations. Embracing multiple approaches and various ways of integrating them, while confronting the philosophical challenges that emerge from doing so, is necessary to ensuring that the various cultural psychologies start with questions (ontology) and proceed to methods (epistemology). Successful integration would also help to lower the barriers between subdisciplines, advancing the vision of an integrated and interdisciplinary cultural psychology (e.g., Shweder, 1990). Practical implementation of mixed-methods approaches in the psychological study of culture will require institutional changes, not merely ideological ones. Both quantitative and qualitative methods require extensive training, and there is little point in rushing to MMR if one (or more than one) of the components is conducted poorly. At a minimum, psychologists interested in culture should develop competence in understanding how to read critically both qualitative and quantitative research so as to take advantage of the knowledge generated by both. Doing so would allow for lines of inquiry to be pursued by several teams of researchers from different methodological traditions who are able to at least respond to each others’ work. Ideally, however, single research teams and some single researchers would be able to pursue MMR in a fully integrated way. The studies reviewed here represent some promising beginnings in this direction. REFERENCES Allwood, C. (2012). The distinction between qualitative and quantitative research methods is problematic. Quality & Quantity: International Journal of Methodology, 46, 1417–1429. Bazeley, P. (2009). Editorial: Integrating data analyses in mixed methods research. Journal of Mixed Methods Research, 3, 203–207. http://dx.doi.org/10.1177/ 1558689809334443 Beagan, B. L., Etowa, J., & Bernard, W. T. (2012). “With God in our lives he gives us the strength to carry on”: African Nova Scotian women, spirituality, and racismrelated stress. Mental Health, Religion & Culture, 15, 103–120. http://dx.doi.org/ 10.1080/13674676.2011.560145 Bergman, M. M. (2010). On concepts and paradigms in mixed methods research. Journal of Mixed Methods Research, 4, 171–175. http://dx.doi.org/10.1177/ 1558689810376950 mixed-methods research     

165

Bergman, M. M. (2011). The good, the bad, and the ugly in mixed methods research and design. Journal of Mixed Methods Research, 5, 271–275. http://dx.doi.org/ 10.1177/1558689811433236 Bryman, A. (2007). Barriers to integrating quantitative and qualitative research. Journal of Mixed Methods Research, 1, 8–22. http://dx.doi.org/10.1177/ 2345678906290531 Cacioppo, J. T. (2007). Psychology is a hub science. Observer, 20(8), 5, 42. Castro, F. G., & Coe, K. (2007). Traditions and alcohol use: A mixed-methods analysis. Cultural Diversity and Ethnic Minority Psychology, 13, 269–284. http://dx.doi. org/10.1037/1099-9809.13.4.269 Cole, M. (1996). Cultural psychology: A once and future discipline. Cambridge, MA: Belknap. Creswell, J. W. (2009). Editorial: Mapping the field of mixed methods research. Journal of Mixed Methods Research, 3, 95–108. http://dx.doi.org/10.1177/ 1558689808330883 Deacon, Z., Pendley, J., Hinson, W. R., & Hinson, J. D. (2011). Chokka-Chaffa’ Kilimpi’, Chikashshiyaakni’ Kilimpi’: Strong family, strong nation. American Indian and Alaska Native Mental Health Research, 18, 41–63. http://dx.doi.org/ 10.5820/aian.1802.2011.41 Greene, J. C. (2008). Is mixed methods social inquiry a distinctive methodology? Journal of Mixed Methods Research, 2, 7–22. http://dx.doi.org/10.1177/ 1558689807309969 Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11, 255–274. http://dx.doi.org/10.3102/01623737011003255 Harré, R., & Moghaddam, F. M. (2012). Psychology for the third millennium: Integrating cultural and neuroscience perspectives. Thousand Oaks, CA: Sage. Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33, 14–26. http://dx.doi. org/10.3102/0013189X033007014 Karasz, A., Garcia, N., & Ferri, L. (2009). Conceptual models of depression in primary care patients: A comparative study. Journal of Cross-Cultural Psychology, 40, 1041–1059. http://dx.doi.org/10.1177/0022022109348782 Karasz, A., & Singelis, T. M. (2009). Qualitative and mixed methods research in cross-cultural psychology: Introduction to the special issue. Journal of CrossCultural Psychology, 40, 909–916. http://dx.doi.org/10.1177/0022022109349172 Leech, N., & Onwuegbuzie, A. (2009). A typology of mixed methods research designs. Quality & Quantity: International Journal of Methodology, 43, 265–275. http://dx.doi.org/10.1007/s11135-007-9105-3 Martin, J., & Sugarman, J. (1999). The psychology of human possibility and constraint. Albany: State University of New York Press.

166       doucerain, vargas, and ryder

Mendenhall, E., & Jacobs, E. A. (2012). Interpersonal abuse and depression among Mexican immigrant women with type 2 diabetes. Culture, Medicine and Psychiatry, 36, 136–153. http://dx.doi.org/10.1007/s11013-011-9240-0 Polanyi, M. (1968). Logic and psychology. American Psychologist, 23, 27–43. http:// dx.doi.org/10.1037/h0037692 Richardson, J. (Ed.). (1996). Handbook of qualitative research methods for psychology and the social sciences. Oxford, England: Wiley-Blackwell. Ritsher, J. E. B., Ryder, A. G., Karasz, A., & Castille, D. M. (2002). Methodological issues in the study of psychopathology across cultures. In P. Boski, F. J. R. van de Vijver, & A. M. Chodynicka (Eds.), New directions in cross-cultural psychology (pp. 129–145). Warsaw, Poland: Polish Psychological Association. Rosen, D. C., Miller, A. B., Nakash, O., Halperin, L., & Alegría, M. (2012). Interpersonal complementarity in the mental health intake: A mixed-methods study. Journal of Counseling Psychology, 59, 185–196. http://dx.doi.org/10.1037/ a0027045 Ryder, A. G., Ban, L. M., & Chentsova-Dutton, Y. E. (2011). Towards a cultural– clinical psychology. Social and Personality Psychology Compass, 5, 960–975. http://dx.doi.org/10.1111/j.1751-9004.2011.00404.x Shweder, R. A. (1990). Cultural psychology: What is it? In J. W. Stigler, R. A. Shweder, & G. Herdt (Eds.), Cultural psychology: Essays on comparative human development (pp. 1–44). http://dx.doi.org/10.1017/CBO9781139173728.002 Shweder, R. A., & Sullivan, M. A. (1993). Cultural psychology: Who needs it? Annual Review of Psychology, 44, 497–523. http://dx.doi.org/10.1146/annurev. ps.44.020193.002433 Tashakkori, A., & Creswell, J. W. (2007). Editorial: Exploring the nature of research questions in mixed methods research. Journal of Mixed Methods Research, 1, 207–211. http://dx.doi.org/10.1177/1558689807302814 Taylor, C. (1973). Peaceful coexistence in psychology. Social Research, 40, 55–82. Tsai, T.-I., Morisky, D. E., Kagawa-Singer, M., & Ashing-Giwa, K. T. (2011). Acculturation in the adaptation of Chinese-American women to breast cancer: A mixed-method approach. Journal of Clinical Nursing, 20, 3383–3393. http:// dx.doi.org/10.1111/j.1365-2702.2011.03872.x Woolgar, S. (1988). Science: The very idea. London, England: Tavistock. Woolley, C. M. (2009). Meeting the mixed methods challenge of integration in a sociological study of structure and agency. Journal of Mixed Methods Research, 3, 7–25. http://dx.doi.org/10.1177/1558689808325774 Zukoski, A. P., Harvey, S. M., Oakley, L., & Branch, M. (2011). Exploring power and sexual decision making among young Latinos residing in rural communities. Women’s Health Issues, 21, 450–457. http://dx.doi.org/10.1016/j.whi.2011.05.002

mixed-methods research     

167

8 CULTURAL ADAPTATIONS IN PSYCHOTHERAPY FOR ETHNIC MINORITIES: STRATEGIES FOR RESEARCH ON CULTURALLY INFORMED EVIDENCE-BASED PSYCHOLOGICAL PRACTICES NOLAN ZANE, JIN E. KIM, GUILLERMO BERNAL, AND CATRINA GOTUACO

The persistent and intractable nature of mental health disparities that exist among ethnic minority populations has pointed to the need for culturally informed psychological treatments. This need is especially relevant for sociobehavioral interventions that are considered to be the best practices in mental health, namely, empirically supported treatments (ESTs). ESTs are still considered the gold standard for mental health treatment even though some have questioned whether the EST process of validation and generalization is really an effective way to reduce mental health disparities for ethnic minority populations (e.g., Bernal & Scharró-del-Río, 2001; Hall, 2001). Over the past 10 years, researchers and practitioners have worked together to develop and test various types of cultural adaptations to ESTs. These adaptations have taken many forms and have addressed various types of processes and aspects of psychological treatment. To guide this important

http://dx.doi.org/10.1037/14940-009 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

169

work, a number of promising approaches to designing and formulating adaptations have been developed and applied (e.g., Hwang, 2006; Lau, 2006). These advances notwithstanding, the measured progress made in developing culturally informed evidence-based treatments (EBTs) for ethnic minority clientele suggests that a reappraisal of these approaches as well as a refinement of the strategies used to guide this work is needed. The purpose of this chapter is threefold. First, we review the extant work done to culturally adapt ESTs for culturally diverse populations. Second, we conduct a content analysis of these adaptations to determine what domains have been the foci of these efforts and what other domains may need to be addressed to advance culturally competent or culturally informed treatment. Finally, we propose an alternative approach or strategy that may be helpful in advancing this work. APPROACHES TO CULTURALLY ADAPT EMPIRICALLY SUPPORTED TREATMENTS Significant progress has been made in developing culturally adapted interventions to better serve the mental health needs of ethnic minority patients and clients. Adaptations have varied in form, content, and the manner in which they are implemented. Cultural adaptation may be defined as making systematic changes with regard to culture, language, and context to EST or intervention manuals with the goal of attuning the protocol to the client’s worldview, cultural perspective, and values (Bernal, Jiménez-Chafey, & Domenech Rodríguez, 2009). Researchers have proposed a number of approaches to facilitate the systematic development of adaptations to empirically supported interventions. Lau (2006) proposed that adaptations must proceed only if certain conditions have been met. First, adaptations should be considered if there might be culturally distinct factors that may alter the risk or protective factors or the manifested symptoms of the disorder for a particular ethnic minority community. Second, adaptations may be needed when there is empirical support that the orthodox intervention may not be acceptable or socially appropriate for the cultural group at risk, such that clients may not access treatment and/or respond to the intervention. If either condition applies, the proposed adaptation must then be grounded in empirical findings that suggest it either can effectively address ethnic-specific risk and protective factors or culturally specific symptomatology or can ameliorate problems in treatment engagement or response. Interventions that reflect the principles articulated in Lau’s model either have garnered empirical support for their effectiveness with Native Americans (Whitbeck, Chen, Hoyt, & Adams, 2004) or appear to have generated a creative set of socially valid adaptations 170       zane et al.

for use with Mexican American (McCabe, Yeh, Garland, Lau, & Chavez, 2005; Weisman, 2005) and African American families (Coard, Wallace, Stevenson, & Brotman, 2004). Hwang (2006) suggested that adapted interventions can focus on certain promising domains or areas that are relevant to cultural issues. In his psychotherapy adaptation and modification framework (PAMF), he indicated that adaptations may occur within six specific domains and offered 25 principles that they can address to provide more culturally informed treatment to ethnic minority clients. These domains include (a) addressing dynamic issues and cultural complexities that are often salient to the life experiences of ethnic minority individuals; (b) orienting clients to the processes and goals of mental health care; (c) understanding cultural beliefs about mental illness, mental health, and helpseeking; (d) strengthening the working alliance or relationship between clients and therapists; (e) accommodating cultural variations in how clients manifest, express, and manage emotional distress; and (f) addressing cultural issues and life stressors that may be especially salient for ethnic minority clientele. In their efforts to advance culturally adapted treatments, Bernal, Bonilla, and Bellido (1995) proposed an ecological validity model (EVM) that identified the following aspects of the treatment that could be modified: (a) language—using culturally appropriate and culturally syntonic language; (b) persons—cultural variations in shaping the therapy relationship; (c) metaphors—symbols and concepts shared with the population; (d) content— knowledge of the client’s culture; (e) concepts—treatment concepts consonant with culture and context; (f) goal—supporting adaptive values from the culture of origin; (g) methods—development and/or cultural adaptation of treatment methods; and (h) context—consideration of changing contexts in assessment during treatment or intervention. The EVM approach is one of the most widely used approaches to cultural adaptations and it has received empirical support from a variety of studies including a recent meta-analysis (Smith, Rodríguez, & Bernal, 2011). It appears that these models have facilitated the development of cultural adaptations that have the potential to better address the specific needs of ethnic minority clientele. In the meta-analytic review of cultural adaptation research, Smith, Rodríguez, and Bernal (2011) found that culturally adapted treatments tended to outperform nonadapted treatments, achieving moderate differences (mean effect size = 0.46, k = 76, n = 25,225). Also, the outcomes tended to improve as the number of adaptations increased, and culturally adapting for client goals and metaphors that matched client worldviews was significantly associated with positive outcomes. In another metaanalytic study, Benish, Quintana, and Wampold (2011) found that adapted interventions outperformed original ESTs, although the differences between the treatments tended to be small (mean effect size = 0.32, k = 21, n = 472). In a cultural adaptations in psychotherapy     

171

meta-analysis with children and adolescents, Huey and Polo (2008) found moderate to low effects for adaptations (mean effect size = 0.38, k = 14, n = 1,056). In a more recent meta-analytic review of the treatment of depression with adults (van Loon, van Scheik, Dekker, & Beekman, 2013), culturally adapted treatments significantly outperformed nonadapted ones (mean effect size = 1.06, k = 9, n = 918). Research on cultural adaptations for ESTs has clearly generated an impressive array of guidelines and approaches. However, this creative set of strategies also presents a major challenge for adaptation researchers in terms of determining which approach should be applied for which EST, clinical problem, and clientele. With the numerous strategies and approaches available to guide the development of an adaptation or set of adaptations to an EST, it is not surprising that the adaptations themselves have been quite heterogeneous. This heterogeneity raises the important question as to the exact nature of these adaptations. In other words, it would be informative to know which ESTs actually have been adapted, for which ethnic minority clientele, and involving which aspects of the EST. Specifically, we determine which empirically supported interventions have been adapted, for whom the adaptation has been made, and the treatment component or process that the adaptation has addressed. Presumably, the adaptation will work or make the treatment more effective for certain ethnic minority clientele because it has changed a critical component or process of treatment. In this way, we can gauge the progress made in developing adaptations and the systematic trends that seem promising. Moreover, we can determine the extent to which the adaptations have focused on or involve processes empirically linked to treatment outcomes. On the basis of this analysis, we also can better determine what refinements or alternative strategies may be needed to advance research in this area. CONTENT ANALYSIS STUDY OF CULTURAL ADAPTATIONS Great progress has been made in developing culturally adapted treatments to better serve the mental health needs and preferences of ethnic minority and culturally diverse clients. However, great variations exist in the type, form, and focus of these adaptations. To determine the exact nature of these adaptations, we conducted a content analysis study of interventions that involved adaptations of ESTs to make them more culturally syntonic to the clinical and mental health needs of a particular ethnic minority clientele. The content analysis examined only adaptations made to ESTs and categorized each adaptation in terms of the treatment process or processes that the adaptation addressed. It should be noted that the analysis focused on all of the various efforts made to adapt ESTs (see Table 8.1) and not on whether or not 172       zane et al.

TABLE 8.1 Treatment Approaches Included in Content Analysis by Racial/Ethnic and Age Group

Target group

cultural adaptations in psychotherapy     

African American adults African American children, youth, family

Asian American adults

Study

Adapted treatment

Population of interest

Kohn et al. (2002)

Group CBT

Ginsburg & Drake (2002)

Group CBT

Myers et al. (1992)

PT

Inner-city African American families

Nicolas, Arntz, Hirsch, & Schmiedigen (2009) Coard et al. (2007)

Group CBT

Haitian American adolescents

PT

Ashworth et al. (2011)

CBT

Low-income African American parents and caregivers Korean immigrant

Chu et al. (2012)

Problem-solving therapy CBT

Dai et al. (1999)

Low-income African American women African American adolescents

Older Chinese American adults Elderly Chinese Americans

Mental or behavioral health issue Depression Anxiety disorders Child behavior problems and parenting skills Depression Child behavior problems and parenting skills Social anxiety disorder Depression Minor depressive symptoms and depression prevention

Type of study Treatment description and trial RCT Quasi-experimental study Treatment adaptation process; focus group findings RCT Case study Treatment adaptation process; pilot testing RCT

(continues)

173

174       zane et al.

TABLE 8.1 Treatment Approaches Included in Content Analysis by Racial/Ethnic and Age Group   (Continued )

Target group

Study Hinton et al. (2005)

CBT

Cambodian refugees

Hinton et al. (2013)

CBT with acceptance and mindfulness One-session exposure treatment Group PT

Southeast Asian and Latino refugees Asian Americans

Huey & Pan (2006); Pan et al. (2011) Asian American children, youth, family Latino/a American adults

Adapted treatment

Population of interest

Lau et al. (2010); Lau et al. (2011) Chavez-Korell et al. (2012) Dwight-Johnson et al. (2011)

IMPACT collaborative care model Telephone-based CBT

Hinton et al. (2011)

CBT

Interian, Allen, Gara, & Escobar (2008)

CBT

Immigrant Chinese families Latino elders Rural Latino primary care patients Latina women Hispanics in primary care

Mental or behavioral health issue

Type of study

Treatment-resistant PTSD and panic attacks PTSD

RCT

Specific phobias

Pilot study and 6-month follow-up

Child behavior problems and parenting skills Depression

Treatment description with case examples; RCT Treatment description with case example

Depression

RCT

Treatment-resistant PTSD Depression

RCT

Treatment description with case examples

Pre-/posttreatment trial

cultural adaptations in psychotherapy     

Latino/a American children, youth, family

Interian et al. (2010)

MI

Latinos on an antidepressant regimen Latinos

Antidepressant adherence

Kopelowicz (1998)

Skills training

Le, Zmuda, Perry, & Muñoz (2010) Lee et al. (2011)

CBT MI

Low-income Latina immigrants Hispanics

Perinatal depression Alcohol use/abuse

Patterson et al. (2005); Mausbach et al. (2008) Santiago-Rivera et al. (2008); Kanter et al. (2010) Shea et al. (2012)

Skills training

Older Latinos

Psychotic disorders

Behavioral activation

Latinos

Depression

CBT

Mexican American women

Binge-eating disorders

Burrow-Sanchez & Wrona (2012) D’Angelo et al. (2009)

Group CBT

Latino adolescents

Preventive intervention

Low-income Latino families

Substance use disorders Depression

CBT

Latino adolescent

Depression

Group PT

Latino parents of preschoolers

Child behavior problems and parenting skills

Duarté-Vélez et al. (2010) Dumas, Arriaga, Begle, & Longoria (2011)

Schizophrenia

Treatment adaptation process with case example Treatment description; RCT Treatment adaptation process Treatment description; preliminary study Treatment adaptation process with focus groups; RCT Treatment description and pilot evaluation; case example Treatment adaptation process with focus groups RCT Treatment adaptation process and open trial Case study Treatment description and pilot study (continues)

175

176       zane et al.

TABLE 8.1 Treatment Approaches Included in Content Analysis by Racial/Ethnic and Age Group   (Continued )

Target group

Study

Adapted treatment

Garza & Bratton (2005)

Child-centered play therapy

Kopelowicz et al. (2003)

Skills training and disease management PT

Martinez & Eddy (2005) Matos et al. (2009)

PCIT

Matos et al. (2006)

PCIT

McCabe et al. (2005)

PCIT

Parra Cardona et al. (2012)

PT

Population of interest

Mental or behavioral health issue

Type of study

Hispanic children referred for school counseling Latino patients and relatives

Child behavior problems

Treatment trial

Schizophrenia

RCT

Spanish-speaking Latino parents of youths Puerto Rican preschool children and parents

Risk for problem behaviors

RCT

Attention-deficit/ hyperactivity disorder and behavior problems Hyperactivity and behavior problems Child behavior problems Child behavior problems

RCT

Puerto Rican children aged 4–6 and parents Mexican American families Latino immigrant families

Treatment adaptation process Treatment adaptation process Treatment adaptation process

cultural adaptations in psychotherapy     

Native American adults Native American children, youth, families

Rosselló & Bernal (1996); Rosselló & Bernal (1999) Wood et al. (2008)

CBT and interpersonal treatment CBT

Venner et al. (2007)

MI

BigFoot & Schmidt (2010)

CBT

BigFoot & Funderburk (2011)

PCIT

Morsette, van den Pol, Schuldberg, Swaney, & Stolle (2012) Dionne, Davis, Sheeber, & Madrigal (2009)

CBT

MI and PT

Puerto Rican adolescents

Depression

RCT

Mexican American students in urban schools Native Americans

Anxiety disorders

Treatment description

Substance problems Trauma

Treatment adaptation process Treatment description with case example

Child behavior problems and parenting skills Trauma

Treatment description

American Indian and Alaska Native children American Indian and Alaska Native families American Indian youth American Indian children and families

Child behavior problems

Preliminary study using quasiexperimental design Open trial with MI; RCT with parent training

Note. CBT = cognitive–behavioral therapy; IMPACT = Improving Mood–Promoting Access to Collaborative Treatment; MI = motivational interviewing; PCIT = parent–child interaction therapy; PT = parent training; PTSD = posttraumatic stress disorder; RCT = randomized controlled trial.

177

these adaptations resulted in more effective interventions. The analysis differed from previous adaptation studies or reviews that determined if cultural adaptations were effective and/or incrementally effective relative to the conventional treatments (e.g., Benish et al., 2011; Coard, Foy-Watson, Zimmer, & Wallace, 2007). We were more interested in the form and content of the adaptation to better assess the progress made in developing more culturally syntonic interventions. Essentially, the analysis also identified what the adaptation changed in terms of specific core components or treatment processes. The analysis included only adaptations made to empirically supported interventions and excluded adaptations made to treatments, which did not have any empirical basis for their effectiveness. Selection of Relevant Studies for the Content Analysis We identified studies for the content analysis by initially conducting a literature search via PsycINFO for all publications related to cultural adaptations of EBTs published from 1990 to 2013 (July). In addition, we examined reference lists of representative review articles and meta-analyses (e.g., Huey & Polo, 2008) for additional studies. We identified any study that examined a cultural adaptation applied to an EBT or EST. We used the following inclusion criteria: First, the study had to have been conducted in the United States. Second, the study had to address a specific change in the treatment procedure or in the treatment context or personnel that implemented the EBT. Third, the intervention addressed a mental health issue (versus a physical health issue). Fourth, the adaptation had to be applied to an empirically supported or empirically validated treatment as defined by Division 12 (Society of Clinical Psychology) of the American Psychological Association. Finally, the adaptation had to involve a change to a sociobehavioral intervention that excluded all medical, pharmaceutical, and other biologically based interventions that treated psychological disorders (e.g., adaptations to how psychotropic medications are administered or modifications to the dosage level of these medications to address ethnic variations in metabolism rates). We excluded publications that focused on (a) an EBT given to an ethnic minority group without cultural modifications, (b) an adaptation of a service delivery setting rather than the treatment itself, (c) an adaptation to enhance medical treatment adherence (e.g., cultural adaptation to diabetes management), (d) an adaptation of instruments or measures, or (e) theoretical frameworks and general recommendations related to cultural adaptations. The literature search yielded 46 publications reflecting 41 treatment approaches in which cultural adaptations occurred. In five instances, two publications were related to the same type of cultural adaptation to the treatment approach (e.g., cultural adaptation of group parent training in Chinese 178       zane et al.

American families as discussed in Lau, Fung, & Yung, 2010, and also Lau, Fung, Ho, Liu, & Gudiño, 2011). For these cases, we grouped related publications together as a single culturally adapted treatment approach. Overall, the publications were diverse in their type and reflected the various stages of cultural adaptations to treatment. Studies ranged from delineating the development process of culturally adapting an EBT to describing case studies or case examples to reporting on results from randomized controlled trials. Categorizing Culturally Adapted Therapeutic Processes Before we undertook the content analysis study, it was necessary to generate the coding scheme for categorizing the treatment components or therapeutic processes addressed by the cultural adaptations. The adaptations could address changes in either surface structure or deep structure (Resnicow, Baranowski, Ahluwalia, & Braithwaite, 1999). The former refers to changing the structural aspects or the contextual aspects of the treatment, such as the type of personnel who provided the treatment or conducting therapy and using materials in the client’s native language. The latter refers to changing a treatment’s purported core components that actually stimulate or produce substantial behavioral, attitudinal, cognitive, and emotional changes in the client. We drew mainly from Grencavage and Norcross’s (1990) review of common therapeutic factors and examined change processes related to therapist qualities (e.g., empathic understanding) and client characteristics (e.g., positive expectation) as well as structural features of the treatment (e.g., a healing setting). We omitted change processes that would not be directly relevant to the study of cultural adaptations (e.g., placebo mechanisms). Although not a change process per se, we added a category pertaining to the adaptation of language and use of cultural metaphors, as this reflects an important area in the study of cultural adaptations (Bernal et al., 1995). Our final list contained 25 therapeutic change processes (see Table 8.2). In addition, we created working definitions of each of these 25 categories (available from Nolan Zane) to minimize variance in the interpretations that may arise in the coding process. Jin E. Kim and Catrina Gotuaco conducted the content analysis. Initially, five publications were randomly chosen for dual review. Kim and Gotuaco discussed how the categorization scheme of therapeutic processes in Table 8.2 would be applied to specific examples from the culturally adapted treatment approaches. Any initial disagreements were discussed and resolved with Nolan Zane. After this process, Kim and Gotuaco independently read and coded each publication to categorize areas of adaptations. The percentage of disagreements across the categories was 15%. These discrepancies were resolved through a consensus discussion with Nolan Zane. If it was unclear cultural adaptations in psychotherapy     

179

TABLE 8.2 Cultural Adaptations of 25 Therapeutic Processes by Racial/Ethnic Group Therapeutic processes or components addressed by cultural adaptations

African American (n = 5)

Asian American (n = 7)

Latino American (n = 24)

Native American (n = 5)

Total (n = 41)

4

4

13

5

26 (63%)

0

0

1

1

  2 (5%)

0

1

5

1

  7 (17%)

1

1

7

2

11 (27%)

1 0

0 2

1 15

3 1

  5 (12%) 18 (44%)

0

0

3

1

  4 (10%)

0

1

3

0

  4 (10%)

0

3

2

2

  7 (17%)

1

4

5

1

11 (27%)

2

1

10

1

14 (34%)

4

2

15

4

25 (61%)

0 2

0 7

0 16

0 4

  0 (0%) 29 (71%)

2

5

7

2

16 (39%)

0

2

2

2

  6 (15%)

0

5

11

1

17 (42%)

0

2

3

0

  5 (12%)

1

3

4

0

  8 (20%)

0 0 0

0 4 2

1 4 3

0 1 1

  1 (2%)   9 (22%)   6 (15%)

0

0

0

0

  0 (0%)

Acquisition and practice of new behaviors (D) Catharsis/ ventilation (D) Cultivating hope; enhancing expectancies (D) Contingency management (D) Cultural ritual use (D) Development of working alliance (D) Elicit selfdisclosure (D) Empathic understanding (D) Extinction; exposure; reconditioning (D) Fostering insight and awareness (D) Healing setting characteristics (S) Participants involved in treatment (S) Persuasion (D) Provision of information; education Provision of rationale/ explanation (D) Reality testing; normalization (D) Relabeling; destigmatizing (D) Socially sanctioned role of healer (D) Success and mastery experiences (D) Suggestion (D) Tension reduction (D) Therapist modeling (D) Transference; relationship projection (D)

180       zane et al.

TABLE 8.2 Cultural Adaptations of 25 Therapeutic Processes by Racial/Ethnic Group   (Continued ) Therapeutic processes or components addressed by cultural adaptations

African American (n = 5)

Asian American (n = 7)

Latino American (n = 24)

Native American (n = 5)

Total (n = 41)

4

4

23

4

35 (85%)

0

0

6

1

  7 (17%)

Use of appropriate language and metaphors (S) Warmth; positive regard; acceptance (D)

Note. D = deep structure change; S = surface structure change.

as to whether or not cultural adaptations had been made to a treatment, we deemed that the publication contained insufficient information and did not count such cases. Results and Discussion Table 8.1 summarizes the 41 culturally adapted treatment approaches that were included in the current content analysis, divided by target racial/ ethnic group and age group (i.e., adults or children, youth, family). As shown in Table 8.1, five of the 41 treatment approaches were culturally adapted for African Americans (12.2%), including one treatment approach for African American adults and four treatment approaches for children, youths, and/or families. There were seven culturally adapted treatments for Asian Americans (17.1%), with six treatment approaches for adults and one treatment approach for children and families. There were 24 culturally adapted treatment approaches for Latinos/as (58.5%), with 11 treatments for adults and 13 treatments for children, youths, and/or families. There were five treatment approaches for Native Americans (12.2%), including one treatment approach for Native American adults and four treatment approaches for children, youths, and/or families. Cognitive and behavioral therapies (including problem-solving therapy, behavioral activation, and motivational interviewing) accounted for the majority of the adaptations (56.1%). As indicated in Table 8.1, the large majority of the studies focused on adaptations to treatments for Hispanic or Latino/a clients. A sizable proportion of adaptations were designed for Asian American clients, and relatively few were focused on African American clients. The relatively large number of adaptations for Latino/a clients determined many of the following trends cultural adaptations in psychotherapy     

181

identified by the following analysis. A substantial number of the adaptations addressed changes to cognitive–behavioral treatments for depression or posttraumatic stress disorder (39%) or modifications to parent training interventions designed to change parenting practices and family dynamics (27%). Adaptation efforts did not address some of the most commonly used therapies used to treat some of the more prevalent mental disorders. For example, anxiety disorders constitute the most common mental disorder in the United States (Kessler et al., 2005). There also is evidence that cultural variations exist in risk for social anxiety (Okazaki, Liu, Longworth, & Minn, 2002; Zane & Song, 2007) and in the experience or interpretation of anxiety and bodily symptoms for Asian and Hispanic Americans (Weems, Hayward, Killen, & Taylor, 2002). Exposure therapies (e.g., systematic desensitization, flooding) along with cognitive–behavioral treatments are the most frequently used empirically supported treatments for anxiety disorders. Only four adaptations (10%) addressed treatments for anxiety problems, and three of the four involved changes to cognitive–behavioral therapy (CBT), with only one adaptation devoted to an exposure approach. Similarly, substance use disorders are the second most prevalent mental condition in the United States, but only two adaptations (5%) addressed this disorder—one involved motivational interviewing and the other addressed changes to a cognitive– behavioral approach. CBT was by far the most frequent focus of the adaptations, with close to half of the studies (46%) addressing this type of treatment. This result is not surprising because CBT is the most widely used EBT for many mental disorders. However, other widely used treatments for certain disorders such as behavioral therapy for the treatment of depression, motivational interviewing for the treatment of substance use disorders, and exposure therapies for the treatment of posttraumatic stress disorder and anxiety disorders received little consideration for adaptation efforts. In other words, some of the most frequently used therapies often have not been adapted for the four major ethnic minority client groups even when empirical evidence suggests such treatments may need modifications to adequately serve ethnic minority clientele. The relative lack of attention from adaptation researchers to these commonly used therapies may reflect the belief that these treatments are so well established that no adaptations are needed or necessary. However, as we argue later, there seem to be a number of compelling reasons why adaptation efforts should be directed toward these psychotherapies. Table 8.2 shows the aspect or component of the EBT that was adapted. Four types of adaptations were used in the majority of the interventions. These included (a) the use of appropriate language and metaphors (85%), (b) provision of education and information (71%), (c) acquisition or practice of new behaviors (63%), and (d) the participants involved in treatment 182       zane et al.

(61%). Note that three of these four focused on surface structural changes to a treatment (Resnicow et al., 1999). A substantial number of interventions applied adaptations in four other areas: (a) development of working alliance (44%), (b) relabeling or destigmatizing (42%), (c) provision of rationale for treatment or explanation of the problem (39%), and (d) healing setting characteristics (34%). In contrast to the four most frequent adaptations, most of these adaptations focused on deep structure changes. The pattern that emerged from examining the most frequent type of adaptations strongly suggests that changes to both surface and deep structure of the treatment appear needed if treatments are to be optimally effective with ethnic minority clientele. A closer inspection of the deep structure changes revealed that many of the treatment processes addressed involved processes that occurred earlier in treatment such as providing a culturally meaningful explanation of the mental disorder or building and strengthening the working or therapeutic alliance. The most frequent deep structure change involved developing skill sets or acquiring new behaviors. Close to two-thirds of the adapted treatments (63%) had this type of adaptation. Some of these approaches positioned these learning experiences in the context of coping with specific stressors or life circumstances often experienced by ethnic minority individuals. For example, Myers et al. (1992) adapted a parent training treatment for African American parents and families that focused on teaching their children how to deal with incidents they may encounter involving racism, discrimination, prejudice, and stereotyping. Similarly, an adaptation of a behavioral activation treatment of depression for Latina women focused “directly on the environmental problems faced by Latinos/as and how to change these life circumstances” (Santiago-Rivera et al., 2008, p. 179). However, in this domain, the most common adaptation approach involved tailoring the learning experiences to be congruent with certain salient cultural values for that particular ethnic group. A culturally adapted treatment for major depression among Hispanic patients modified assertion training by addressing the cultural value placed on showing respect for others (Interian, Martinez, Rios, Krejci, & Guarnaccia, 2010). The treatment taught patients how to respond assertively while at the same time emphasized the respect the person had for the other person. Chu, Hynuh, and Areán (2012) adapted problem-solving therapy for Chinese clients. They noted that in conventional problem solving, the therapist usually assumes a nondirective approach to encourage the client to generate his or her own solutions, but this process may not be credible to Chinese clients who may expect more direction from the therapist given the cultural value placed on social hierarchy and respect for authority figures. The cultural accommodation involved having the therapist first model problem solving by actively solving the first problem with the client. cultural adaptations in psychotherapy     

183

As the second most frequent adaptation to deep structure, close to half of the adapted treatments (44%) involved efforts to enhance the working alliance or to minimize cultural problems in this relationship. Most of these adaptations capitalized on salient cultural values or role expectations that presumably influenced the relationship between clients and therapists. For example, an adapted treatment for skills training treatment for Latino/a patients with psychotic disorders emphasized the importance of simpatia and personalismo to facilitate the working alliance (Patterson et al., 2005). Similarly, Kopelowicz, Zarate, Gonzalez Smith, Mintz, and Liberman (2003) addressed these values in a skills training treatment for Hispanic clients with schizophrenia through the sharing of food and by initiating casual conversation before and after the training sessions. Intervention researchers have been very creative and highly responsive to addressing cultural issues, values, life experiences and circumstances, and stressors. The diverse and rich set of adaptations along with concerted efforts to make changes that go beyond surface structure modifications underscores the substantial progress made to develop more culturally informed and culturally responsive EBTs that are nuanced for culturally diverse clientele. Nevertheless, the analysis revealed a number of challenges that remain and that may explain why the effectiveness of many adaptations has been modest. First, the majority of the studies still focused on surface structure changes rather than on deep structure changes. In our coding system, we purposely focused on the treatment processes that could be addressed by adaptations and reserved only four categories for surface structure modifications. Despite this, three of the four most frequent types of adaptation involved surface structure changes (i.e., use of appropriate language and metaphors, education and information provision, and participants involved). On average, 63% of the adaptations involved some type of surface structure modification. In contrast, when we examined the four most frequent deep structure changes (i.e., acquisition and practice of new behaviors, development of working alliance, relabeling or destigmatizing, and provision of rationale or explanation), on average, only 47% of the adaptations involved these changes. It seems that both surface and deep structure changes are needed in cultural adaptations. However, less emphasis on deep structure changes may have limited the impact of adaptations because the deep structure changes tend to affect treatment processes that are often empirically linked to treatment outcomes. Second, it appears that adaptations often have not addressed some basic treatment processes so essential to effective therapy. For example, a strong therapeutic or working alliance has been consistently linked to positive treatment outcomes (Martin, Garske, & Davis, 2000; Orlinsky, Rønnestad, & Willutzki, 2004). Often because of positive therapeutic bonds, client self-disclosure also is considered to be an essential condition for 184       zane et al.

effective treatment (Jourard, 1964; Stricker & Fisher, 1990). Moreover, Ridley (1984) noted how many African American clients may not selfdisclose to White therapists out of distrust and “cultural paranoia” because of the long history of discrimination and racism perpetuated against African Americans in the United States. The growing evidence that cultural factors such as these may hinder therapeutic conditions (e.g., Zane & Ku, 2014) points to the importance of addressing them in therapies for ethnic minority clientele. However, fewer than half of the adaptations (44%) addressed the working alliance and only a few studies (10%) focused on enhancing client self-disclosure. Finally, even when the adaptations addressed deep structure changes, it often is unclear whether the adaptations actually achieved their purported goals because of the lack of empirical evidence showing changes in the treatment process targeted by the adaptation. For example, Lee et al. (2011) conducted one of the few adaptations designed to facilitate client self-disclosure in their modification of motivational interviewing to treat Hispanics with alcohol problems. However, the study did not provide evidence that increases in self-disclosure had actually occurred. It appears that the areas of focus for the adaptations occurred because researchers assumed or nominated these domains on the basis of some empirical evidence that substantial ethnic or cultural variations existed in the variables. As the basis for documenting these cultural variations, investigators primarily used two methodologies. Focus groups were the most frequently used mechanism for soliciting and obtaining support for documenting or affirming that cultural variations involving this adaptation actually did exist. Other investigators used systematic literature reviews to determine if there were important cultural and ethnic differences in the variables they chose for adaptation. However, none of the investigations provided any type of evidence that demonstrated the adaptation actually had changed a critical treatment process related to outcomes. There appears to be a major disconnect between research on cultural adaptations to treatment and research on what makes psychotherapy effective. Moreover, this disconnect may contribute to the measured progress in reducing mental health disparities in the quality of care for ethnic minority clientele. In sum, researchers have made significant progress in developing culturally informed adaptations of empirically supported interventions so that these treatments can better address the specific mental health issues of ethnic minority clients. The majority of the adaptations were made for Latino/a clientele, underscoring the need for more adaptation research for the other three major ethnic minority groups. Many of the adaptations still focused on surface structure changes, but there is a growing trend toward more deep structure changes. There clearly is a need to develop more adaptations that cultural adaptations in psychotherapy     

185

target deep structure changes as these tend to influence key treatment processes that are empirically linked to clinical improvement. Finally, it appears that more of the deep structure changes could be directed toward important treatment conditions or processes that are necessary for treatment to be effective such as self-disclosure, the working alliance, and treatment credibility. It also would be important to know if the cultural adaptation actually enhances these critical processes and/or the purported change mechanisms involved in a particular treatment. CRITICAL PROCESS APPROACH TO CULTURALLY INFORMED TREATMENT ADAPTATION On the basis of the findings from our content analysis of adaptation research, we propose the following approach for designing and testing culturally informed treatment adaptations for ethnic minority clients, particularly with respect to deep structure changes. Adaptation researchers assume that culture maps onto specific social psychological factors that can influence treatment (Betancourt & Lopez, 1993), such as stigma, face concerns, and self-construals. These factors are related to the cultural experiences of ethnic minority clients, but they also can affect critical processes of treatment (e.g., treatment credibility, self-disclosure), which, in turn, affect treatment outcomes. However, adaptation researchers still have a major challenge in determining which cultural factor or factors should be addressed with respect to which key specific processes or components in treatment. A plethora of cultural factors can be addressed by an adaptation and these factors, in turn, can affect a large array of treatment processes. For example, cultural variations in communication style can affect treatment processes of client engagement and self-disclosure as well as behavioral change strategies. Variations in value orientation can influence most of these treatment processes such as treatment goals, client engagement, and self-disclosure. The critical process approach (CPA) can help researchers sort out and select the most promising variables that are addressed by a particular adaptation. CPA underscores the need to systematically use an empirically based approach to design adaptations for evidence-based practices so that these adaptations can be developed in a more systematic way that is directly tied to processes that are empirically associated with treatment outcomes. When this approach is used, three conditions must be met. First, there must be empirical evidence that substantial variation on the variable being adapted exists between different cultural or ethnic groups. In other words, the effect size associated with group differences on this variable should be large enough to justify its inclusion in a clinical intervention. Granted, 20 years ago, 186       zane et al.

research was insufficient to make this determination but now it is possible. Second, the variable adapted must also be empirically related to some type of critical process in treatment or directly to a treatment outcome. In this way, adaptations that are selected and used or applied have a greater likelihood of actually contributing to the effectiveness of the treatment. Third, there must be a rationale or framework that explains how the adaptation is culturally significant. Cultural significance as applied to an intervention refers to the extent to which an intervention or, in this case, an adaptation has features that resonate with, match with, or in some way fit with certain cultural tendencies of clients from a specific culture. In the studies reviewed for the content analysis, the rationales presented to justify the adaptation selected largely relied on claims of face validity. In other words, researchers often argued that the cultural factors addressed by the adaptation attended to issues important to the client with respect to mental health care. Too often it was unclear how or why the adaptation was culturally appropriate or culturally informed. Investigators often have not explained why an adaptation culturally fits or matches the cultural lifestyles, norms, values, and so on of a certain cultural group. Gallimore, Goldenberg, and Weisner (1993) proposed a conceptual framework for determining and assessing cultural fit or match. They noted that cultural lifestyles or cultural contexts constitute a set of regular activities that characterize how individuals meet and cope with the demands of daily life. When the activity of the human service (in this case, learning new skills or helping people with their emotional problems) is isomorphic, that is, similar in form and function to a daily activity setting of that person’s culture, the service will tend to be more culturally syntonic or culturally appropriate. This framework is only one way of conceptualizing cultural fit or match. More such frameworks are needed to provide a better theoretical basis for proposing a cultural adaptation. Last, and related to the previous condition, there must be some evidence that the adaptation is considered or perceived as culturally appropriate or socially valid by individuals from that cultural or ethnic group. For an intervention to be socially valid for clients from a particular cultural group, the clinical problem must be perceived as important, the procedures used to change behavior and/or attitudes must be considered culturally acceptable or appropriate, and the treatment goals or targeted changes must be valued or considered functional and important (Kazdin, 1977). Previous adaptation efforts have relied largely on focus groups and careful reviews of the relevant research to identify the cultural factor(s) addressed by the intervention. However, none of the adaptations reviewed here provided evidence to empirically demonstrate the intervention’s social validity. We now offer some preliminary work to show how the critical process approach can be applied to begin the development of effective cultural cultural adaptations in psychotherapy     

187

adaptations. Much of our work has focused on face concern as a pivotal variable in the treatment of culturally diverse clients from more collectivistic or interdependent cultures. On the basis of various accounts of face in both East Asian and Western psychology, it appears that face has the following psychosocial parameters: First, as social beings, people are invested in presenting to others, either implicitly or explicitly, certain claims about their character in terms of traits, attitudes, and values. This set of claims constitutes that person’s face (Ho, 1991). Others come to recognize and accept the person’s face, or line, that the person claims for him- or herself. Second, according to Hu (1944), face represents the person’s social position or prestige gained by performing one or more specific social roles that are well recognized by others or, as Goffman (1955) noted, “face is an image of self delineated in terms of approved social attributes” (p. 213). The line or face that one can claim is constrained or parameterized by the social roles ascribed and assumed by that person. Third, as Ho (1991) observed, face is very salient in East Asian social relations because of its function as a mechanism to maintain group harmony. Great value is placed in East Asian cultures on maintaining harmonious relationships among in-group members and protecting the integrity of the group. Face-saving behaviors and the avoidance of face-loss interactions enhance smooth interactions among group members and help minimize disruptions to the social order. Thus, face can be defined as essentially a person’s set of socially sanctioned claims concerning one’s social character and social integrity in which this set of claims or this line is defined largely by certain prescribed roles that one carries out as a member and representative of a group. The fact that face has esteem implications that extend beyond the individual to that individual’s group references is probably the main reason it has such psychological power in certain shame-based social groups such as East Asian and Asian American cultures. Research on face concern strongly suggests that cultural adaptations focused on this issue may be quite efficacious for collectivistic clients such as Asian American patients and clientele. First, there is consistent and convergent evidence that cultural differences exist in face concern among Asian American and White American individuals. Zane and Yeh (2002) found that Asian Americans were more concerned with face issues than were White American individuals, and the ethnic difference was fairly substantial at the moderate effect size level. The ethnic group effect size for the face concern tended to be greater than effect sizes found for other personality measures in earlier studies (e.g., Abe & Zane, 1990). Moreover, ethnic differences on face concern persisted after controlling for various demographic, personality, and distress differences between Asian Americans and Whites. These robust, ethnic differences have been found in studies focused on other Asian groups (e.g., Mak, Chen, Lam, & Yiu, 2009). As such, the first condition of 188       zane et al.

the critical process approach appears to have been met in light of the robust and substantial ethnic differences in face concern between Asian Americans and White Americans. There also is evidence that face concern influences important therapeutic processes. Self-disclosure is a critical aspect of counseling in psychotherapy; clients must disclose or share important, private, and confidential concerns with their therapists for therapy to be effective. Little therapeutic progress can occur if the client does not feel safe and trust the therapist sufficiently to divulge private and personal information that may cause great guilt, embarrassment, or shame. Zane and Ku (2014) experimentally varied ethnic match and gender match to determine their effects on self-disclosure among Asian American participants. They also assessed acculturation, cultural identity, and face concern prior to the experiment to determine if any of these organismic variables predicted self-disclosure. Four types of self-disclosure were assessed including disclosure about one’s personality, intimate relationships, personal attitudes and feelings, and sex life. Gender match affected disclosure about sex life whereas ethnic match did not affect any type of self-disclosure. Face concern was the most consistent predictor of self-disclosure in that it was a negative predictor of all four types of self-disclosure. Acculturation was a negative predictor of disclosure concerning close relationships, but this effect was mediated by face concern. Less acculturated individuals were more concerned about face, and those concerned about face tended to disclose less about their close relationships. In another study, face concern influenced another treatment process— the credibility of the therapeutic approach. Here, credibility refers to the client’s perceptions of the therapist as an effective, trustworthy, and expert helper (Kazdin & Wilcoxon, 1976; Sue & Zane, 1987; Tracey, Glidden, & Kokotovic, 1988). Previous studies have found that Asian American clients tend to prefer a more directive counseling style by the therapist to a nondirective one (e.g., Beutler et al., 2004; B. S. K. Kim, Li, and Liang, 2002). Park, Kim, and Zane (2015) hypothesized that face concern would moderate the relative preference for a directive versus nondirective counseling style. They also tested an alternative hypothesis that counseling expectations would account for this preference. Consistent with previous studies, Asian American participants preferred the directive style over the nondirective one. Face concern moderated this effect whereas counseling expectations did not. For Asian Americans with low face concern, there was no preference for the directive approach relative to the nondirective one. For Asian Americans with high face concern, there was a significant preference for the directive approach over the nondirective one. It appears that face concern can affect client response to different types of therapeutic approaches used in counseling and psychotherapy. cultural adaptations in psychotherapy     

189

Another study pointed to how cultural variations in symptomatology may be a promising focus for cultural adaptations. Somatic symptoms can involve cultural idioms of distress or culturally sanctioned help-seeking behaviors, especially among Asians and Asian Americans (Hwang, Myers, Abe-Kim, & Ting, 2008). This somatic tendency has often been attributed to the stigma of mental illness (Yen, Robins, & Lin, 2000) and to Asian cultural traditions that emphasize a more holistic view of the body and mind, leading to a greater focus on physical rather than emotional symptoms (Lin & Cheung, 1999). J. E. Kim, Zane, and Blozis (2012) determined whether or not somatic and other client tendencies were predictive of clinical outcomes. They found that client somatic tendency was a negative predictor of psychosocial functioning after short-term, outpatient treatment. Collectively, these studies suggest that both face concern and somatic tendencies may be opportunistic issues on which cultural adaptations can focus to develop more culturally informed interventions for ethnic minority clients, especially those from collectivistic or interdependent cultures such as Asian Americans. Plans are under way to pilot-test an adapted intervention of CBT for depression among Asian Americans. In light of the previous research and following the CPA, the adaptation uses face-saving strategies to enhance self-disclosure and client engagement as well as specific CBT procedures for pain management to address the somatic tendencies of Asian American clients. For the former, the adaptation focuses on reframing personal issues into solving specific problems as opposed to reducing personal symptoms or managing a personal disease (Littlejohn & Domenici, 2006). In this way, the orientation fosters demonstrations of mastery and competence that can enhance face restoration and minimize further face loss. For the latter, the adaptation capitalizes on empirically supported CBT procedures such as relaxation training and cognitive restructuring to reduce catastrophizing over somatic symptoms so that the intervention directly addresses the somatic tendencies of Asian American clients. The incremental yield to adaptation research from the critical process approach remains an empirical question. Specifically, there needs to be empirical support indicating that the interventions generated by this approach are actually more socially valid than the conventional interventions. Also needed is evidence that these more socially valid interventions can outperform the convention interventions when used with ethnic minority clients. Essentially, the critical process approach embeds evidence-based criteria throughout the cultural adaptation process from the development of the intervention through its clinical trial. Arguably, this approach should increase the likelihood that such adaptations will be culturally informed as well as have major clinical effects. 190       zane et al.

REFERENCES Abe, J. S., & Zane, N. W. S. (1990). Psychological maladjustment among Asian and White American college students: Controlling for confounds. Journal of Counseling Psychology, 37, 437–444. http://dx.doi.org/10.1037/0022-0167.37.4.437 Ashworth, K. J., Randall, J., Millen, A., & Rosqvist, J. (2011). Culturally competent CBT: Treating SAD in a Korean immigrant—A single case analysis. Clinical Case Studies, 10, 449–465. http://dx.doi.org/10.1177/1534650111435163 Benish, S. G., Quintana, S., & Wampold, B. E. (2011). Culturally adapted psychotherapy and the legitimacy of myth: A direct-comparison meta-analysis. Journal of Counseling Psychology, 58, 279–289. http://dx.doi.org/10.1037/a0023626 Bernal, G., Bonilla, J., & Bellido, C. (1995). Ecological validity and cultural sensitivity for outcome research: Issues for the cultural adaptation and development of psychosocial treatments with Hispanics. Journal of Abnormal Child Psychology, 23, 67–82. http://dx.doi.org/10.1007/BF01447045 Bernal, G., Jiménez-Chafey, M. I., & Domenech Rodríguez, M. M. (2009). Cultural adaptation of evidence-based treatments for ethno-cultural youth. Professional Psychology: Research and Practice, 40, 361–368. Bernal, G., & Scharró-del-Río, M. R. (2001). Are empirically supported treatments valid for ethnic minorities? Toward an alternative approach for treatment research. Cultural Diversity and Ethnic Minority Psychology, 7, 328–342. http:// dx.doi.org/10.1037/1099-9809.7.4.328 Betancourt, H., & Lopez, S. R. (1993). The study of culture, ethnicity, and race in American psychology. American Psychologist, 48, 629–637. http://dx.doi. org/10.1037/0003-066X.48.6.629 Beutler, L. E., Malik, M., Alimohamed, S., Harwood, T. M., Talebi, H., Noble, S., & Wong, E. (2004). Therapist effects. In M. J. Lambert (Ed.), Bergin and Garfield’s handbook of psychotherapy and behavior change (5th ed., pp. 227–306). New York, NY: Wiley. BigFoot, D. S., & Funderburk, B. W. (2011). Honoring children, making relatives: The cultural translation of parent-child interaction therapy for American Indian and Alaska Native families. Journal of Psychoactive Drugs, 43, 309–318. http://dx.doi.org/10.1080/02791072.2011.628924 BigFoot, D. S., & Schmidt, S. R. (2010). Honoring children, mending the circle: Cultural adaptation of trauma-focused cognitive-behavioral therapy for American Indian and Alaska native children. Journal of Clinical Psychology, 66, 847–856. http://dx.doi.org/10.1002/jclp.20707 Burrow-Sanchez, J. J., & Wrona, M. (2012). Comparing culturally accommodated versus standard group CBT for Latino adolescents with substance use disorders: A pilot study. Cultural Diversity and Ethnic Minority Psychology, 18, 373–383. http://dx.doi.org/10.1037/a0029439 Chavez-Korell, S., Rendón, A. D., Beer, J., Rodriguez, N., Garr, A. D., Pine, C. A., . . . Malcolm, E. (2012). Improving access and reducing barriers to depression cultural adaptations in psychotherapy     

191

treatment for Latino elders: Un nuevo amanecer (a new dawn). Professional Psychology: Research and Practice, 43, 217–226. http://dx.doi.org/10.1037/a0026695 Chu, J. P., Huynh, L., & Areán, P. (2012). Cultural adaptation of evidence-based practice utilizing an iterative stakeholder process and theoretical framework: Problem solving therapy for Chinese older adults. International Journal of Geriatric Psychiatry, 27, 97–106. http://dx.doi.org/10.1002/gps.2698 Coard, S. I., Foy-Watson, S., Zimmer, C., & Wallace, A. (2007). Considering culturally relevant parenting practices in intervention development and adaptation: A randomized controlled trial of the Black Parenting Strengths and Strategies (BPSS) program. The Counseling Psychologist, 35, 797–820. http://dx.doi. org/10.1177/0011000007304592 Coard, S. I., Wallace, S. A., Stevenson, H. C., Jr., & Brotman, L. M. (2004). Towards culturally relevant preventive interventions: The consideration of racial socialization in parent training with African American families. Journal of Child and Family Studies, 13, 277–293. http://dx.doi.org/10.1023/B:JCFS.0000022035.07171.f8 Dai, Y., Zhang, S., Yamamoto, J., Ao, M., Belin, T. R., Cheung, F., & Hifumi, S. S. (1999). Cognitive behavioral therapy of minor depressive symptoms in elderly Chinese Americans: A pilot study. Community Mental Health Journal, 35, 537–542. http://dx.doi.org/10.1023/A:1018763302198 D’Angelo, E. J., Llerena-Quinn, R., Shapiro, R., Colon, F., Rodriguez, P., Gallagher, K., & Beardslee, W. R. (2009). Adaptation of the preventive intervention program for depression for use with predominantly low-income Latino families. Family Process, 48, 269–291. http://dx.doi.org/10.1111/j.1545-5300.2009.01281.x Dionne, R., Davis, B., Sheeber, L., & Madrigal, L. (2009). Initial evaluation of a cultural approach to implementation of evidence-based parenting interventions in American Indian communities. Journal of Community Psychology, 37, 911–921. http://dx.doi.org/10.1002/jcop.20336 Duarté-Vélez, Y., Bernal, G., & Bonilla, K. (2010). Culturally adapted cognitivebehavior therapy: Integrating sexual, spiritual, and family identities in an evidence-based treatment of a depressed Latino adolescent. Journal of Clinical Psychology, 66, 895–906. http://dx.doi.org/10.1002/jclp.20710 Dumas, J. E., Arriaga, X. B., Begle, A. M., & Longoria, Z. N. (2011). Child and parental outcomes of a group parenting intervention for Latino families: A pilot study of the CANNE program. Cultural Diversity and Ethnic Minority Psychology, 17, 107–115. http://dx.doi.org/10.1037/a0021972 Dwight-Johnson, M., Aisenberg, E., Golinelli, D., Hong, S., O’Brien, M., & Ludman, E. (2011). Telephone-based cognitive-behavioral therapy for Latino patients living in rural areas: A randomized pilot study. Psychiatric Services, 62, 936–942. http://dx.doi.org/10.1176/ps.62.8.pss6208_0936 Gallimore, R., Goldenberg, C. N., & Weisner, T. S. (1993). The social construction and subjective reality of activity settings: Implications for community psychology. American Journal of Community Psychology, 21, 537–560. http://dx.doi. org/10.1007/BF00942159

192       zane et al.

Garza, Y., & Bratton, S. C. (2005). School-based child-centered play therapy with Hispanic children: Outcomes and cultural consideration. International Journal of Play Therapy, 14(1), 51–80. http://dx.doi.org/10.1037/h0088896 Ginsburg, G. S., & Drake, K. L. (2002). School-based treatment for anxious African American adolescents: A controlled pilot study. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 768–775. http://dx.doi.org/ 10.1097/00004583-200207000-00007 Goffman, E. (1955). On face-work: An analysis of ritual elements in social interaction. Psychiatry: Journal for the Study of Interpersonal Processes, 18, 213–231. Grencavage, L. M., & Norcross, J. C. (1990). Where are the commonalities among the therapeutic common factors? Professional Psychology: Research and Practice, 21, 372–378. http://dx.doi.org/10.1037/0735-7028.21.5.372 Hall, G. C. N. (2001). Psychotherapy research with ethnic minorities: Empirical, ethical, and conceptual issues. Journal of Consulting and Clinical Psychology, 69, 502–510. http://dx.doi.org/10.1037/0022-006X.69.3.502 Hinton, D. E., Chhean, D., Pich, V., Safren, S. A., Hofmann, S. G., & Pollack, M. H. (2005). A randomized controlled trial of cognitive-behavior therapy for Cambodian refugees with treatment-resistant PTSD and panic attacks: A crossover design. Journal of Traumatic Stress, 18, 617–629. http://dx.doi.org/10.1002/ jts.20070 Hinton, D. E., Hofmann, S. G., Rivera, E., Otto, M. W., & Pollack, M. H. (2011). Culturally adapted CBT (CA-CBT) for Latino women with treatmentresistant PTSD: A pilot study comparing CA-CBT to applied muscle relaxation. Behaviour Research and Therapy, 49, 275–280. http://dx.doi.org/10.1016/ j.brat.2011.01.005 Hinton, D. E., Pich, V., Hofmann, S. G., & Otto, M. W. (2013). Acceptance and mindfulness techniques as applied to refugee and ethnic minority populations with PTSD: Examples from “culturally adapted CBT.” Cognitive and Behavioral Practice, 20, 33–46. http://dx.doi.org/10.1016/j.cbpra.2011.09.001 Ho, D. (1991). The concept of “face” in Chinese American interaction. In W.-C. Hu & C. L. Grove (Eds.), Encountering the Chinese: A guide for Americans (pp. 111–124). Yarmouth, ME: Intercultural Press. Hu, H. C. (1944). The Chinese concepts of “face.” American Anthropologist, 46, 45–64. http://dx.doi.org/10.1525/aa.1944.46.1.02a00040 Huey, S. J., Jr., & Pan, D. (2006). Culture-responsive one-session treatment for phobic Asian Americans: A pilot study. Psychotherapy: Theory, Research, Practice, Training, 43, 549–554. http://dx.doi.org/10.1037/0033-3204.43.4.549 Huey, S. J., Jr., & Polo, A. J. (2008). Evidence-based psychosocial treatments for ethnic minority youth. Journal of Clinical Child and Adolescent Psychology, 37, 262–301. http://dx.doi.org/10.1080/15374410701820174 Hwang, W.-C. (2006). The psychotherapy adaptation and modification framework: Application to Asian Americans. American Psychologist, 61, 702–715. http:// dx.doi.org/10.1037/0003-066X.61.7.702 cultural adaptations in psychotherapy     

193

Hwang, W.-C., Myers, H. F., Abe-Kim, J., & Ting, J. Y. (2008). A conceptual paradigm for understanding culture’s impact on mental health: The cultural influences on mental health (CIMH) model. Clinical Psychology Review, 28, 211–227. http://dx.doi.org/10.1016/j.cpr.2007.05.001 Interian, A., Allen, L. A., Gara, M. A., & Escobar, J. I. (2008). A pilot study of culturally adapted cognitive behavior therapy for Hispanics with major depression. Cognitive and Behavioral Practice, 15, 67–75. http://dx.doi.org/10.1016/ j.cbpra.2006.12.002 Interian, A., Martinez, I., Rios, L. I., Krejci, J., & Guarnaccia, P. J. (2010). Adaptation of a motivational interviewing intervention to improve antidepressant adherence among Latinos. Cultural Diversity and Ethnic Minority Psychology, 16, 215–225. http://dx.doi.org/10.1037/a0016072 Jourard, S. M. (1964). Risking a leap. PsycCRITIQUES, 9, 394–396. http://dx.doi. org/10.1037/007383 Kanter, J. W., Santiago-Rivera, A. L., Rusch, L. C., Busch, A. M., & West, P. (2010). Initial outcomes of a culturally adapted behavioral activation for Latinas diagnosed with depression at a community clinic. Behavior Modification, 34, 120–144. http://dx.doi.org/10.1177/0145445509359682 Kazdin, A. E. (1977). Assessing the clinical or applied importance of behavior change through social validation. Behavior Modification, 1, 427–452. http://dx.doi.org/ 10.1177/014544557714001 Kazdin, A. E., & Wilcoxon, L. A. (1976). Systematic desensitization and nonspecific treatment effects: A methodological evaluation. Psychological Bulletin, 83, 729–758. http://dx.doi.org/10.1037/0033-2909.83.5.729 Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime prevalence and age-of-onset distributions of DSM–IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62, 593–602. http://dx.doi.org/10.1001/archpsyc.62.6.593 Kim, B. S. K., Li, L. C., & Liang, T. H. (2002). Effects of Asian American client adherence to Asian cultural values, session goal, and counselor emphasis of client expression on career counseling process. Journal of Counseling Psychology, 49, 342–354. http://dx.doi.org/10.1037/0022-0167.49.3.342 Kim, J. E., Zane, N. W., & Blozis, S. A. (2012). Client predictors of short-term psychotherapy outcomes among Asian and White American outpatients. Journal of Clinical Psychology, 68, 1287–1302. http://dx.doi.org/10.1002/jclp.21905 Kohn, L. P., Oden, T., Muñoz, R. F., Robinson, A., & Leavitt, D. (2002). Adapted cognitive behavioral group therapy for depressed low-income African American women. Community Mental Health Journal, 38, 497–504. http://dx.doi. org/10.1023/A:1020884202677 Kopelowicz, A. (1998). Adapting social skills training for Latinos with schizophrenia. International Review of Psychiatry, 10, 47–50. http://dx.doi.org/10.1080/ 09540269875087

194       zane et al.

Kopelowicz, A., Zarate, R., Gonzalez Smith, V., Mintz, J., & Liberman, R. P. (2003). Disease management in Latinos with schizophrenia: A family-assisted, skills training approach. Schizophrenia Bulletin, 29, 211–228. http://dx.doi.org/10.1093/ oxfordjournals.schbul.a006999 Lau, A. S. (2006). Making the case for selective and directed cultural adaptations of evidence-based treatments: Examples from parent training. Clinical Psychology: Science and Practice, 13, 295–310. http://dx.doi.org/10.1111/ j.1468-2850.2006.00042.x Lau, A. S., Fung, J. J., Ho, L. Y., Liu, L. L., & Gudiño, O. G. (2011). Parent training with high-risk immigrant Chinese families: A pilot group randomized trial yielding practice-based evidence. Behavior Therapy, 42, 413–426. http://dx.doi. org/10.1016/j.beth.2010.11.001 Lau, A. S., Fung, J. J., & Yung, V. (2010). Group parent training with immigrant Chinese families: Enhancing engagement and augmenting skills training. Journal of Clinical Psychology, 66, 880–894. http://dx.doi.org/10.1002/jclp.20711 Le, H. N., Zmuda, J., Perry, D. F., & Muñoz, R. F. (2010). Transforming an evidencebased intervention to prevent perinatal depression for low-income Latina immigrants. American Journal of Orthopsychiatry, 80, 34–45. http://dx.doi.org/ 10.1111/j.1939-0025.2010.01005.x Lee, C. S., López, S. R., Hernández, L., Colby, S. M., Caetano, R., Borrelli, B., & Rohsenow, D. (2011). A cultural adaptation of motivational interviewing to address heavy drinking among Hispanics. Cultural Diversity and Ethnic Minority Psychology, 17, 317–324. http://dx.doi.org/10.1037/a0024035 Lin, K. M., & Cheung, F. (1999). Mental health issues for Asian Americans. Psychiatric Services, 50, 774–780. http://dx.doi.org/10.1176/ps.50.6.774 Littlejohn, S., & Domenici, K. (2006). A Facework frame for mediation. In M. S. Herrman (Ed.), The Blackwell handbook of mediation: Bridging theory, research, and practice (pp. 228–246). Malden, MA: Blackwell. Mak, W. W. S., Chen, S. X., Lam, A. G., & Yiu, V. F. L. (2009). Understanding distress: The role of face concern among Chinese Americans, European Americans, Hong Kong Chinese, and Mainland Chinese. The Counseling Psychologist, 37, 219–248. http://dx.doi.org/10.1177/0011000008316378 Martin, D. J., Garske, J. P., & Davis, M. K. (2000). Relation of the therapeutic alliance with outcome and other variables: A meta-analytic review. Journal of Consulting and Clinical Psychology, 68, 438–450. http://dx.doi.org/10.1037/ 0022-006X.68.3.438 Martinez, C. R., Jr., & Eddy, J. M. (2005). Effects of culturally adapted parent management training on Latino youth behavioral health outcomes. Journal of Consulting and Clinical Psychology, 73, 841–851. Matos, M., Bauermeister, J. J., & Bernal, G. (2009). Parent-child interaction therapy for Puerto Rican preschool children with ADHD and behavior problems: A pilot efficacy study. Family Process, 48, 232–252. http://dx.doi.org/10.1111/ j.1545-5300.2009.01279.x cultural adaptations in psychotherapy     

195

Matos, M., Torres, R., Santiago, R., Jurado, M., & Rodríguez, I. (2006). Adaptation of parent-child interaction therapy for Puerto Rican families: A preliminary study. Family Process, 45, 205–222. http://dx.doi.org/10.1111/j.1545-5300.2006.00091.x Mausbach, B. T., Bucardo, J., McKibbin, C. L., Goldman, S. R., Jeste, D. V., Patterson, T.  L., . . . Barrio, C. (2008). Evaluation of a culturally tailored skills intervention for Latinos with persistent psychotic disorders. American Journal of Psychiatric Rehabilitation, 11, 61–75. http://dx.doi.org/10.1080/15487760701853102 McCabe, K. M., Yeh, M., Garland, A. F., Lau, A. S., & Chavez, G. (2005). The GANA program: A tailoring approach to adapting parent child interaction therapy for Mexican Americans. Education & Treatment of Children, 28, 111–129. Morsette, A., van den Pol, R., Schuldberg, D., Swaney, G., & Stolle, D. (2012). Cognitive behavioral treatment for trauma symptoms in American Indian youth: Preliminary findings and issues in evidence-based practice and reservation culture. Advances in School Mental Health Promotion, 5, 51–62. http://dx.doi.org/ 10.1080/1754730X.2012.664865 Myers, H. F., Alvy, K. T., Arlington, A., Richardson, M. A., Marigna, M., Huff, R., . . . Newcomb, M. D. (1992). The impact of a parent training program on inner-city African American families. Journal of Community Psychology, 20, 132–147. http://dx.doi.org/10.1002/1520-6629(199204)20:23.0.CO;2-Z Nicolas, G., Arntz, D. L., Hirsch, B., & Schmiedigen, A. (2009). Cultural adaptation of a group treatment for Haitian American adolescents. Professional Psychology: Research and Practice, 40, 378–384. http://dx.doi.org/10.1037/a0016307 Okazaki, S., Liu, J. F., Longworth, S. L., & Minn, J. Y. (2002). Asian American– White American differences in expressions of social anxiety: A replication and extension. Cultural Diversity and Ethnic Minority Psychology, 8, 234–247. http:// dx.doi.org/10.1037/1099-9809.8.3.234 Orlinsky, D. E., Rønnestad, M. H., & Willutzki, U. (2004). Fifty years of psychotherapy process-outcome research: Continuity and change. In M. J. Lambert (Ed.), Bergin and Garfield’s handbook of psychotherapy and behavior change (5th ed., pp. 307–389). New York, NY: Wiley. Pan, D., Huey, S. J., Jr., & Hernandez, D. (2011). Culturally adapted versus standard exposure treatment for phobic Asian Americans: Treatment efficacy, moderators, and predictors. Cultural Diversity and Ethnic Minority Psychology, 17, 11–22. http://dx.doi.org/10.1037/a0022534 Park, S., Kim, J., & Zane, N. (2015). Effects of ethnicity, counseling expectations, and face concern on counseling style credibility and the working alliance. Asian American Journal of Psychology. Manuscript submitted for publication. Parra Cardona, J. R., Domenech-Rodriguez, M., Forgatch, M., Sullivan, C., Bybee, D., Holtrop, K., . . . Bernal, G. (2012). Culturally adapting an evidence-based parenting intervention for Latino immigrants: The need to integrate fidelity and cultural relevance. Family Process, 51, 56–72. http://dx.doi.org/10.1111/ j.1545-5300.2012.01386.x

196       zane et al.

Patterson, T. L., Bucardo, J., McKibbin, C. L., Mausbach, B. T., Moore, D., Barrio, C., . . . Jeste, D. V. (2005). Development and pilot testing of a new psychosocial intervention for older Latinos with chronic psychosis. Schizophrenia Bulletin, 31, 922–930. http://dx.doi.org/10.1093/schbul/sbi036 Resnicow, K., Baranowski, T., Ahluwalia, J. S., & Braithwaite, R. L. (1999). Cultural sensitivity in public health: Defined and demystified. Ethnicity & Disease, 9(1), 10–21. Ridley, C. R. (1984). Clinical treatment of the nondisclosing Black client. A therapeutic paradox. American Psychologist, 39, 1234–1244. http://dx.doi.org/ 10.1037/0003-066X.39.11.1234 Rosselló, J., & Bernal, G. (1996). Adapting cognitive-behavioral and interpersonal treatments for depressed Puerto Rican adolescents. In E. D. Hibbs & P. S. Jensen (Eds.), Psychosocial treatments for child and adolescent disorders: Empirically based strategies for clinical practice (pp. 157–185). http://dx.doi.org/10.1037/10196-007 Rosselló, J., & Bernal, G. (1999). The efficacy of cognitive-behavioral and interpersonal treatments for depression in Puerto Rican adolescents. Journal of Consulting and Clinical Psychology, 67, 734–745. http://dx.doi.org/10.1037/ 0022-006X.67.5.734 Santiago-Rivera, A., Kanter, J., Benson, G., Derose, T., Illes, R., & Reyes, W. (2008). Behavioral activation as an alternative treatment approach for Latinos with depression. Psychotherapy: Theory, Research, Practice, Training, 45, 173–185. http://dx.doi.org/10.1037/0033-3204.45.2.173 Shea, M., Cachelin, F., Uribe, L., Striegel, R. H., Thompson, D., & Wilson, G. T. (2012). Cultural adaptation of a cognitive behavior therapy guided self-help program for Mexican American women with binge eating disorders. Journal of Counseling & Development, 90, 308–318. http://dx.doi.org/10.1002/ j.1556-6676.2012.00039.x Smith, T. B., Rodríguez, M. D., & Bernal, G. (2011). Culture. Journal of Clinical Psychology, 67, 166–175. http://dx.doi.org/10.1002/jclp.20757 Stricker, G., & Fisher, M. (Eds.). (1990). Self-disclosure in the therapeutic relationship. New York, NY: Plenum Press. Sue, S., & Zane, N. (1987). The role of culture and cultural techniques in psychotherapy. A critique and reformulation. American Psychologist, 42, 37–45. http:// dx.doi.org/10.1037/0003-066X.42.1.37 Tracey, T. J., Glidden, C. E., & Kokotovic, A. M. (1988). Factor structure of the Counselor Rating Form—Short. Journal of Counseling Psychology, 35, 330–335. http://dx.doi.org/10.1037/0022-0167.35.3.330 van Loon, A., van Schaik, A., Dekker, J., & Beekman, A. (2013). Bridging the gap for ethnic minority adult outpatients with depression and anxiety disorders by culturally adapted treatments. Journal of Affective Disorders, 147, 9–16. Venner, K. L., Feldstein, S. W., & Tafoya, N. (2007). Helping clients feel welcome: Principles of adapting treatment cross-culturally. Alcoholism Treatment Quarterly, 25(4), 11–30. http://dx.doi.org/10.1300/J020v25n04_02 cultural adaptations in psychotherapy     

197

Weems, C. F., Hayward, C., Killen, J., & Taylor, C. B. (2002). A longitudinal investigation of anxiety sensitivity in adolescence. Journal of Abnormal Psychology, 111, 471–477. http://dx.doi.org/10.1037/0021-843X.111.3.471 Weisman, A. (2005). Integrating culturally based approaches with existing interventions for Hispanic/Latino families coping with schizophrenia. Psychotherapy: Theory, Research, Practice, Training, 42, 178–197. http://dx.doi.org/10.1037/ 0033-3204.42.2.178 Whitbeck, B. L., Chen, X., Hoyt, D. R., & Adams, G. W. (2004). Discrimination, historical loss and enculturation: Culturally specific risk and resiliency factors for alcohol abuse among American Indians. Journal of Studies on Alcohol, 65, 409–418. http://dx.doi.org/10.15288/jsa.2004.65.409 Wood, J. J., Chiu, A. W., Hwang, W.-C., Jacobs, J., & Ifekwunigwe, M. (2008). Adapting cognitive–behavioral therapy for Mexican American students with anxiety disorders: Recommendations for school psychologists. School Psychology Quarterly, 23, 515–532. http://dx.doi.org/10.1037/1045-3830.23.4.515 Yen, S., Robins, C. J., & Lin, N. (2000). A cross-cultural comparison of depressive symptom manifestation: China and the United States. Journal of Consulting and Clinical Psychology, 68, 993–999. http://dx.doi.org/10.1037/0022-006X.68.6.993 Zane, N., & Ku, H. (2014). Effects of ethnic match, gender match, acculturation, cultural identity, and face concerns on self-disclosure in counseling for Asian Americans. Asian American Journal of Psychology, 5, 66–74. http://dx.doi. org/10.1037/a0036078 Zane, N., & Song, A. (2007). Interpersonal effectiveness among Asian Americans: Issues of leadership, career advancement, and social competence. In F. T. L. Leong, A. G. Iman, A. Ebreo, L. Yang, L. Kinoshita, & M. Fu (Eds.), Handbook of Asian American psychology (2nd ed., pp. 283–301). Thousand Oaks, CA: Sage. Zane, N., & Yeh, M. (2002). Use of culturally-based variables in assessment: Studies on loss of face. In K. S. Kurasaki, S. Okazaki, & S. Sue (Eds.), Asian American mental health: Assessment, theories, and methods (pp. 123–138). http://dx.doi.org/ 10.1007/978-1-4615-0735-2_9

198       zane et al.

9 COMMUNITY-BASED PARTICIPATORY RESEARCH FOR COCREATING INTERVENTIONS WITH NATIVE COMMUNITIES: A PARTNERSHIP BETWEEN THE UNIVERSITY OF NEW MEXICO AND THE PUEBLO OF JEMEZ LORENDA BELONE, JANICE TOSA, KEVIN SHENDO, ANITA TOYA, KEE STRAITS, GREG TAFOYA, REBECCA RAE, EMMA NOYES, DOREEN BIRD, AND NINA WALLERSTEIN

Community-based participatory research (CBPR) is recognized as an important research approach for reducing disparities and improving health status within communities of color (communities of color refers to communities of people who are not White) and other communities of social identity that have faced histories and patterns of discrimination or stigmatization (Minkler & Wallerstein, 2008; National Congress of American Indians [NCAI] Policy Research Center & Montana State University Center for Native Health Partnerships, 2012). CBPR has been defined “not simply as a community outreach strategy but rather a systematic effort to incorporate community This research was supported by funding from the Native American Research Centers for Health (a joint program by the National Institutes of Health and the Indian Health Service), No. U26IHS30009, through the Albuquerque Area Indian Health Board. Principal Investigator: Nina Wallerstein. We deeply appreciate our partners and thank them for their involvement throughout the process: Pueblo of Jemez tribal leadership; the Hemish of Walatowa Family Circle Program Advisory Council and Facilitators; the Pueblo of Jemez Departments of Education and Health and Human Services; and Walatowa Charter High School. http://dx.doi.org/10.1037/14940-010 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

199

participation and decision making, local theories of etiology and change, and community practices into the research effort” (Wallerstein & Duran, 2006, p. 313). Key elements of CBPR include (a) community ownership, (b) coalition building with internal and external partners, (c) capacity building, (d) promotion of interdependence that facilitates colearning, (e) application of research findings to action, and (f) long-term commitment to communities (Israel, Eng, Schulz, & Parker, 2013). It holds promise to enhance translational science because of its capacity to promote external validity and reach of interventions to diverse communities; recognizes the importance of implementation context; and increases ownership and sustainability through grounding interventions within community cultures (Belone et al., 2014; Wallerstein & Duran, 2010). American Indian/Alaska Native (AI/AN) communities, in particular, have increasingly expected and demanded the use of CBPR strategies because of historic genocide and federal institutional policies, such as assimilation through boarding schools (Duran & Duran, 2000; Duran, Duran, & Yellow Horse Braveheart, 1998). Research abuses, even recently (Mello & Wolf, 2010), have led tribes to reject being “surveyed to death,” without the return of data or receiving benefits of the research. The term tribal participatory research has grown in use (Baldwin, Johnson, & Benally, 2009; Burhansstipanov, Christopher, & Schumacher, 2005; Fisher & Ball, 2003; LaVeaux & Christopher, 2009; Teufel-Shone, Siyuja, Watahomigie, & Irwin, 2006; Thomas et al., 2009), acknowledging the importance of tribal governance in oversight of research and growing numbers of tribal institutional review boards (Becenti-Pigman, White, Bowman, Palmanteer-Holder, & Duran, 2008). AI/AN communities are particularly at risk for health disparities, facing high rates of historical and intergenerational trauma as well as structural inequities, such as high unemployment. American Indians suffer high rates of alcoholism and suicide as significant causes of death (LeMaster, Beals, Novins, Manson, & AI-SUPERPFP Team, 2004; Mullany et al., 2009; National Center for Health Statistics, 2011). A 2011 Centers for Disease Control and Prevention (CDC) report of Youth Risk Behavior Survey data between 2001 and 2009 noted that AI/AN youth reported higher rates for “ever smoked cigarettes” (71.2%) compared with White youth (54.6%); “ever drank alcohol” (78.8%) compared with Black youth (69.2%); and binge drinking (30.9%) compared with Black youth (12.9%; Jones, Anderson, Lowry, & Conner, 2011). New Mexico has a significant American Indian presence at 11% of the population, with rich historical traditions, including Pueblo ancestry, since time immemorial (four major language groups); three Apache nations; and close to half of the Navajo Nation, the largest tribe in the United States. Native youth in New Mexico (more than 32% were 17 years old and younger) have excess rates of risky behavior. In 2009, 24% of New Mexico 200       belone et al.

high school AI youth were current cigarette smokers, 25% binge drank, and 9.7% reported a suicide attempt in the past year (New Mexico State Center for Health Statistics, 2009). In addition, 40% of AI youth reported parents/ adults not setting boundaries, 68% stated they lacked meaningful community participation, and only 44% reported social competencies to negotiate negative opportunities (New Mexico Department of Health, 2003). These social and health disparities, along with the strengths of their cultural and language continuity, provide an optimal environment for engaging in participatory research with New Mexican tribal communities, based on authentic partnership and tribal oversight of research processes. This chapter provides an overview of principles and strategies for engaging in CBPR with AI/AN communities. To illustrate these strategies, we provide an example of a 13-year tribal–academic partnership between the University of New Mexico Center for Participatory Research (UNM CPR) and the Pueblo of Jemez (POJ). We showcase our processes for codeveloping a culturally centered prevention and intervention program with a tribal community partner, with a focus on strengthening families, language, and culture. We describe how we blended indigenous/Western theory to cocreate and implement a culturally centered prevention curriculum, the Family Listening Program (FLP), rather than a tailored or adapted program. By illustrating the importance of grounding methodologies in culturally centered principles that resonate with the community’s values, we hope to directly address historical wrongs and explicitly acknowledge and positively engage conflict to produce growth for all collaborators in the research process. We discuss some of the challenges in engaging in CBPR with tribes and the implications for incorporating CBPR principles into psychological research and practice. In closing, we present lessons learned for extending translational culturally centered research into other communities of color and having an impact on reducing health inequities within diverse ethnocultural populations. GUIDING PRINCIPLES OF COMMUNITY ENGAGEMENT WITH NATIVE COMMUNITIES In research, adherence to certain principles provides guidance toward ethical actions throughout the research process. Ethical health research with ethnocultural groups must acknowledge and address two significant challenges: (a) historical mistrust of Western Eurocentric systems, institutions, and methodologies that have a legacy of harm; and (b) the power differential between Western Eurocentric and ethnocultural persons, perspectives, and systems (Trimble & Fisher, 2005) that often lead to conflict and discrediting of non-Western cultural values. These issues are largely unaddressed in community-based participatory research     

201

the enforceable code of conduct for psychologists (American Psychological Association [APA], 2002), although the addition of multicultural guidelines (APA, 2003) encourages psychologists to acknowledge their cultural, social, and professional position with a responsibility to uphold social justice and racial equity. More comprehensively, Trimble and Fisher (2005) offered contrasting ethics for research with ethnocultural communities that include relationship-based research, collaborative and participatory approaches, and individual, community, and institutional rights and responsibilities. Participatory research has a growing positive history with communities of color. In particular, CBPR outlines aspirational principles that more closely align with ethnocultural community values. For example, general research and psychological ethical principles specify beneficence, protection, and responsibility to the individual (APA, 2002) whereas CBPR recognizes the community as a principal unit of consideration in research (Minkler, Garcia, Rubin, & Wallerstein, 2012). Thus, CBPR recognizes the interconnectedness of ethnic/cultural individuals with their communities of origin and the communal impact of research findings. Significant for addressing historical mistrust and power imbalance, CBPR specifies that equitable research partnerships should “attend to social inequalities” and “openly address issues of race, ethnicity, racism, and social class” (Minkler et al., 2012, p. 12). Although generic CBPR principles may have great fit or adaptability for research with ethnocultural communities, community-based researchers also need to be guided by the values and perspectives of the specific community with whom they work. In the past decade, researchers, research centers, and their AI/AN partners have begun to define principles or guidelines specific to AI/AN communities (Christopher, Watts, McCormick, & Young, 2008; Fisher & Ball, 2003; LaVeaux & Christopher, 2009; NCAI Policy Research Center, 2012; Straits et al., 2012; Walters et al., 2008). Some common threads include the explicit recognition and impact of tribal sovereignty and historical trauma on research processes. The guidelines also share a common appreciation of existing knowledge within Native communities and the right of Native people to use their own knowledge and people from which to generate research. In addition, concepts of continual dialogue, time, decolonization, and tribal diversity all require in-depth understanding within a specific AI/AN cultural context. Each concept emphasizes the diversity among Native American communities (e.g., tribe, Pueblo, nation, federally recognized, state recognized, urban Indian, communities within communities such as Navajo chapters) and acknowledges that each community may have distinct values. Christopher et al. (2011) demonstrated that the application of research principles specific to AI/AN communities enhances researchers’ and partner communities’ ability to positively confront and advance through issues of power and trust. 202       belone et al.

HISTORY OF UNIVERSITY OF NEW MEXICO–TRIBAL PARTNERSHIPS UNM CPR consolidated in 2006 to create a unified mission and “core values of community partnership, health equity, and participatory engagement in order to co-create new knowledge and translate existing knowledge, to improve quality of life among New Mexico’s diverse population.” (For more information about UNM CPR, visit http://cpr.unm.edu.) UNM CPR’s participatory practices have evolved over time, in particular with tribes, to include recognition of tribal sovereignty in terms of oversight and ownership of research data; commitment to working with community advisory councils (CACs) or tribal research teams, which represent both tribal program staff and community members; and dedication to supporting culturally centered prevention and intervention programs that promote language and cultural connection as protective factors for community well-being. UNM CPR started its partnership with the Jemez in 1999 with a CDC CBPR grant to assess and strengthen community and cultural capacities to improve tribal health systems and health status. Jemez is a federally recognized tribe located 50 miles northwest of Albuquerque, New Mexico, with more than 3,400 enrolled tribal members living in a single village, known by its members as Walatowa (POJ, 2013). With a young population of 38% adolescents and 40% young adults, Jemez faces challenges similar to those of many other tribes: excess rates of childhood overweight and obesity, substance abuse problems, and lower educational achievement. Yet, the POJ is also rich in culture, language, and capacities. According to the 2002 Jemez tribal census, 95% of tribal members spoke their native language, Towa, a language that no other tribal community is known to speak. The Jemez Department of Education (DOE) founded a charter high school in 2002, which strengthened the Towa language programming from Head Start through high school and increased numbers of youth entering the university. The Jemez Health and Human Services Department (HHSD), which took over its own health care from the Indian Health Service in 1999, has initiated and collaborated on multiple prevention initiatives, such as a growers’ market, bicycle and runners clubs, and nutrition classes focusing on traditional foods. The CDC study focused its inquiry, through key informant interviews and focus group discussions, on cultural capacities and community members’ interpretations of their sociocultural strengths (rather than on health disparities) and identified the interconnectedness between the built, sociocultural, and natural environments and the importance of maintaining cultural integrity for promoting good health (Wallerstein et al. 2003). The study produced Community Voices reports, outlining people’s visions for their future as well as highlighting several concerns: (a) a breakdown in communication community-based participatory research     

203

intergenerationally, (b) a lack of family support and unity, (c) an increase in substance abuse, (d) desire for greater engagement between youth and elders, and (e) desire to promote use of the Towa language and knowledge of traditions. The study participants articulated that when tribal members had a direct relationship with alcohol or substance abuse they soon became marginalized, losing critically important collective Jemez strengths and support. Participants shared that interventions to prevent substance abuse needed to include tribal values, participation and knowledge of cultural practices, support for the Towa language, and increased sense of community. The literature also recognizes the importance of passing on traditional and cultural wisdom as protective factors that must be considered for interventions in a tribal context (Belone et al., 2012; Duran & Duran, 2000; Goodkind et al., 2010; Mmari, Blum, & TeufelShone, 2010). Similar concerns were also expressed by a different tribe with whom UNM CPR started partnering in 2000 (Oetzel et al., 2011). Findings led to a request by both tribes for UNM CPR to codevelop a research grant to address concerns within a family context, which received funding from 2004 to 2009. THE FAMILY LISTENING PROGRAM/FAMILY CIRCLE PROGRAM The empirical finding of miscommunication between elders and youth and desires to support cultural renewal led to the development of a National Institutes of Health (NIH) research grant to prevent substance use and initiation in 4th and 5th graders by strengthening cultural values and communication skills through development of an intergenerational prevention program. In searching nationally for successful tribal prevention and/or intervention programs, the NIH funded the Anishinabe Bii-Zin-Da-De-Dah program, a partnership with the University of Nebraska (UN). Translated as Listening to Each Other, this program was identified as the only, at the time, intergenerational family intervention to reduce drug and alcohol abuse with evidence of effectiveness. As a psychocultural and psychoeducational prevention intervention, resulting in the development of a curriculum that combined both cultural messages and sanctions with mainstream parenting communication skills, the Bii-Zin-Da-De-Dah program found that culturally embedded prevention messages were more effective and retained with the Anishinabe youth (Whitbeck, 2001). Although the prevention program was initiated with middle school youth, parents, and elders, the patterns of alcohol use among this age group led the Anishinabe–UN partnership to reconsider aiming their program to elementary school age children and their families. The UNM CPR and Jemez partnership built upon the learnings and consultations from Whitbeck and his Anishinabe partners and received a 204       belone et al.

Native American Research Centers for Health (NARCH III) research grant (2004–2009) to use a CBPR approach in the development, piloting, and implementation of a culturally centered intergenerational intervention, initially called the FLP, to reduce alcohol and other drug initiation, use, and abuse among Jemez late elementary school youth. The NARCH national funding initiative, a partnership between the Indian Health Service and the NIH, enabled us to write a grant with a recommitment to tribal oversight and a participatory approach in each step of the research process. NARCH was created in 2000 with three major goals: to reduce research mistrust among tribes and academic institutions, to increase the pipeline of American Indian researchers, and to reduce health disparities in Indian Country. The uniqueness of NARCH is that the principal investigator of the NIH-funded grant must be a tribal entity, which supports tribal research capacity development and allows them to choose their academic research partners. The development of the FLP was based on an in-depth CBPR cultural centering process (discussed in the next section) as well as a commitment to incorporate three perspectives: cultural centeredness and indigenous theory, a public health model of risk and protective factors, and empowerment theory based on the educational philosophy of Paulo Freire (1970). Starting from the Anishinabe curriculum, which involved multiple generations in an after-school dinner-based program with interactive activities, we proposed to spend the first year in recentering and recreating the prevention curriculum within the cultural values, history, visions, and communication skills traditional to Jemez. Cultural centeredness, in contrast to culturally tailored approaches, reflects an understanding of health and disease processes as deeply embedded in complex and dynamic cultural contexts (Dutta, 2007). Dutta (2007) considered cultural centeredness to be a set of cultural processes and noted the importance of community voice and agency in decision making, reflecting a perspective parallel to that of CBPR. Moran and Reaman (2002) found that many American Indian youth programs attempt to adapt mainstream programs by adding cultural elements, but this add-on approach is inadequate. Beyond tailoring, cultural centeredness seeks to create knowledge and reciprocal learning, where interventions can integrate culturally supported indigenous practices and community dialogue (Dutta, 2007). Indigenous knowledge theory, according to which culture, language, and community are central to learning (Cajete, 1995; Pankratz et al., 2006), can promote protective factors of prosocial relationships and cultural identity and values of tribal interdependence and responsibility. Programs based on this theory support youth self-efficacy, with AI youth more likely to seek advice or support from adults other than parents (Beebe et al., 2008). Indigenous knowledge applies cultural mentorship to center beliefs and practices by explaining community-based participatory research     

205

problems and offering solutions within prevention programs (Cajete, 1999; King, Smith, & Gracey, 2009; Tuhiwai-Smith, 2005). Indigenous theory suggests that specific behavioral sanctions against deviant drug and alcohol behaviors have been weakened through discrimination and disruption of cross-generational teachings on traditional behaviors (Duran & Duran, 2000; Duran et al., 1998). Various literatures find cultural connectedness and identity positively associated with health, Hopi traditional practices as inversely associated with smoking and obesity, and cultural adult role models and peer groups as associated with AI alcohol abstinence (Beals et al., 2005; Spicer, Novins, Mitchell, & Beals, 2003) and with increasing AI youth commitment to preventing substance abuse (Ringwalt & Bliss, 2006), strengthening antidrug norms (Kulis, Napoli, & Marsiglia, 2002), and enhancing family communication (Whitbeck, Hoyt, Stubben, & LaFromboise, 2001). The second perspective used included the application of the public health socioecologic multilevel model to guide the intervention, which embraces individual and family risk and protective behaviors in their broader cultural, social, and economic contexts. The FLP Conceptual Model (see Figure 9.1) incorporates the literature on risk factors and integrates evidence-based theories of behavior change with cultural practices, adult mentors, and values to support child, family, and community outcomes, such as increased child resiliencies and empowerment, increased parent–child communication, increased community and cultural participation, and support of kids to live biculturally in two worlds. As reflected in the conceptual model, we also drew from the child and adolescent literature that parenting behaviors have the most impact during childhood, including from the multistage social learning model, social development model (Hawkins & Weis, 1985), and family interaction theory (Brook, Brook, Gordon, Whiteman, & Cohen, 1990). We incorporated characteristics of effective prevention programs for youth substance use, suicide ideation, school failure, and violence with sufficient dosage, theory-driven, and culturally and socially relevant messages with well-trained staff and mentors and appropriately timed in children’s lives for maximum impact (Nation et al., 2003). More simply stated, we used information from programs that had ageappropriate peer interactions, caregiver–parent components, and key skill development exercises (Catalano, Hawkins, Berglund, Pollard, & Arthur, 2002; Perry et al., 2007; Perry, Stigler, Arora, & Reddy, 2009). The FLP was intended to fit the above criteria and add tribal language and cultural dimensions to increase cultural communication and relationship skills between three generations of children, parents, and elders and therefore reinforces identification with protective cultural norms and values. Finally, the FLP expanded beyond the Anishinabe model by integrating empowerment theory. Empowerment, defined as a social action process in 206       belone et al.

Figure 9.1.  Family Listening/Circle Program conceptualized model.

which individuals gain mastery over their lives as they act to make changes in their social environment to improve equity and quality of life (Wallerstein, 2002), has been linked with psychological and community health outcomes, especially with youth (Holden, Crankshaw, Nimsch, Hinnant, & Hund, 2004; Wallerstein, 2006). FLP went beyond strengthening internal family communication—it engaged families in community action projects (CAPs), allowing service back to their community. CAPs resonated with existing cultural systems of community responsibility and accountability. Based on the reflexive processes of Freire (1970), FLP integrated the listening–dialogue– action cycle of children with their parents and elders listening to each other’s visions and concerns, having dialogue about how they could address these concerns to reach their visions, and structuring concrete community actions they could take. OUR CBPR PROCESSES At the time of receipt of the NARCH III grant, the UNM CPR team consisted of the principal investigator, a White Jewish CBPR faculty researcher, and a research scientist, a (Navajo) master of public health (MPH) student who went on to receive her doctorate and a faculty position. Both were involved in the 1999–2003 CDC-funded Jemez CBPR grant with its CAC to guide the research process and had established a relationship built on trust; community CAC members stated they valued that they were in the driver’s community-based participatory research     

207

seat in creating the interview guides and collecting and coanalyzing the data. Over the 4 years of the FLP, the research team expanded to include highly skilled native graduate researchers, including an MPH student who was a member of Santa Clara Pueblo and a master in community and regional planning student who was a member of the Jicarilla Apache nation, both of whom became research scientists with the CPR; an MPH Navajo student who later entered medical school; and an undergraduate Jicarilla student. Other students participated at different times. This unique, primarily American Indian UNM CPR team and Jemez partner entered into a 4-year research project. The first year’s aim was to strengthen and expand the Jemez CAC to include tribal health providers, educators, parents, elders, high school youth, community leaders, and others. The CAC of 10 to 12 core members, often expanding to 20 people including five or six core elders, met monthly and at times weekly to (a) review the CDC Community Voices reports to identify core issues to be included in the program; (b) conduct focus groups with parents, elders, youth, and service providers on age-appropriate cultural stories, history, and values for elementary school age children; and (c) produce a Jemez version of the Anishinabe Bii-Zin-Da-De-Dah curriculum centered in Jemez history and values. Youth participated in creating a video for the program by interviewing tribal leaders and bringing in an explanation of the tribal seal. The CAC named their curriculum the Hemish of Walatowa Family Circle Program (FCP), which was completed after a year and half of numerous iterations. This name change reflected the importance of Hemish from their own Towa language, rather than the imposition of Jemez from the Spanish language; the idea of a family circle reflected the life circle of all ages; and the name Walatowa was their village of origin. To center the curriculum within Jemez, the CAC with UNM CPR painstakingly reviewed each of the Anishinabe sessions and incorporated their own focus group data and personal knowledge of Jemez culture and values. This process was extensive, as cultural centering delves deeper than curriculum tailoring through a quick rewrite or adding community-appropriate images. With the unique UNM CPR Native researchers, the team recognized the importance of listening carefully to the elders during the CAC meetings and providing as much time as needed for them to speak about their issues in their own language. For example, during the re-creation of the anger management session, lengthy discussion in Towa ensued on the many terms for expressing anger in their language and on the traditional ways for helping kids with their anger. Most of this discussion was not translated back to the research team; ultimately, the elders led the decision of what to include during that week’s session. The end product was a collective work of a detailed 202-page familystrengthening curriculum consisting of 14 weekly sessions, embedding state educational standards, drug and alcohol prevention messages, and other health 208       belone et al.

promotion information for children, families, and community, while reinforcing Hemish traditions, history, and knowledge and the Towa language. Whereas the original Bii-Zin-Da-De-Dah focused on internal family dynamics, the Jemez CAC expanded the curriculum to include empowerment-based strategies for community change by incorporating a community visioning process, an analysis of community concerns, and inclusion of a community action project, allowing families the opportunity to identify, discuss, and plan to address a community concern, such as littering or speeding. Children took pictures of their community action project using PhotoVoice to create an educational display for larger community viewing. A facilitator’s manual was also developed to guide a facilitator through each session. Although the FLP curriculum was written in English, facilitation by CAC members was predominantly conducted in the Towa language. Facilitators often practiced verbal translation of each session during guided facilitation training prior to each session. In addition to developing a curriculum, the CAC and UNM CPR took the time to review and adapt process and pre- and posttest outcome measures, incorporating national scales of substance use and abuse, depression, and anxiety. Scales were also developed on levels of cultural participation, resulting in an increase in internal validity of the measures for Jemez. By the third year, the CAC recruited 10 families with fourth and fifth graders (with their parents and/or grandparents) to pilot-test the curriculum. A second pilot-test of the curriculum was conducted during the fourth year of the project. Because of the intensive work it took to run the program and to recruit the children, the FCP transitioned from the Jemez HHSD to the Jemez DOE, recognizing its greater connection to the schools, parents, and families. The Jemez DOE also had direct access to teachers, who became key players as facilitators in delivering the curriculum. This chapter focuses on CBPR partnering processes, yet preliminary outcome data have been published showing efficacy in the pilots and promise of a culturally centered and evidence-based approach (Shendo et al., 2012). Our initial work in Jemez found Hawe, Shiell, and Riley’s (2009) framework of complex interventions within diverse communities helpful: They asserted that adapting or integrating prevention programs into dynamic contexts demands a focus on the underlying functions of the program. The actual form that the program components will take in any given community needs to be unwrapped and translated by each CAC as they incorporate their own assessment and planning processes to situate the intervention within their own cultural context, which includes day-to-day relationships in tribal communities, which is beyond the view of academic researchers. Privileging difference between communities and encouraging different forms (i.e., specific composition of CACs; different presentations of values, history, or stories; different instructions for activities) avoids recolonization by researchers lacking this community-based participatory research     

209

insight. The core evidence base of the program, for example, may be to provide a history session or a problem-solving session, but how this is done will differ depending on cultural applicability. Hawe et al.’s (2009) understanding leads prevention intervention researchers to tackle the quandary of how to promote more external validity, even as it is important to test internal validity (Cargo & Mercer, 2008; Glasgow & Emmons, 2007; Green & Glasgow, 2006). The UNM CPR has implemented this participatory process with two other tribes through NARCH funding. Similar to Jemez, each tribe recentered the curriculum using their own reflections of their CACs, their own focus group responses, and other resources from within their communities, such as their traditional language programs. Participatory research demands flexibility of the research team if the goal is to embed programs within specific cultures with equalizing distribution of power (Muhammad et al., 2015). In one community, for example, rather than a short focus group, the elders met for 6 hours listening to each other tell their stories and then, after that day, stated that they hadn’t finished and wanted time for a second and third focus group. The three focus groups were taped in their language, so translation was needed, which took several months. The different timeframe standard produced a rich inclusion of community perspectives on childhood developmental life cycles and important historical moments for the tribe, which would not have been known if the elders’ wisdom had not been solicited (Belone et al., 2012). CHALLENGES AND LESSONS LEARNED Several important lessons concerning ownership, adaptability, and sustainability emerged from the design and implementation of the Hemish of Walatowa FCP. Tribal ownership of the FCP was an intended outcome from the onset of the study based on the utilization of a CBPR approach, which includes the community partner in every step of the research process, from development of research questions to intervention implementation and finally analysis and dissemination. The commitment to patience and flexibility was important because of the time (including extensive travel time to meet in the community) needed to center the curriculum in Jemez culture and values by listening to the elders and involving youth. This process took close to a year and a half in the creation of the Jemez-based curriculum. Partnering alone, however, may not ensure a sense of community ownership. In this case, the combination of several formal agreements was important, including a memorandum of agreement, project approval processes (i.e., by tribal government and health board), a CAC of community representatives (service providers, teachers, elders, parents, and youth) to provide guidance and wisdom, and ultimately recognition of tribal sovereignty regarding ownership of data and 210       belone et al.

the project itself. Meeting in the community monthly (and sometimes weekly during intense development and implementation times) demonstrated that the research team was committed to authentic communication and integration of community voices. Although other ethnocultural communities might not share the authority of tribal sovereignty, participatory research teams and partnerships that honor the time it takes to create or re-create culturally centered programs can be developed. Because of its growth out of a participatory process to center the curriculum within Jemez values and across multiple programs, other opportunities arose for FCP to become sustainable within the culture and community. For example, the Jemez DOE incorporated traditional Jemez foods in the second FCP pilot. The meals, which were prepared by the Jemez Nutrition Program, with the assistance of youth from the Native American Youth Empowerment program, exposed participants to forgotten recipes and reinforced traditional and cultural values, community norms, and tribal history. A direct outcome was a traditional foods and recipes booklet distributed to community members, which also strengthened organizational linkages between Jemez HHSD and the Jemez DOE. An unexpected outcome regarding sustainability involved further adaptation of the FCP curriculum by the Jemez Summer Youth Program. Over an 8-week period in the summer of 2011, nearly 100 children, ages 7 to 16, were reached, which extended FCP’s impact far beyond the smaller family dinner-based structure. The UNM CPR team was asked to provide evaluation assistance for the Summer Youth Program, continuing an active partnership. FCP facilitators who were elementary school teachers also have reported using different components in their classrooms. The capacity for FCP to be adapted for priority community needs has resulted in continued tribal use, increased cultural connections by the children and youth as outcomes (as expressed in the posttests), and application in new settings. Sustainability is a challenge for any new intervention competing with existing programs for resources. The decision to integrate elements of the curriculum into various venues has been positive, yet the need to consider sustainability remains. One illustration of this challenge is around publication of results. Tribal ownership of data has been respected and clearly outlined in this partnership; therefore, publication of results has to be part of tribal priorities that typically include informing leadership, program managers, and the tribal community at large, rather than publishing to the external world. It is only now after 3 years of the grant ending, and the deepening of ownership of the FCP within Jemez, that the CAC is seeking to share the effectiveness of its approach. The willingness to publish also reflects a deepening partnership and trust between UNM CPR and the Jemez Advisory Council as well as other tribal programs. It is important that we continue in a manner that supports mutual respect and colearning. community-based participatory research     

211

Financing has also been a challenge, as NIH grants typically go to universities, with fewer resources going to the community. Funder time lines can often be shorter than community time lines, though NARCH grants have typically understood the importance of tribal planning processes. Although we advocate for memoranda of agreement, simple administrative processes can also be a challenge in implementing agreements. These need to be clearly defined and always renegotiated to smooth exchange of funds and responsibilities. In sum, our experience adds to the experience of others who have documented the challenges to consider when engaging in CBPR with Native American communities (Allen et al., 2006; Ball & Janyst, 2008; Thomas, Rosa, Forcehimes, & Donovan, 2011). These challenges include (a) time lines of funders versus community time lines, which may include unanticipated community events; (b) extent of approval processes, which differ by individual tribes and can include health boards, program directors, tribal administration, tribal governments, tribal councils, and tribal institutional review boards; (c) extensive travel often required for meeting attendance and program implementation; (d) challenges of maintaining funding; and (e) the methodological challenge of research objectivity while building rapport and authentic relationships. Some funding also prohibits payment for food, which is a critically important component to honor community member participation, especially if community members are volunteering to be part of community advisory boards. However, we believe the additional challenge to spend the time to fully integrate and center the culture and context into the interventions is essential for community ownership and sustainability. Sustainability is essential to see health outcomes improve over time, far beyond any specific grant cycles. These lessons learned have provided key points for consideration and action for research partnerships in AI/AN communities as well as other ethnocultural communities in the development of prevention and intervention programs. IMPLICATIONS FOR PSYCHOLOGICAL PRACTICE AND RESEARCH WITH ETHNOCULTURAL COMMUNITIES As illustrated above, participatory research with ethnocultural communities is an intensely rich process that involves equitable partnerships with interdisciplinary teams of professionals and community members, recentering paradigms within a specific community and cultural context, constant reflective communication, and deepening relationships over an extended period. This example highlights the complexities of CBPR that necessitate a principled approach to research. UNM CPR’s work with Jemez Pueblo and other AI communities in the Southwest reflects a deep respect for tribal sovereignty 212       belone et al.

and self-determination essential to AI-specific principled research (Fisher & Ball, 2003, 2005; NCAI Policy Research Center, 2012), as demonstrated by tribal oversight, tribal decisions on research steps (in their own language), and a primarily AI research team. The FLP team incorporated many of the key principles for working with Native communities, such as honoring community time frames and processes (LaVeaux & Christopher, 2009; Straits et al., 2012). From initiating a project with the community beginning in 1999 continuing through the present to conducting daylong focus groups and then to respecting the community’s readiness and manner of dissemination and publication, even if publication occurs multiple years after the project ends, a principled participatory research approach presents an alternate methodology for engaging with ethnocultural groups that better addresses historical mistrust and values differences while still maintaining rigorous research standards. Implications for participatory research with other ethnocultural groups would be to identify the values underlying the research methodologies and how they may better honor and uphold cultural and community values to promote social justice and reduce the likelihood of harm. For example, although tribal sovereignty may not apply to other groups, CBPR with immigrant Latino communities may incorporate a parallel principle that respects immigration status, citizenship, and documentation. Culturally relevant research principles also provide communities with a standard of conduct to which they can hold researchers. Culturally centered participatory research has vast implications for impacting mental health disparities experienced by racial/ethnic groups including poor access to and quality of care, lower utilization, greater stigma and discrimination, lack of culturally and linguistically competent providers, and underrepresentation in research and clinical trials (McGuire & Miranda, 2008). CBPR provides needed data on culturally centered interventions for improving access and quality of care for ethnic minorities. Although further research is needed to validate the assumptions in the use and positive impact of participatory research methods, they represent a research tool that may offer greater suitability to the mental health needs of ethnocultural communities who otherwise are not provided a voice in their own health care. Inviting the community to participate at the research level of developing mental health interventions initiates a process of change that becomes more sustainable and enduring as community members and organizations become more invested, the research process generates community health education (e.g., FLP research process led to a traditional foods and recipes booklet), community members’ expertise becomes elevated and enhances outreach to more individuals, and the community’s culture shapes interventions to become an integral part of community function. To more fully engage with and address sociocultural determinants of mental health and community-based participatory research     

213

reduce mental health disparities, the research methodologies used must have the capacity to embrace community through the equitable and active engagement of all research partners that increases access to, quality of, and utilization of mental health services. In regard to psychological practice, we need to reconceptualize our notions of psychological interventions and how we use research to generate evidencebased practices. As Kazdin and Blase (2011) and others have noted, individual, couples, or family therapy may be effective at a one-on-one level for a small subset of the population but has not helped to reduce mental health disparities. Kazdin and Blase called for “a portfolio of models of delivery” (p. 507), including mixing prevention with intervention, collaborating across professions, and providing opportunities for nonprofessionals to provide psychological interventions. Chin, Walters, Cook, and Huang (2007) reviewed promising interventions for reducing health disparities that included multifactorial and culturally tailored approaches creating linkages between communities and health care systems that provide insight into implementation in realworld settings. CBPR takes a grassroots approach to mental health change, integrating multidisciplinary research experts, community organizations, and community members and experts and combining research with intervention to create an environment for change that may be more likely to have immediate and widespread effect, while also developing individual or group interventions backed by data in real-world settings with proven effectiveness for ethnic minority groups. Revisioning participatory research and interventions for specific communities holds significant promise in reducing mental health disparities by combining the psychological knowledge of our field, the discipline and data of research, and the cultural wisdom of our communities to produce culturally and scientifically validated interventions with positive effects at multiple socio-ecological levels.

REFERENCES Allen, J., Mohatt, G. V., Rasmus, S. M., Hazel, K. L., Thomas, L., & Lindley, S. (2006). The tools to understand: Community as coresearcher on culture-specific protective factors for Alaska Natives. Journal of Prevention & Intervention in the Community, 32(1–2), 41–59. http://dx.doi.org/10.1300/J005v32n01_04 American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. American Psychologist, 57, 1060–1073. http://dx.doi. org/10.1037/0003-066X.57.12.1060 American Psychological Association. (2003). Guidelines on multicultural education, training, research, practice, and organizational change for psychologists. Retrieved from http://www.apa.org/pi/oema/resources/policy/multicultural-guideline.pdf

214       belone et al.

Baldwin, J. A., Johnson, J. L., & Benally, C. C. (2009). Building partnerships between indigenous communities and universities: Lessons learned in HIV/AIDS and substance abuse prevention research. American Journal of Public Health, 99(Suppl. 1), S77–S82. http://dx.doi.org/10.2105/AJPH.2008.134585 Ball, J., & Janyst, P. (2008). Enacting research ethics in partnerships with indigenous communities in Canada: “Do it in a good way”. Journal of Empirical Research on Human Research Ethics: An International Journal, 3(2), 33–51. http://dx.doi. org/10.1525/jer.2008.3.2.33 Beals, J., Manson, S. M., Whitesell, N. R., Mitchell, C. M., Novins, D. K., Simpson, S., & Spicer, P. (2005). Prevalence of major depressive episode in two American Indian reservation populations: Unexpected findings with a structured interview. The American Journal of Psychiatry, 162, 1713–1722. http://dx.doi. org/10.1176/appi.ajp.162.9.1713 Becenti-Pigman, B., White, K., Bowman, B., Palmanteer-Holder, N. L., & Duran, B. (2008). Research policies, processes, and protocol: The Navajo Nation Human Research Review Board. In M. Minkler & N. Wallerstein (Eds.), Communitybased participatory research for health: From process to outcomes (2nd ed., pp. 441–445). San Francisco, CA: Wiley. Beebe, L. A., Vesely, S. K., Oman, R. F., Tolma, E., Aspy, C. B., & Rodine, S. (2008). Protective assets for nonuse of alcohol, tobacco, and other drugs among urban American Indian youth in Oklahoma. Maternal and Child Health Journal, 12(Suppl. 1), 82–90. http://dx.doi.org/10.1007/s10995-008-0325-5 Belone, L., Lucero, J. E., Duran, B., Tafoya, G., Baker, E. A., Chan, D., . . . Wallerstein, N. (2014). Community-based participatory research conceptual model: Community partner consultation and face validity. Qualitative Health Research. Advance online publication. http://dx.doi.org/10.1177/1049732314557084 Belone, L., Oetzel, J. G., Wallerstein, N., Tafoya, G., Rae, R., Rafelito, A., . . . Thomas, A. (2012). Using participatory research to address substance use in an American Indian community. In L. R. Frey & K. Carragee (Eds.), Communication activism (6th ed., pp. 403–434). Cresskill, NJ: Hampton Press. Brook, J. S., Brook, D. W., Gordon, A. S., Whiteman, M., & Cohen, P. (1990). The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs, 116, 111–267. Burhansstipanov, L., Christopher, S., & Schumacher, S. A. (2005, November). Lessons learned from community-based participatory research in Indian Country. Cancer Control, 12(Suppl. 2), 70–76. Cajete, G. (1995). Look to the mountain: An ecology of Indigenous education. Durango, CO: Kivaki Press. Cajete, G. (1999). Ignite the sparkle: An Indigenous science model. Skyland, NC: Kivaki Press. Cargo, M., & Mercer, S. L. (2008). The value and challenges of participatory research: Strengthening its practice. Annual Review of Public Health, 29, 325–350. http:// dx.doi.org/10.1146/annurev.publhealth.29.091307.083824 community-based participatory research     

215

Catalano, R. F., Hawkins, J. D., Berglund, M. L., Pollard, J. A., & Arthur, M. W. (2002). Prevention science and positive youth development: Competitive or cooperative frameworks? Journal of Adolescent Health, 31(Suppl. 6), 230–239. http://dx.doi.org/10.1016/S1054-139X(02)00496-2 Chin, M. H., Walters, A. E., Cook, S. C., & Huang, E. S. (2007). Interventions to reduce racial and ethnic disparities in health care. Medical Care Research and Review, 64(Suppl. 5), 7S–28S. http://dx.doi.org/10.1177/1077558707305413 Christopher, S., Saha, R., Lachapelle, P., Jennings, D., Colclough, Y., Cooper, C., . . . Webster, L. (2011). Applying indigenous community-based participatory research principles to partnership development in health disparities research. Family & Community Health, 34, 246–255. http://dx.doi.org/10.1097/FCH. 0b013e318219606f Christopher, S., Watts, V., McCormick, A. K. H. G., & Young, S. (2008). Building and maintaining trust in a community-based participatory research partnership. American Journal of Public Health, 98, 1398–1406. http://dx.doi.org/10.2105/ AJPH.2007.125757 Duran, B., & Duran, E. (2000). Applied postcolonial research and clinical strategies. In M. Battiste (Ed.), Reclaiming Indigenous voice and vision (pp. 86–100). Vancouver, Canada: University of British Columbia Press. Duran, B., Duran, E., & Yellow Horse Braveheart, M. (1998). Native Americans and the trauma of history. In R. Thornton (Ed.), Studying Native America: Problems and prospects in Native American studies (pp. 60–76). Madison: University of Wisconsin Press. Dutta, M. J. (2007). Communicating about culture and health: Theorizing culturecentered and cultural sensitivity approaches. Communication Theory, 17, 304–328. http://dx.doi.org/10.1111/j.1468-2885.2007.00297.x Fisher, P. A., & Ball, T. J. (2003). Tribal participatory research: Mechanisms of a collaborative model. American Journal of Community Psychology, 32, 207–216. http://dx.doi.org/10.1023/B:AJCP.0000004742.39858.c5 Fisher, P. A., & Ball, T. J. (2005). Balancing empiricism and local cultural knowledge in the design of prevention research. Journal of Urban Health, 82(Suppl. 3), iii44–iii55. Freire, P. (1970). Pedagogy of the oppressed. New York, NY: The Seabury Press. Glasgow, R. E., & Emmons, K. M. (2007). How can we increase translation of research into practice? Types of evidence needed. Annual Review of Public Health, 28, 413–433. http://dx.doi.org/10.1146/annurev.publhealth.28.021406.144145 Goodkind, J. R., Ross-Toledo, K., John, S., Hall, J. L., Ross, L., Freeland, L., . . . Lee, C. (2010). Promoting healing and restoring trust: Policy recommendations for improving behavioral health care for American Indian/Alaska Native adolescents. American Journal of Community Psychology, 46, 386–394. http://dx.doi. org/10.1007/s10464-010-9347-4 Green, L. W., & Glasgow, R. E. (2006). Evaluating the relevance, generalization, and applicability of research: Issues in external validation and translation meth-

216       belone et al.

odology. Evaluation & the Health Professions, 29, 126–153. http://dx.doi.org/ 10.1177/0163278705284445 Hawe, P., Shiell, A., & Riley, T. (2009). Theorising interventions as events in systems. American Journal of Community Psychology, 43, 267–276. http://dx.doi. org/10.1007/s10464-009-9229-9 Hawkins, J. D., & Weis, J. G. (1985). The social development model: An integrated approach to delinquency prevention. Journal of Primary Prevention, 6, 73–97. http://dx.doi.org/10.1007/BF01325432 Holden, D. J., Crankshaw, E., Nimsch, C., Hinnant, L. W., & Hund, L. (2004). Quantifying the impact of participation in local tobacco control groups on the psychological empowerment of involved youth. Health Education & Behavior, 31, 615–628. http://dx.doi.org/10.1177/1090198104268678 Israel, B. A., Eng, E., Schulz, A. J., & Parker, E. A. (Eds.). (2013). Methods for community-based participatory research for health (2nd ed.). San Francisco, CA: Jossey-Bass. Jones, S. E., Anderson, K., Lowry, R., & Conner, H. (2011). Risks to health among Indian/Alaska Native high school students in the United States. Preventing Chronic Disease, 8(A76). Retrieved from http://www.cdc.gov/pcd/issues/2011/ jul/10_0193.htm Kazdin, A. E., & Blase, S. L. (2011). Rebooting psychotherapy research and practice to reduce the burden of mental illness. Perspectives on Psychological Science, 6, 21–37. http://dx.doi.org/10.1177/1745691610393527 King, M., Smith, A., & Gracey, M. (2009). Indigenous health part 2: The underlying causes of the health gap. The Lancet, 374, 76–85. http://dx.doi.org/10.1016/ S0140-6736(09)60827-8 Kulis, S., Napoli, M., & Marsiglia, F. F. (2002). Ethnic pride, biculturalism, and drug use norms of urban American Indian adolescents. Social Work Research, 26, 101–112. http://dx.doi.org/10.1093/swr/26.2.101 LaVeaux, D., & Christopher, S. (2009). Contextualizing CBPR: Key principles of CBPR meet the Indigenous research context. Pimatisiwin, 7(1), 1–25. LeMaster, P. L., Beals, J., Novins, D. K., Manson, S. M., & AI-SUPERPFP Team. (2004). The prevalence of suicidal behaviors among Northern Plains American Indians. Suicide and Life-Threatening Behavior, 34, 242–254. McGuire, T. G., & Miranda, J. (2008). New evidence regarding racial and ethnic disparities in mental health: Policy implications. Health Affairs, 27, 393–403. http://dx.doi.org/10.1377/hlthaff.27.2.393 Mello, M. M., & Wolf, L. E. (2010). The Havasupai Indian tribe case—Lessons for research involving stored biologic samples. The New England Journal of Medicine, 363, 204–207. http://dx.doi.org/10.1056/NEJMp1005203 Minkler, M., Garcia, A. P., Rubin, V., & Wallerstein, N. (2012). Community-based participatory research: A strategy for building healthy communities and promoting health through policy change. Oakland, CA: PolicyLink. community-based participatory research     

217

Minkler, M., & Wallerstein, N. (Eds.). (2008). Community-based participatory research for health: From process to outcomes (2nd ed.). San Francisco, CA: Jossey-Bass. Mmari, K. N., Blum, R. W., & Teufel-Shone, N. (2010). What increases risk and protection for delinquent behaviors among American Indian youth? Findings from three tribal communities. Youth & Society, 41, 382–413. http://dx.doi. org/10.1177/0044118X09333645 Moran, J. R., & Reaman, J. A. (2002). Critical issues for substance abuse prevention targeting American Indian youth. The Journal of Primary Prevention, 22, 201–233. http://dx.doi.org/10.1023/A:1013665604177 Muhammad, M., Wallerstein, N., Sussman, A. L., Avila, M., Belone, L., & Duran, B. (2015). Reflections on researcher identity and power: The impact of positionality on community based participatory research (CBPR) processes and outcomes. Critical Sociology, 41, 1045–1063. http://dx.doi.org/10.1177/0896920513516025 Mullany, B., Barlow, A., Goklish, N., Larzelere-Hinton, F., Cwik, M., Craig, M., & Walkup, J. T. (2009). Toward understanding suicide among youths: Results from the White Mountain Apache tribally mandated suicide surveillance system, 2001–2006. American Journal of Public Health, 99, 1840–1848. http://dx.doi. org/10.2105/AJPH.2008.154880 Nation, M., Crusto, C., Wandersman, A., Kumpfer, K. L., Seybolt, D., MorrisseyKane, E., & Davino, K. (2003). What works in prevention: Principles of effective prevention programs. American Psychologist, 58, 449–456. http://dx.doi. org/10.1037/0003-066X.58.6-7.449 National Center for Health Statistics. (2011). Health, United States, 2011: With special feature on socioeconomic status and health. Hyattsville, MD: Author. National Congress of American Indians Policy Research Center, & Montana State University Center for Native Health Partnerships. (2012). ‘Walk softly and listen carefully’: Building research relationships with tribal communities. Washington, DC, and Bozeman, MT: Authors. New Mexico Department of Health, New Mexico Public Education Department, & The University of New Mexico Center for Health Promotion and Disease Prevention. (2003). New Mexico Youth Risk and Resiliency Survey: 2003 report of state results. Retrieved from http://www.youthrisk.org/pdf/YRRS-2003Report.pdf New Mexico State Center for Health Statistics, Bureau of Vital Records and Health Statistics, Statistics and Epidemiology Unit. (2009). New Mexico selected health statistics: Annual report 2009. Retrieved from http://nmhealth.org/data/view/vital/561/ Oetzel, J., Wallerstein, N., Solimon, A., Garcia, B., Siemon, M., Adeky, S., . . . Tafoya, G. (2011). Creating an instrument to measure people’s perception of community capacity in American Indian communities. Health Education & Behavior, 38, 301–310. Pankratz, M. M., Jackson-Newsom, J., Giles, S. M., Ringwalt, C. L., Bliss, K., & Bell, M. L. (2006). Implementation fidelity in a teacher-led alcohol use prevention

218       belone et al.

curriculum. Journal of Drug Education, 36, 317–333. http://dx.doi.org/10.2190/ H210-2N47-5X5T-21U4 Perry, C. L., Lee, S., Stigler, M., Farbakhsh, K., Komro, K., Gewirtz, A., & Williams, C. (2007). The impact of Project Northland on selected MMPI-A problem behavior scales. The Journal of Primary Prevention, 28, 449–465. http://dx.doi.org/ 10.1007/s10935-007-0105-9 Perry, C. L., Stigler, M. H., Arora, M., & Reddy, K. S. (2009). Preventing tobacco use among young people in India: Project MYTRI. American Journal of Public Health, 99, 899–906. http://doi.org/10.2105/AJPH.2008.145433 Pueblo of Jemez. (2013). Governor’s Office. Retrieved from http://www.jemezpueblo.org/ Governors_Office.aspx Ringwalt, C., & Bliss, K. (2006). The cultural tailoring of a substance use prevention curriculum for American Indian youth. Journal of Drug Education, 36, 159–177. http://dx.doi.org/10.2190/369L-9JJ9-81FG-VUGV Shendo, K., Tosa, J., Tafoya, G., Belone, L., Rae, R., & Wallerstein, N. (2012). The Family Listening Program, the process and outcomes from a CBPR universitytribal partnership. Indian Health Service Provider, 37(8), 185–191. Spicer, P., Novins, D. K., Mitchell, C. M., & Beals, J. (2003). Aboriginal social organization, contemporary experience, and American Indian adolescent alcohol use. Journal of Studies on Alcohol, 64, 450–457. http://dx.doi.org/10.15288/ jsa.2003.64.450 Straits, K. J. E., Bird, D. M., Tsinajinnie, E., Espinoza, J., Goodkind, J., Spenser, O., . . . Guiding Principles Workgroup. (2012). Guiding principles for engaging in research with Native American communities: Version 1. Albuquerque: University of New Mexico Center for Rural and Community Behavioral Health, & Albuquerque Area Southwest Tribal Epidemiology Center. Retrieved from http://psychiatry. unm.edu/centers/crcbh/naprogram/guidingprinciples.html Teufel-Shone, N. I., Siyuja, T., Watahomigie, H. J., & Irwin, S. (2006). Communitybased participatory research: Conducting a formative assessment of factors that influence youth wellness in the Hualapai community. American Journal of Public Health, 96, 1623–1628. http://dx.doi.org/10.2105/AJPH.2004.054254 Thomas, L. R., Donovan, D. M., Sigo, R. L., Austin, L., Marlatt, G. A., & Suquamish Tribe. (2009). The community pulling together: A tribal community–university partnership project to reduce substance abuse and promote good health in a reservation tribal community. Journal of Ethnicity in Substance Abuse, 8, 283–300. http://dx.doi.org/10.1080/15332640903110476 Thomas, L. R., Rosa, C., Forcehimes, A., & Donovan, D. M. (2011). Research partnerships between academic institutions and American Indian and Alaska Native Tribes and organizations: Effective strategies and lessons learned in a multisite CTN study. The American Journal of Drug and Alcohol Abuse, 37, 333–338. http:// dx.doi.org/10.3109/00952990.2011.596976 Trimble, J. E., & Fisher, C. B. (2005). The handbook of ethical research with ethnocultural populations and communities. Thousand Oaks, CA: Sage. community-based participatory research     

219

Tuhiwai-Smith, L. (2005). Decolonizing methodologies: Research and Indigenous peoples. London, England: Zed Books. Wallerstein, N. (2002). Empowerment to reduce health disparities. Scandinavian Journal of Public Health, 30(Suppl. 59), 72–77. http://dx.doi.org/10.1177/ 14034948020300031201 Wallerstein, N. (2006, February). What is the evidence on effectiveness of empowerment to improve health? Retrieved from World Health Organization Regional Office for Europe website: http://www.euro.who.int/__data/assets/pdf_file/0010/74656/ E88086.pdf Wallerstein, N. B., & Duran, B. (2006). Using community-based participatory research to address health disparities. Health Promotion Practice, 7, 312–323. http://dx.doi.org/10.1177/1524839906289376 Wallerstein, N., & Duran, B. (2010). Community-based participatory research contributions to intervention research: The intersection of science and practice to improve health equity. American Journal of Public Health, 100 (Suppl. 1), S40–46. http://dx.doi.org/10.2105/AJPH.2009.184036 Wallerstein, N., Duran, B. M., Aguilar, J., Joe, L., Loretto, F., Toya, A., . . . Shendo, K. (2003). Jemez Pueblo: Built and social–cultural environments and health within a rural American Indian community in the Southwest. American Journal of Public Health, 93, 1517–1518. http://dx.doi.org/10.2105/AJPH.93.9.1517 Walters, K., Stately, A., Evans-Campbell, T., Simoni, J., Duran, B., Shultz, K., . . . Guerreo, D. (2008). “Indigenist” collaborative research efforts in Native American communities. In A. Stiffman (Ed.), The nitty-gritty of managing field research (pp. 146–173). Oxford, England: University Press. Whitbeck, L. B. (2001). Bii-zin-da-de-da: The listening to one another prevention program. Paper presented at the Second National Conference on Drug Abuse Prevention Research: A Progress Update, Washington, DC. Whitbeck, L. B., Hoyt, D. R., Stubben, J. D., & LaFromboise, T. (2001). Traditional culture and academic success among American Indian children in the upper Midwest. Journal of American Indian Education, 40, 48–60.

220       belone et al.

IV Treatment and Interventions

10 A CULTURALLY INFORMED APPROACH TO AMERICAN INDIAN/ ALASKA NATIVE YOUTH SUICIDE PREVENTION TERESA D. LaFROMBOISE AND SAIMA S. MALIK

According to scholarly accounts, suicide has disproportionately affected the American Indian/Alaska Native (AI/AN) population for well over half a century (Havighurst, 1971). The Indian Health Service (IHS) has reported that the AI/AN youth suicide rate is 3.5 times higher for those 15 to 24 years old than the national average for this age group (IHS, 2002). Suicide accounts for 26.5% of deaths among AI/ANs 15 to 19 years old. Suicide also accounts for 13.5% of deaths among AI/ANs 10 to 14 years old, nearly double the rate for all races in that age group (Centers for Disease Control and Prevention, 2007). Since the introduction of the Garrett Lee Smith Act of 2004, various forms of evidence-based intervention (EBI) efforts within AI/AN communities have attempted to combat this devastating epidemic (Goldston et al., 2010). Though well-intentioned and scientifically sound, these interventions have historically been culturally disconnected and therefore have failed to bring about any real and long-lasting change (Wexler et al., 2015; Wexler & http://dx.doi.org/10.1037/14940-011 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

223

Gone, 2012). In addition, in the exploration of potential solutions to the problem of AI/AN youth suicide, there has been an unfortunate tendency to focus primarily on risk factors associated with AI/AN experiences and to overlook valuable protective factors embedded within AI/AN cultural practices. We begin this chapter with a discussion of the unique risk and protective factors of the subgroup of AI/AN adolescents who struggle with being suicidal. We assert that to be successful, the content of AI/AN suicide prevention interventions should be based first and foremost on comprehensive knowledge of culturally unique factors. Second, intervention developers should be responsive to the needs of the AI/AN community and actively engage its members in intervention design, evaluation, and implementation. We present the development and evaluation of the American Indian Life Skills (AILS) curriculum as an account of one suicide prevention intervention that acknowledges the importance of both of these tenets. The implementation and subsequent evaluation of the AILS reflect the particular challenges associated with addressing the evidence-based practice mandate of federally funded programs in AI/AN communities (Walker & Bigelow, 2011). Although the gold standard for determining evidence-based practice involves a strict experimental approach with random assignment of participants to treatment groups, doing so becomes challenging in cases where a community is averse to a randomization scheme. Consideration of these challenges is particularly important in light of the current dialogue surrounding tensions between conventional intervention evaluation protocol and maintaining respect for AI/AN life ways. Information within this chapter is shared with a caveat, out of respect for the complexities of youth suicide within a heterogeneous population. There are 2.9 million AI/ANs representing 565 federally recognized tribes. When those who affiliate with more than one racial group are included, this estimate expands to 5.2 million people (Humes, Jones, & Ramirez, 2011). Each tribal community has its own unique history, practices, cultural identification, and economic resources. For instance, recent writings comparing suicidal behavior among AI/AN youth have found differential rates and correlates of suicide based on residence (Freedenthal & Stiffman, 2004; Pettingell et al., 2008). There exists a widening gap in resources and social status of individuals between different AI/AN communities (Harvard Project on American Indian Economic Development, 2008). These features challenge the validity of generalizing suicide patterns across AI/AN groups. Adolescent suicide often occurs in contexts such as family conflict, academic and disciplinary difficulties, and disruption in peer relationships. Unlike trends in the U.S. mainstream population, in which the risk of suicide increases with age, death by suicide is particularly salient for AI/ANs during adolescence 224       lafromboise and malik

into the mid-20s, at which point risk begins to decline. Although the reasons for such high rates occurring during this developmental stage are unclear, speculation regarding this phenomenon abound. For instance, AI/AN adolescents may feel that they have been disappointed, hurt, or abandoned by loved ones dealing with a host of hardships. The future may seem bleak for those who anticipate disrespect and discrimination. There may be an additional burden of role captivity, the perceived inability to detach from family and community turmoil. Alternatively, those who have access to cultural resources such as community support, high-functioning family members, and rituals that increase a sense of belonging may be able to assist others who are undergoing distress. Reviewing studies of AI/AN youth suicide reveals a multitude of individual-level risk factors including depression, hopelessness, strained interpersonal communication, social isolation, and school difficulties (HowardPitney, LaFromboise, Basil, September, & Johnson, 1992; Wexler, 2006). Mullany et al. (2009) found that interpersonal factors such as family or partner conflict precipitated suicide attempts among Apache youth. Experiences such as family members having attempted suicide, being a victim of physical or sexual abuse, being placed in special education classes, being involved in a gang, and having guns available have also been linked to an increased likelihood of AI/AN adolescent suicide attempts (Borowsky, Resnick, Ireland, & Blum, 1999; Medoff, 2007). In addition, age, gender, negative life events, perceived discrimination, and drug use have been associated with the likelihood of thinking about suicide (Yoder, Whitbeck, Hoyt, & LaFromboise, 2006). Alcohol or drug consumption along with depression has also been shown to strongly predict suicidal ideation and suicidal behavior (LaFromboise, Medoff, Lee, & Harris, 2007). CULTURALLY UNIQUE RISK FACTORS FOR SUICIDE Although many of the risk factors noted above may be present in the lives of other ethnic minority adolescents, AI/AN youth currently have the highest suicide rates (Goldston et al., 2008), and the prevalence and presentation of these factors in the lives of AI/ANs are distinct. We next consider the impact of historical trauma, acculturation stress, and community violence on individual adaptation. Given the robust nature of substance abuse as either a unique or co-occurring risk factor for suicide, as well as its high prevalence among AI/ANs, we include substance abuse in our discussion. Historical Trauma Historical trauma has been described as trauma resulting from successive, compounding events perpetrated on a community over generations to eliminate a culturally informed approach     

225

the cultural practices and cultural identity of its members (Brave Heart, 2003). These events have included reservation internment, annihilation of ceremonial practices, and “loss of land to which American Indians were spiritually and emotionally tied” (Brave Heart & DeBruyn, 1998, p. 361). Walters et al. (2011) asserted that stress associated with historical loss may not only have a stifling effect on development but also may negatively impact health over generations. An assimilationist policy most detrimental to cultural identity and family functioning was the forced removal of AI children from their families to boarding schools (Meriam, 1928). According to Evans-Campbell (2008), at least four generations of survivors may have transmitted the trauma they experienced in boarding schools to their children, grandchildren, extended family members, and nondescendants. Historical trauma may explain a mechanism by which AI/ANs are at greater risk of experiencing psychological turbulence. Whitbeck, Walls, Johnson, Morrisseau, and McDougall (2009) hypothesized that growing up in a context of constant reminders of loss associated with colonization contributes to adolescent depression and demoralization. They found that daily thoughts of historical loss by AI adolescents (11 to 13 years old) were similar to, and sometimes exceeded, those of their adult caretakers. Walls, Chapple, and Johnson (2007) found that several stressors, including coercive parenting and caretaker rejection, were related to suicidal behavior among AI adolescents. More recently, in a study of trauma and suicidal behavior among Canadian Aboriginal people, Elias et al. (2012) found that a youth’s personal history of negative boarding school experiences was associated with suicidal thoughts. Acculturation Stress Defined as a systemic overload associated with navigating differences between two or more cultures, acculturation stress is most often manifested within the AI/AN population as personal resistance or struggle to maintain AI/AN cultural integrity when functioning in mainstream society. Chadwick and Strauss (1975) found that many AI/AN people living in Seattle since the 1950s were as culturally engaged as those who had recently migrated from their reservations to the city. The impact of resistance to acculturation on AI/AN mental health continues to be evident today (Waldram, 2004). AI/AN youth who leave their reservations often face increasing challenges in the urban environment. They feel the tension between remaining in a culturally familiar and supportive environment and venturing into an unfamiliar milieu that privileges individual achievement. For example, Dickerson and Johnson (2011) found that AI/AN urban youth experience stress associated with less frequent opportunities to retain traditional knowledge and engage in 226       lafromboise and malik

traditional activities. Furthermore, they are burdened by challenges associated with negative stereotypes, racism, violence, and gang activity. The relationship of acculturation and suicide was established in studies of suicide patterns in New Mexico by Van Winkle and May (1986, 1993) and was further buttressed by Lester (1999). Both studies found that suicide rates were positively associated with acculturation stress and negatively associated with traditional integration in 18 AI/AN tribes. Community Violence AI/AN youth are 2.5 times more likely than other youth to encounter trauma (National Center for Children in Poverty, 2007). It important to note that although AI/ANs are twice as likely as their non-AI/AN counterparts to experience violent victimization, 70% of the time that violence is inflicted upon them by non-AI/AN perpetrators (Greenfield & Smith, 1999). It is notable that AI/ANs suffer from high rates of posttraumatic stress disorder (PTSD), depression, anxiety symptoms, substance abuse, antisocial behavior, social withdrawal, and academic problems (Boyd-Ball, Manson, Noonan, & Beals, 2006; Goodkind, LaNoue, & Milford, 2010). Childhood exposure to interparental violence has been found to result in moderate to clinical levels of conduct problems, emotional problems, and lower levels of social functioning among youth (Fantuzzo et al., 1991). Although precise estimates of AI/AN child sexual abuse do not exist, risk factors associated with this form of abuse are greater among AI/ANs than any other racial group (L. EchoHawk, 2001). Untreated victims have a higher risk of continuing this cycle of abuse as adults. Although little is known about the prevalence of self-cutting, burning, or hitting within this subgroup, a 2011 study of Apache youth indicated greater involvement in nonsuicidal self-injury as compared with youth in the general population. Precipitants of self-injury included conflict with one’s family or partner, peer pressure, and mimicking others’ self-injurious behavior (Cwik et al., 2011). Substance Abuse Adolescents who are exposed to adverse childhood experiences often undergo severe emotional problems and may be more inclined to engage in substance use (Koss et al., 2003). AI/AN youth are an especially high-risk group for the propensity for binge drinking and problems associated with alcohol consumption (Hawkins, Cummins, & Marlatt, 2004; Walls, Whitbeck, Hoyt, & Johnson, 2007). They are more likely than their non-AI/AN peers to report lifetime alcohol use, indicating earlier and more frequent and a culturally informed approach     

227

problematic alcohol use in this subgroup (Beauvais, Jumper-Thurman, Helm, Plested & Burnside, 2004). Because the abuse of alcohol is so deeply intertwined with the abuse of other drugs, it is nearly impossible to examine the effects of each separately. According to Brave Heart, Chase, Elkins, and Altschul (2011), AI/ANs are increasingly abusing inhalants, methamphetamines, and IV drugs. Data from the Monitoring the Future annual survey (Wallace et al., 2003) indicate a high prevalence of inhalant abuse among AI/AN youth (9.4% compared with 6.6% in all other ethnic groups). In a recent study on binge substance use, Apache youth reported engaging in this behavior to avoid problems or reduce negative feelings (Tingey et al., 2012). In light of their salience with the AI/AN population, the risk factors for suicide discussed thus far are essential considerations when testing models for the etiology of suicide and multiple, co-occurring conditions (e.g., bullying, substance use and other forms of self-injury). Furthermore, findings from research on AI/AN youth regarding the correlates of engagement in life-threatening behavior should guide the development and refinement of suicide prevention efforts. PSYCHOLOGICAL FUNCTIONING AI/AN youth who struggle with suicide also have unique experiences according to the family and community to which they belong, their cultural orientation, and their psychological functioning. Unique risk factors for suicide may be compounded by the presence of mental disorders such as depressive disorder, bipolar disorder, schizophrenia, and personality disorder (Goldsmith, Pellmar, Kleinman, & Bunney, 2002). According to Beals et al. (1997), 29.4% of AI adolescents 14 to 16 years of age from a Northern Plains reservation met the diagnostic criteria for at least one mental disorder. However, Whitbeck, Yu, Johnson, Hoyt, and Walls (2008) found that 44.8% of early adolescents, 13 to 15 years of age, from a tribe located in the Upper Midwest had a lifetime psychological disorder and 26.6% had a 12-month disorder. These variable rates of mental disorders among AI/AN youth complicate one’s understanding of the impact of psychological functioning on suicide. One diagnostic category that is quite common among AI/AN youth is PTSD. This disorder was more common in the two reservation populations sampled in the AI-SUPERPFP study1 than in other populations (Beals et al., 1The AI-SUPERPFP was the first comprehensive assessment funded by the National Institute of Mental Health of the prevalence of alcohol, drug, and mental health problems in two distinct and heavily populated AI/AN groups.

228       lafromboise and malik

2005). Rates of full PTSD as high as 10% and rates of subthreshold PTSD as high as 14% were noted in a clinical sample of AI/ANs by Deters, Novins, Fickenscher, and Beals (2006). CULTURALLY UNIQUE PROTECTIVE FACTORS FOR SUICIDE Despite Borowsky et al.’s (1999) original discussion of protective factors in suicide attempts among AI/AN youth, scant attention has been paid to the factors that buffer against AI/AN youth suicide. Their large-scale study found that emotional health, discussing problems with friends or family, and connectedness to family protected against suicide attempts. This research also identified that increasing protective factors was more effective than decreasing risk factors in reducing the probability of a suicide attempt. A population-based study of AI/AN and non-AI/AN youth attending schools off the reservation found a buffering effect of certain protective factors against suicide. These factors included self-esteem, self-efficacy, positive mood/emotional health, family support/connectedness, and parental prosocial norms (Mackin, Perkins, & Furrer, 2012). Not only were AI/ANs at a higher risk for suicide attempts than were non-AI/ANs, but they also had a higher threshold of risk factors marking a suicide attempt. These protective factors had a greater impact for individuals who exhibited a higher level of risk for suicide than for those who exhibited a lower level of risk. Another key protective factor is enculturation. Enculturation refers to the extent to which individuals are grounded in their traditional beliefs, engage in cultural practices, and examine and internalize their cultural identity. Wolsko, Lardon, Mohatt, and Orr (2007) found that Yup’ik people (Alaska) who identified more with their traditional way of life experienced greater happiness, more adherence to spiritual ways of coping, and less frequent drug and alcohol use to manage stress. In an investigation of suicidal ideation among AI/AN youth from a tribe in the Upper Midwest, Yoder et al. (2006) found that enculturation was a strong predictor of suicidal thoughts and plans such that individuals with higher levels of enculturation were less likely to suffer from suicidal ideation. LaFromboise, Medoff, Lee, and Harris’s study (2007) of AI youth living on a Northern Plains reservation also noted the protective role of enculturation and school belonging in deterring suicidal ideation. In addition, LaFromboise, Albright, and Harris (2010) found associations between stronger ethnic/cultural identity and lower levels of hopelessness, a psychological outcome often linked to depression and suicide. Fluency in one’s tribal language, an essential feature of enculturation, was not measured in this study because of the complex influence of English and tribal language a culturally informed approach     

229

use on ethnic identity among AI/AN youth (Moran, Fleming, Somervell, & Manson, 1999). However, the work of Chandler and colleagues in examining youth suicide within First Nation communities in British Columbia has considered conversational knowledge of one’s Aboriginal language along with other cultural continuity factors (Chandler & Proulx, 2006). Hallett, Chandler, and Lalonde (2007) noted that conversational knowledge of language was associated with lower rates of youth suicide above and beyond the presence of other cultural continuity factors (e.g., evidence of a band—the Canadian equivalent of a U.S. tribe or nation—having taken back from government agencies certain rights to self-government). Although this portfolio of studies reveals a robust relationship between enculturation and psychological well-being, the extent to which cultural identity has a direct impact on suicide remains unclear. However, these findings bolster the need for AI/AN prevention intervention programs to draw on AI/AN cultures and traditions. SCHOOL-BASED SUICIDE PREVENTION EFFORTS The risk of youth suicide is compounded when resources offering suicide intervention within communities are inadequate. The IHS receives woefully inadequate funding, particularly for mental health services. In many communities, the infrastructure to support mental health is weak, if present at all. Furthermore, when resources are available, AI/AN youth often avoid seeking psychological services because of internal factors such as self-reliance, embarrassment, lack of problem recognition, and a belief that nobody would help them (Freedenthal & Stiffman, 2007). There is reluctance on the part of many AI/AN families to seek help or engage in therapy. Instead, parents may enlist the help of a family member or close friend to talk with a suicidal child. We believe that schools can be an effective venue for the delivery of suicide prevention with AI/AN youth, especially when used in combination with other prevention strategies such as gatekeeper training, extensive community outreach, and social–emotional learning programs. May, Serna, Hurt, and DeBruyn (2005) validated the success of such a combined approach with an AI reservation community in the Southwest. School-based suicide prevention programs typically provide education and awareness about suicide and encourage a positive attitude toward seeking necessary help. School-based suicide prevention programs are generally brief in duration and some incorporate screening for suicide and mental health needs as part of the prevention protocol (Steele & Doey, 2007). Whereas schoolbased suicide prevention programs have been found to increase knowledge 230       lafromboise and malik

and improve attitudes concerning mental health and suicide (Breton et al., 2002), too often the content is disconnected from the community they serve (Muehlenkamp, Marrone, Gray, & Brown, 2009). This disconnect is apparent in the identification of the following three EBI prevention programs implemented in schools that have been found to reduce suicidal behaviors: the Good Behavior Game (GBG), Coping and Support Training (CAST), and Sources of Strength. GBG is a universal primary prevention program targeting elementary school age students (Barrish, Saunders, & Wolf, 1969), with the goal of less aggressive and disruptive behavior among children. GBG has demonstrated long-term effects on decreased impulsive or disruptive behavior, substance use, and lower rates of suicidal ideation and attempts among participants (Kellam et al., 2008). However, the program includes a component of competition and hence may be incongruent with the values of a number of tribal groups. CAST, a program adapted from Reconnecting Youth, utilizes a skills training approach and targets high school students (ages 14–19) who have been identified through an initial suicide assessment interview screen. The goals of CAST include increased mood management, better school performance, and decreased drug involvement. Participants have demonstrated increased problem-solving skills, increase perceived family support and selfcontrol, decreased symptoms of depression and hopelessness, and a significant reduction in anger and suicide risk behavior (Eggert, Thompson, Randell, & Pike, 2002; Thompson, Eggert, Randell, & Pike, 2001). However, similar to GBG, the intervention does not explicitly address cultural factors that may influence suicidal ideation. Sources of Strength is a universal, strength-based comprehensive wellness program that was initially developed for tribal and rural settings and later expanded for the general population. Peer leaders are trained to respond to students who display risk factors for suicide and direct them to a trusting adult. Participants in the program reported reduced suicide attempts and increased knowledge about suicide whereas peer leaders reported increased adaptability in attitudes toward suicide and other mental illnesses and enhanced ability to refer a suicidal friend to a trusted adult (Aseltine & DeMartino, 2004; Aseltine, James, Schilling, & Glanovsky, 2007). Although evaluations reveal decreases in students’ suicidal behavior, Sources of Strength does not refer specifically to tribal culture and no outcome studies of the program have been conducted with an adequate AI/AN population. Only 1% of the sample in the outcome evaluation of this intervention was reported to be of AI/AN heritage (Wyman et al., 2010). To our knowledge these three interventions are the only programs that have met the criteria for effectiveness in decreasing suicide risk when following randomized clinical trial procedures. Although these interventions have been received by schools serving AI/ANs, all three have yet to target an adequate sample of AI/ANs in their a culturally informed approach     

231

evaluation efforts. Thus they have yet to demonstrate effectiveness in reducing suicide risk with American Indian youth or widespread appeal among AI/AN communities, possibly in part because of their inattention to AI/AN culturally specific suicide risk and protective factors. Furthermore, many AI/ AN communities prefer culturally sensitive programs over ones apparently packaged for mainstream consumption. In the following section, we share information about the evolution and evaluation of a community-driven, school-based suicide prevention approach initiated by leaders of the Zuni tribe who believe that the foundation of healthy individuals, families, and communities relies on shared valuing of life. A CASE EXAMPLE: ZUNI LIFE SKILLS EVALUATION Developing the Intervention The Zuni Life Skills Development Curriculum (ZLS) emerged as a response to the specific demands of an AI/AN community and actively engaged community stakeholders during the phases of development, implementation, and evaluation. Intervention content was based on a comprehensive knowledge of both cognitive and behavioral theory as it pertains to adolescence as well as the specific AI/AN risk and protective factors highlighted in this chapter. Evaluation results support the benefits of this intervention with participants. Achieving EBI model program status based on criteria for evidencebased practice in this experience included the following challenges: a lack of understanding on the part of local behavioral health staff regarding the need for school-based prevention efforts, the existence of a relatively small and unique sample population, an initial rejection by community leaders of randomized assignment to treatment, a dearth of qualified AI/AN interventionists, and off-site supervision of the intervention. Although the positive student outcomes reported here reflect the importance of a culturally informed approach to suicide prevention, challenges faced along the way provide insight into the difficulty of implementing evidence-based practice with AI/AN communities. This struggle between evidence-based practice and the healing value of indigenous culture has been hotly contested within the field of behavioral health in recent years (H. Echo-Hawk, 2011; Smith, 1999; Whitbeck, Walls, & Welch, 2012). Between 1980 and 1987, there were 13 deaths by suicide among youth in the Zuni pueblo, a reservation made up of 9,000 members located in the high desert of northwest New Mexico. At the request of tribal leaders, Teresa LaFromboise was asked to assist in addressing this tragic problem in the Zuni Public School District. Over a 3-year period, she worked in partnership with 232       lafromboise and malik

community and university educators to develop and evaluate the ZLS, an intervention consisting of life skills training and education about youth suicide prevention (LaFromboise, 1991). During the initial phase of development, community members participated in asset identification and issue selection wherein they contemplated the potential causes of youth depression and suicide in their community and recommended coping strategies they wished to encourage throughout the intervention (LaFromboise & Howard-Pitney, 1993). Structural issues negotiated during the development phase included which grade levels would receive the intervention, who would deliver the intervention, and what the preferred format of delivery would be. It was decided that the ZLS would include a thoughtful balance of Zuni socialization methods and psychological approaches. Cultural values of respect, honesty, wisdom, gratitude, and bravery were carefully considered for integration into the new skills curriculum. Opportunities for the inclusion of spiritual teachings delivered by respected Zuni community members were infused throughout the intervention to support the positive involvement of Zuni cultural teachings and practices in the lives of Zuni youth, their families, and their community (LaFromboise & Lewis, 2008). Keeping Zuni adolescent daily experiences in mind, the curriculum included 44 lessons, 10 of which covered suicide crisis intervention training (e.g., how to listen to and seek help for a suicidal friend); 30 sought to increase protective factors such as positive mood and emotional health through skills training to learn to engage in peer-to-peer support, cope with stress, manage anger and depression, and solve problems or resolve conflicts; and four provided background understanding about self-destructive behavior. The psychological foundations for the intervention stemmed from social–cognitive theory (Bandura, 1986) and cognitive–behavioral theory (Beck, 1976). From a social–cognitive perspective, suicidal behavior is attributed to direct learning or modeling influences (e.g., prevalent suicidal behavior within the community) in conjunction with certain environmental influences (e.g., lifetime exposure to substance abuse) and individual characteristics (e.g., depression and hopelessness) that mediate decisions related to risk behavior (LaFromboise & Rowe, 1983). From a cognitive–behavioral perspective, adolescents are presumed to be at risk for suicide, at least in part, when they are predisposed to having depressive and/or irrational thoughts. Teaching youth new coping mechanisms and strengthening their repertoire of coping responses provides them with some measure of protection against the tendency to avoidant coping (e.g., self-isolation, substance abuse) and other forms of risk behavior. Select material from the Adolescent Coping With Depression Course (Clarke, Lewinsohn, & Hops, 1990) was culturally adapted for inclusion in the intervention. a culturally informed approach     

233

Oftentimes it may be difficult to find an interventionist who is both intimately aware of the cultural nuances of the local context and well-versed in the specific clinical methodology required for delivery. Although careful selection and training of community members as interventionists can help to ease this challenge somewhat, many AI/AN communities are lacking in relevant resources, requiring the curriculum to be taught either by outsiders or by those from within the community who may not have expertise in social–emotional development. Because of a lack of Zuni teachers at the school at the time, finding an interventionist with expert knowledge about Zuni culture posed a challenge. Soon after the ZLS was launched, Zuni IHS mental health technicians offered to colead the intervention with non-AI/ AN teachers/interventionists, providing an ideal solution to this dilemma. Unfortunately, a lack of qualified interventionists remains a challenge in many AI/AN prevention efforts, potentially curtailing the full impact of the program. Evaluation of a school-based suicide prevention intervention for AI/AN communities poses several fundamental difficulties. Zuni High School is a relatively small school with a small population. At the time of this evaluation, community members did not want the researchers to follow procedures for randomization at the individual level. The close connectedness within the Zuni community raised the possibility of control group contamination. Although community refusal to allow randomization rendered a randomized controlled trial impossible, the evaluation, following a quasi-experimental design with two conditions (intervention and no-intervention), found that participation in ZLS reduced suicidal ideation and hopelessness, increased problem-solving ability, increased confidence to manage anger, and increased suicide prevention skills among participants (LaFromboise & Howard-Pitney, 1995). Youth suicide in Zuni ceased almost immediately after the curriculum was implemented in 1991 (Woodard, 2012). As a testament to its wide-scale acceptability and ecological validity, ZLS is a required course in Zuni High School even today. Adapting the Intervention to a New Context In 1990, after completing the ZLS evaluation, Teresa LaFromboise was invited by the Cherokee Nation to implement and evaluate the intervention at Sequoyah High School, a boarding school in Tahlequah, Oklahoma, which served AI/AN students from more than 20 different tribes across Indian Country at that time. To ensure relevance and maximize student engagement in and benefits from the intervention, cultural nuances of the ZLS that were specific to the Zuni tribe were substituted with examples of salient events and cultural teachings from the tribes represented by this 234       lafromboise and malik

student population. Although the curriculum was culturally adapted for the Sequoyah High School context, the core skills and content of the ZLS remained unchanged. Simultaneously, a longitudinal study of mental health among students at Sequoyah High School was being conducted by the National Center for American Indian and Alaska Native Mental Health Research. Because this longitudinal study included assessment of reports of suicidal behavior, students participating in the intervention could be monitored throughout their Sequoyah High School career. Results from this natural experiment indicated a reduction in a 20-year suicide and suicide attempt rate with no deaths by suicide noted since the beginning of implementation of the intervention in 1990 (see the National Academy of Sciences/Institute of Medicine Report edited by Goldsmith et al., 2002). A question that emerged during this particular implementation and evaluation opportunity was that of off-site supervision. We were unsure whether interventionists could balance their role as classroom teacher with their role as prevention intervention facilitator. To tackle this problem, an on-site teaching staff member was appointed to facilitate weekly meetings with the teachers/interventionists to provide technical assistance on intervention content and implementation issues. In this application, all teachers/ interventionists were AI. Their collective insight and understanding of the students and their cultural backgrounds were invaluable. Although off-site supervision from the intervention developer was a challenge at that time, the availability of various forms of technology and media-driven delivery systems has eased this challenge and rendered off-site supervision of the AILS more effective. After the yearlong evaluation was complete, a more tribally heterogeneous version of the ZLS was published as the American Indian Life Skills Development Curriculum (AILS; LaFromboise, 1996). Rather than being a one-size-fits-all intervention, the AILS encourages interventionists to incorporate traditional and contemporary worldviews of the tribes and communities they work with into the curriculum without compromising the core psychological components of the program or displacing the skills training outlined in the manual. Community members interested in culturally adapting the AILS are cautioned during interventionist trainings that spontaneous adaptation of an EBI may alter it to a point that it is rendered unscientific. AILS intervention trainings also include didactic information on strategies covered in the AILS and culturally unique AI/AN risk factors for suicide. Information on how to use the manual and demonstrations to teach core social skills training strategies are shared. Additional topics covered include confidentiality, classroom management, school policy and procedures related to disclosure of suicidal intent or behavior, and the process a culturally informed approach     

235

of referral for treatment. To date, AILS interventionist trainings have been conducted with community members from more than 100 reservations across Indian country.2 An independent evaluation of the AILS was conducted in 2011 by direct service providers working with AI/AN students in New Mexico and involved 90 participants 14 to 18 years of age. This evaluation of an abbreviated version of the AILS employed a quasi-experimental design with an intervention condition. The intervention consisted of 10 AILS lessons purposefully selected to meet the needs of the Navajo youth who participated in the intervention and its evaluation. Pretest–posttest comparisons supported the positive effect of treatment (Salvatore, 2011). The presence of independent evaluations such as this one indicates an effort toward continuous quality improvement of prevention interventions through evaluation of school- and community-based programs (see also the evaluation of AILS in Cozad, 2008). Although only 10 lessons were selected for this study, each of these lessons adhered to the original lesson content in the AILS manual. Modification of AILS dosage may be viewed as a reflection of programmatic restrictions faced by service providers, something not uncommon in low-resourced, school-based work. Refining the AILS for Early Adolescents The staggering rates of suicide among younger children in the AI/AN population led to a growing urgency to extend the benefits of the curriculum to younger adolescents as well. In response to this need, Teresa LaFromboise decided to conduct a pilot trial to evaluate the feasibility and effects of the AILS with a younger student population. The results from these studies have guided modifications of the curriculum toward a version of the AILS that is aimed at addressing the issues of AI/AN early adolescents. In 2007, an evaluation of the original AILS was conducted with 122 middle school students (11–15 years of age) living on a Northern Plains reservation. During the evaluation, the curriculum was offered twice a week, in 35-minute sessions, over 6 weeks. Scores on scales of hopelessness (Kazdin, Rodgers, & Colbus, 1986) and suicidal ideation (Reynolds, 1988) were determined among students who were randomized to the AILS condition, a comparative treatment condition (Reconnecting Youth; Eggert & Nicholas, 2004), or a “learning period as usual” delayed intervention condition. This evaluation found statistically significant reductions in hopelessness and suicidal ideation 2 Locations of trainings include: Anchorage and Bethel, Alaska; Bylas and Phoenix, Arizona; Arcata, California; Detroit, Michigan; Leech Lake and St. Cloud, Minnesota; Wolf Point, Montana; Belcourt and Fort Totten, North Dakota; Macy, Nebraska; Santa Fe, Shiprock, and Zuni, New Mexico; Gardenville and Pyramid Lake, Nevada; Oklahoma City, Oklahoma; Pine Ridge, Rosebud, and Sisseton, South Dakota; and Spokane, Washington.

236       lafromboise and malik

for students in the comparative treatment group. These results, along with classroom observations and feedback from experts in the field of AI/AN child trauma and special education, led to extensive modifications to the AILS curriculum. These modifications included adjustment of content, reduction in number of lessons from 44 to 30, and restructuring of lesson activities, making the intervention more age-appropriate for students and more user-friendly for teachers/interventionists. In 2011, Teresa LaFromboise was invited back to the Pueblo of Zuni to implement the AILS at the middle school level and to conduct an evaluation of the AILS modified version. Participants in this study included 141 students attending Zuni Middle School who were randomized to receive the modified AILS immediately or 6 weeks later as a delayed treatment. Findings indicated significantly positive outcomes for those who had engaged in the AILS intervention on measures of life skills efficacy, depression management, stress management, ability to enlist community support, and ability to enlist social resources as compared with the control group (LaFromboise & Malik, 2012). In this study we were fortunate to find one AI/AN interventionist (the other two interventionists were non-AI/AN educators). Statistical comparisons of student outcomes found no interventionist effect, which we attribute to the strength of the curriculum itself. We believe these findings support continued implementation of the early adolescent version of the AILS with evaluation. CONCLUSION We began this chapter with a discussion of the risk and protective factors associated with suicide among AI/AN youth. We reviewed a number of studies that highlighted the importance of considering these factors in working with AI/AN adolescents and asserted that suicide prevention interventions should be based on comprehensive knowledge of culturally unique factors as well as be responsive to the needs of diverse AI/AN communities. We reviewed three evidence-based prevention programs found to reduce suicide risk that are also used by schools serving AI/AN youth. We then presented the development and evaluation of the AILS as a case example of one suicide prevention intervention that was based on an understanding of unique AI/AN risk and protective factors, sensitive to the needs of the community, and actively partnered with community members at every phase of development. In our discussion of the AILS, we highlighted challenges inherent in evaluating interventions using currently prescribed highly rigorous evaluation techniques with AI/AN communities and provided examples of the practical solutions employed in this case. The specific demands of scientific inquiry may sometimes be at odds with the preservation of traditional ways of life. The requirements of funding a culturally informed approach     

237

agencies regarding evidence-based prevention programs oftentimes overlook the value of such traditional knowledge. This system leaves many AI/AN communities at a disadvantage and may force them to adopt intervention implementation and evaluation styles that are disrespectful of AI/AN protocol and at odds with tribal sovereignty. The design of appropriate measures of intervention outcomes requires a balance between culturally informed AI/AN perspectives of effective practice and conventional ideals of model interventions based on scientific evidence. We hope that examples of our experience developing and evaluating the AILS may contribute toward a more realistic understanding of this form of challenging but very rewarding service. REFERENCES Aseltine, R. H., Jr., & DeMartino, R. (2004). An outcome evaluation of the SOS suicide prevention program. American Journal of Public Health, 94, 446–451. http:// dx.doi.org/10.2105/AJPH.94.3.446 Aseltine, R. H., Jr., James, A., Schilling, E. A., & Glanovsky, J. (2007). Evaluating the SOS suicide prevention program: A replication and extension. BMC Public Health, 7, 161. http://dx.doi.org/10.1186/1471-2458-7-161 Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. Barrish, H. H., Saunders, M., & Wolf, M. M. (1969). Good behavior game: Effects of individual contingencies for group consequences on disruptive behavior in a classroom. Journal of Applied Behavior Analysis, 2, 119–124. http://dx.doi. org/10.1901/jaba.1969.2-119 Beals, J., Manson, S. M., Whitesell, N. R., Spicer, P., Novins, D. K., Mitchell, C. M., & AI-SUPERPFP Team. (2005). Prevalence of DSM–IV disorders and attendant help-seeking in two American Indian reservation populations. Archives of General Psychiatry, 62, 99–108. http://dx.doi.org/10.1001/archpsyc.62.1.99 Beals, J., Piasecki, J., Nelson, S., Jones, M., Keane, E., Dauphinais, P., . . . Manson, S. M. (1997). Psychiatric disorder among American Indian adolescents: Prevalence in Northern Plains youth. Journal of the American Academy of Child & Adolescent Psychiatry, 36, 1252–1259. http://dx.doi.org/10.1097/00004583199709000-00018 Beauvais, F., Jumper-Thurman, P., Helm, H., Plested, B., & Burnside, M. (2004). Surveillance of drug use among American Indian adolescents: Patterns over 25 years. Journal of Adolescent Health, 34, 493–500. http://dx.doi.org/10.1016/ S1054-139X(03)00340-9 Beck, A. T. (1976). Cognitive therapy and the emotional disorders. Oxford, England: International Universities Press.

238       lafromboise and malik

Borowsky, I. W., Resnick, M. D., Ireland, M., & Blum, R. W. (1999). Suicide attempts among American Indian and Alaska Native youth: Risk and protective factors. Archives of Pediatrics & Adolescent Medicine, 153, 573–580. http://dx.doi.org/ 10.1001/archpedi.153.6.573 Boyd-Ball, A. J., Manson, S. M., Noonan, C., & Beals, J. (2006). Traumatic events and alcohol use disorders among American Indian adolescents and young adults. Journal of Traumatic Stress, 19, 937–947. http://dx.doi.org/10.1002/jts.20176 Brave Heart, M. Y. (2003). The historical trauma response among natives and its relationship with substance abuse: A Lakota illustration. Journal of Psychoactive Drugs, 35, 7–13. http://dx.doi.org/10.1080/02791072.2003.10399988 Brave Heart, M. Y., Chase, J., Elkins, J., & Altschul, D. B. (2011). Historical trauma among Indigenous peoples of the Americas: Concepts, research, and clinical considerations. Journal of Psychoactive Drugs, 43, 282–290. http://dx.doi.org/ 10.1080/02791072.2011.628913 Brave Heart, M. Y., & DeBruyn, L. M. (1998). The American Indian Holocaust: Healing historical unresolved grief. American Indian and Alaska Native Mental Health Research, 8, 56–78. Breton, J. J., Boyer, R., Bilodeau, H., Raymond, S., Joubert, N., & Nantel, M. A. (2002). Is evaluative research on youth suicide programs theory-driven? The Canadian experience. Suicide and Life-Threatening Behavior, 32, 176–190. Centers for Disease Control and Prevention. (2007). Deaths: Leading causes for 2004, National Vital Statistics Reports, 56(5). Available at http://www.cdc.gov/men/ lcod/2004/index.htm Chadwick, B. A., & Strauss, J. H. (1975). The assimilation of American Indians into urban society: The Seattle case. Human Organization, 34, 359–369. Chandler, M., & Proulx, T. (2006). Changing selves in changing worlds: Youth suicide on the fault-lines of colliding cultures. Archives of Suicide Research, 10, 125–140. http://dx.doi.org/10.1080/13811110600556707 Clarke, G. N., Lewinsohn, P. M., & Hops, H. (1990). Instructor’s manual for the Adolescent Coping With Depression Course. Eugene, OR: Castalia Press. Cozad, A. R. (2008). Saving the next generation: A “Life skills” curriculum for rural Oklahoma youth. Unpublished manuscript. Indian Health Service Injury Prevention Fellowship. Lawton, OK: Indian Health Service. Cwik, M. F., Barlow, A., Tingey, L., Larzelere-Hinton, F., Goklish, N., & Walkup, J. T. (2011). Nonsuicidal self-injury in an American Indian reservation community: Results from the White Mountain Apache surveillance system, 2007–2008. Journal of the American Academy of Child & Adolescent Psychiatry, 50, 860–869. http://dx.doi.org/10.1016/j.jaac.2011.06.007 Deters, P. B., Novins, D. K., Fickenscher, A., & Beals, J. (2006). Trauma and posttraumatic stress disorder symptomatology: Patterns among American Indian adolescents in substance abuse treatment. American Journal of Orthopsychiatry, 76, 335–345. http://dx.doi.org/10.1037/0002-9432.76.3.335 a culturally informed approach     

239

Dickerson, D. L., & Johnson, C. L. (2011). Design of a behavioral health program for urban American Indian/Alaska Native youths: A community informed approach. Journal of Psychoactive Drugs, 43, 337–342. http://dx.doi.org/10.1080/ 02791072.2011.629152 Echo-Hawk, H. (2011). Indigenous communities and evidence building. Journal of Psychoactive Drugs, 43, 269–275. http://dx.doi.org/10.1080/02791072.2011.628920 EchoHawk, L. (2001). Child sexual abuse in Indian Country: Is the guardian keeping in mind the seventh generation? Journal of Legislation and Public Policy, 5(1), 83–127. Eggert, L. L., & Nicholas, L. J. (2004). Reconnecting Youth: A peer group approach to building life skills. Bloomington, IN: National Educational Service. Eggert, L. L., Thompson, E. A., Randell, B. P., & Pike, K. C. (2002). Preliminary effects of brief school-based prevention approaches for reducing youth suicide— Risk behaviors, depression, and drug involvement. Journal of Child and Adolescent Psychiatric Nursing, 15, 48–64. http://dx.doi.org/10.1111/j.1744-6171.2002. tb00326.x Elias, B., Mignone, J., Hall, M., Hong, S. P., Hart, L., & Sareen, J. (2012). Trauma and suicide behaviour histories among a Canadian indigenous population: An empirical exploration of the potential role of Canada’s residential school system. Social Science & Medicine, 74, 1560–1569. http://dx.doi.org/10.1016/ j.socscimed.2012.01.026 Evans-Campbell, T. (2008). Historical trauma in American Indian/Native Alaska communities: A multilevel framework for exploring impacts on individuals, families, and communities. Journal of Interpersonal Violence, 23, 316–338. http:// dx.doi.org/10.1177/0886260507312290 Fantuzzo, J. W., DePaola, L. M., Lambert, L., Martino, T., Anderson, G., & Sutton, S. (1991). Effects of interparental violence on the psychological adjustment and competencies of young children. Journal of Consulting and Clinical Psychology, 59, 258–265. http://dx.doi.org/10.1037/0022-006X.59.2.258 Freedenthal, S., & Stiffman, A. R. (2004). Suicidal behavior in urban American Indian adolescents: A comparison with reservation youth in a southwestern state. Suicide and Life-Threatening Behavior, 34, 160–171. Freedenthal, S., & Stiffman, A. R. (2007). They might think I was crazy: Young American Indians’ reasons for not seeking help when suicidal. Journal of Adolescent Research, 22, 58–77. http://dx.doi.org/10.1177/0743558406295969 Goldsmith, S. K., Pellmar, T. C., Kleinman, A. M., & Bunney, W. E. (Eds.). (2002). Reducing suicide: A national imperative. Washington, DC: The National Academies Press. Goldston, D. B., Molock, S. D., Whitbeck, L. B., Murakami, J. L., Zayas, L. H., & Hall, G. C. N. (2008). Cultural considerations in adolescent suicide prevention and psychosocial treatment. American Psychologist, 63, 14–31. http://dx.doi. org/10.1037/0003-066X.63.1.14

240       lafromboise and malik

Goldston, D. B., Walrath, C. M., McKeon, R., Puddy, R. W., Lubell, K. M., Potter, L. B., & Rodi, M. S. (2010). The Garrett Lee Smith Memorial Suicide Prevention Program. Suicide and Life-Threatening Behavior, 40, 245–256. Goodkind, J. R., LaNoue, M. D., & Milford, J. (2010). Adaptation and implementation of cognitive behavioral intervention for trauma in schools with American Indian youth. Journal of Clinical Child and Adolescent Psychology, 39, 858–872. http://dx.doi.org/10.1080/15374416.2010.517166 Greenfield, L., & Smith, S. (1999). American Indians and crime. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. Washington, DC: U.S. Government Printing Office. Hallett, D., Chandler, M. J., & Lalonde, C. E. (2007). Aboriginal language knowledge and youth suicide. Cognitive Development, 22, 392–399. http://dx.doi. org/10.1016/j.cogdev.2007.02.001 Harvard Project on American Indian Economic Development. (2008). The state of the Native nations: Conditions under U.S. policies of self-determination. New York, NY: Oxford University Press. Havighurst, R. J. (1971). The extent and significance of suicide among American Indians today. Mental Hygiene, 55, 174–177. Hawkins, E. H., Cummins, L. H., & Marlatt, G. A. (2004). Preventing substance abuse in American Indian and Alaska native youth: Promising strategies for healthier communities. Psychological Bulletin, 130, 304–323. http://dx.doi. org/10.1037/0033-2909.130.2.304 Howard-Pitney, B., LaFromboise, T. D., Basil, M., September, B., & Johnson, M. (1992). Psychological and social indicators of suicide ideation and suicide attempts in Zuni adolescents. Journal of Consulting and Clinical Psychology, 60, 473–476. http://dx.doi.org/10.1037/0022-006X.60.3.473 Humes, K. R., Jones, N. A., & Ramirez, R. R. (2011, March). Overview of race and Hispanic origin: 2010 (2010 Census Briefs No. C2010BR-02). Washington, DC: U.S. Census Bureau. Indian Health Service. (2002). Regional differences in Indian health 2002–2003 edition. Washington, DC: U.S. Department of Health and Human Services. Kazdin, A. E., Rodgers, A., & Colbus, D. (1986). The hopelessness scale for children: Psychometric characteristics and concurrent validity. Journal of Consulting and Clinical Psychology, 54, 241–245. http://dx.doi.org/10.1037/0022-006X.54.2.241 Kellam, S. G., Brown, C. H., Poduska, J. M., Ialongo, N. S., Wang, W., Toyinbo, P., . . . Wilcox, H. C. (2008). Effects of a universal classroom behavior management program in first and second grades on young adult behavioral, psychiatric, and social outcomes. Drug and Alcohol Dependence, 95(Suppl. 1), S5–S28. http:// dx.doi.org/10.1016/j.drugalcdep.2008.01.004 Koss, M. P., Yuan, N. P., Dightman, D., Prince, R. J., Polacca, M., Sanderson, B., & Goldman, D. (2003). Adverse childhood exposures and alcohol dependence

a culturally informed approach     

241

among seven Native American tribes. American Journal of Preventive Medicine, 25, 238–244. http://dx.doi.org/10.1016/S0749-3797(03)00195-8 LaFromboise, T. D. (1991). Zuni life skills development curriculum. Unpublished curriculum. Stanford, CA: School of Education, Stanford University. LaFromboise, T. D. (1996). American Indian life skills development curriculum. Madison: University of Wisconsin Press. LaFromboise, T. D., Albright, K., & Harris, A. (2010). Patterns of hopelessness among American Indian adolescents: Relationships by levels of acculturation and residence. Cultural Diversity and Ethnic Minority Psychology, 16, 68–76. http://dx.doi.org/10.1037/a0016181 LaFromboise, T. D., & Howard-Pitney, B. (1993). The Zuni Life Skills Development Curriculum: A collaborative approach to curriculum development. American Indian and Alaska Native Mental Health Research, The Journal of the National Center, 4, 98–121. LaFromboise, T. D., & Howard-Pitney, B. (1995). The Zuni Life Skills Development Curriculum: Description and evaluation of a suicide prevention program. Journal of Counseling Psychology, 42, 479–486. http://dx.doi.org/10.1037/0022-0167.42.4.479 LaFromboise, T. D., & Lewis, H. A. (2008). The Zuni Life Skills Development Program: A school/community-based suicide prevention intervention. Suicide and Life-Threatening Behavior, 38, 343–353. LaFromboise, T. D., & Malik, S. S. (2012, May). Development of the American Indian Life Skills Curriculum: Middle School Version. Poster presentation, Second Biennial Conference of the Society for the Psychological Study of Ethnic Minority Issues, Ann Arbor, MI. LaFromboise, T. D., Medoff, L., Lee, C., & Harris, A. (2007). Psychosocial and cultural correlates of suicidal ideation among American Indian early adolescents on a Northern Plains reservation. Research in Human Development, 4(1–2), 119–143. LaFromboise, T. D., & Rowe, W. (1983). Skills training for bicultural competence: Rationale and application. Journal of Counseling Psychology, 30, 589–595. http:// dx.doi.org/10.1037/0022-0167.30.4.589 Lester, D. (1999). Native American suicide rates, acculturation stress and traditional integration. Psychological Reports, 84, 398. http://dx.doi.org/10.2466/ pr0.1999.84.2.398 Mackin, J., Perkins, T., & Furrer, C. (2012). The power of protection: A populationbased comparison of Native and non-Native youth suicide attempters. American Indian and Alaska Native Mental Health Research, 19(2), 20–54. http://dx.doi.org/ 10.5820/aian.1902.2012.20 May, P. A., Serna, P., Hurt, L., & DeBruyn, L. M. (2005). Outcome evaluation of a public health approach to suicide prevention in an American Indian tribal nation. American Journal of Public Health, 95, 1238–1244. http://dx.doi.org/10.2105/ AJPH.2004.040410

242       lafromboise and malik

Medoff, L. R. (2007). Suicidal ideation and related factors in Native American adolescents with and without learning disabilities. Dissertation Abstracts Inter­ national: Section B. The Sciences and Engineering 67(9-B), 5414. Meriam, L. (Ed.). (1928). The problem of Indian administration. Baltimore, MD: Johns Hopkins University Press. Moran, J. R., Fleming, C. M., Somervell, P., & Manson, S. M. (1999). Measuring bicultural ethnic identity among American Indian adolescents: A factor analytic study. Journal of Adolescent Research, 14, 405–426. http://dx.doi. org/10.1177/0743558499144002 Muehlenkamp, J. J., Marrone, S., Gray, J. S., & Brown, D. L. (2009). A college suicide prevention model for American Indian students. Professional Psychology: Research and Practice, 40, 134–140. http://dx.doi.org/10.1037/a0013253 Mullany, B., Barlow, A., Goklish, N., Larzelere-Hinton, F., Cwik, M., Craig, M., & Walkup, J. T. (2009). Toward understanding suicide among youths: Results from the White Mountain Apache tribally mandated suicide surveillance system, 2001–2006. American Journal of Public Health, 99, 1840–1848. http://dx.doi. org/10.2105/AJPH.2008.154880 National Center for Children in Poverty. (2007). Facts about trauma for policymakers. Retrieved from http://www.nccp.org/publications/pub_746.html Pettingell, S. L., Bearinger, L. H., Skay, C. L., Resnick, M. D., Potthoff, S. J., & Eichhorn, J. (2008). Protecting urban American Indian young people from suicide. American Journal of Health Behavior, 32, 465–476. http://dx.doi.org/10.5993/AJHB.32.5.2 Reynolds, W. M. (1988). Suicidal Ideation Questionnaire: Professional manual. Lutz, FL: Psychological Assessment Resources. Salvatore, N. F. (2011). American Indian Life Skills Curriculum: Evaluation results in San Juan County local implementation of classes, Fall 2010 and Spring 2011 (Unpublished report). Farmington, NM: Author. Smith, L. T. (1999). Decolonizing methodologies: Research and Indigenous peoples. New York, NY: Zed Books. Steele, M. M., & Doey, T. (2007). Suicidal behaviour in children and adolescents. Part 2: Treatment and prevention. Canadian Journal of Psychiatry, 52(6, Suppl. 1), 35S–45S. Thompson, E. A., Eggert, L. L., Randell, B. P., & Pike, K. C. (2001). Evaluation of indicated suicide risk prevention approaches for potential high school dropouts. American Journal of Public Health, 91, 742–752. http://dx.doi.org/10.2105/ AJPH.91.5.742 Tingey, L., Cwik, M., Goklish, N., Alchesay, M., Lee, A., Strom, R., . . . Barlow, A. (2012). Exploring binge drinking and drug use among American Indians: Data from adolescent focus groups. The American Journal of Drug and Alcohol Abuse, 38, 409–415. http://dx.doi.org/10.3109/00952990.2012.705204 Van Winkle, N. W., & May, P. A. (1986). Native American suicide in New Mexico, 1957–1979: A comparative study. Human Organization, 45, 296–309. http:// dx.doi.org/10.17730/humo.45.4.f1159w1x64k164t4 a culturally informed approach     

243

Van Winkle, N. W., & May, P. A. (1993). An update on American Indian suicide in New Mexico, 1980–1987. Human Organization, 52, 304–315. Waldram, J. B. (2004). Revenge of the Windigo: The construction of the mind and mental health of North American Aboriginal peoples. Toronto, Ontario, Canada: University of Toronto Press. Walker, R. D., & Bigelow, D. A. (2011). A constructive Indian country response to the evidence-based program mandate. Journal of Psychoactive Drugs, 43, 276–281. http://dx.doi.org/10.1080/02791072.2011.628910 Wallace, J. M., Jr., Bachman, J. G., O’Malley, P. M., Schulenberg, J. E., Cooper, S. M., & Johnston, L. D. (2003). Gender and ethnic differences in smoking, drinking and illicit drug use among American 8th, 10th and 12th grade students, 1976–2000. Addiction, 98, 225–234. http://dx.doi.org/10.1046/ j.1360-0443.2003.00282.x Walls, M. L., Chapple, C. L., & Johnson, K. D. (2007). Strain, emotion, and suicide among American Indian youth. Deviant Behavior, 28, 219–246. http://dx.doi. org/10.1080/01639620701233100 Walls, M. L., Whitbeck, L. B., Hoyt, D. R., & Johnson, K. D. (2007). Early-onset alcohol use among Native American youth: Examining female caretaker influence. Journal of Marriage and Family, 69, 451–464. http://dx.doi.org/10.1111/ j.1741-3737.2007.00376.x Walters, K. L., Mohammed, S. A., Evans-Campbell, T., Beltran, R. E., Chae, D. H., & Duran, B. (2011). Bodies don’t just tell stories, they tell histories: Embodiment of historical trauma among American Indians and Alaska Natives. Du Bois Review, 8, 179–189. http://dx.doi.org/10.1017/S1742058X1100018X Wexler, L. M. (2006). Inupiat youth suicide and culture loss: Changing community conversations for prevention. Social Science & Medicine, 63, 2938–2948. http:// dx.doi.org/10.1016/j.socscimed.2006.07.022 Wexler, L. M., Chandler, M., Gone, J. P., Cwik, M., Kirmayer, L. J., LaFromboise, T., . . . Allen, J. (2015). Advancing suicide prevention research with rural American Indian and Alaska Native populations. American Journal of Public Health, 105, 891–899. http://dx.doi.org/10.2105/AJPH.2014.302517 Wexler, L. M., & Gone, J. P. (2012). Culturally responsive suicide prevention in indigenous communities: Unexamined assumptions and new possibilities. American Journal of Public Health, 102, 800–806. http://dx.doi.org/10.2105/ AJPH.2011.300432 Whitbeck, L. B., Walls, M. L., Johnson, K. D., Morrisseau, A. D., & McDougall, C. M. (2009). Depressed affect and historical loss among North American Indigenous adolescents. American Indian and Alaska Native Mental Health Research, 16, 16–41. http://dx.doi.org/10.5820/aian.1603.2009.16 Whitbeck, L. B., Walls, M. L., & Welch, M. L. (2012). Substance abuse prevention in American Indian and Alaska Native communities. The American Journal of Drug and Alcohol Abuse, 38, 428–435. http://dx.doi.org/10.3109/ 00952990.2012.695416

244       lafromboise and malik

Whitbeck, L. B., Yu, M., Johnson, K. D., Hoyt, D. R., & Walls, M. L. (2008). Diagnostic prevalence rates from early to mid-adolescence among indigenous adolescents: First results from a longitudinal study. Journal of the American Academy of Child & Adolescent Psychiatry, 47, 890–900. http://dx.doi.org/10.1097/ CHI.0b013e3181799609 Wolsko, C., Lardon, C., Mohatt, G. V., & Orr, E. (2007). Stress, coping, and wellbeing among the Yup’ik of the Yukon-Kuskokwim Delta: The role of enculturation and acculturation. International Journal of Circumpolar Health, 66(1), 51–61. http://dx.doi.org/10.3402/ijch.v66i1.18226 Woodard, S. (2012). Suicide is epidemic for American Indian youth: What more can be done? Open Channel. Retrieved from http://investigations.nbcnews.com/ _news/2012/10/10/14340090-suicide-is-epidemic-for-american-indian-youthwhat-more-can-be-done Wyman, P. A., Brown, C. H., LoMurray, M., Schmeelk-Cone, K., Petrova, M., Yu, Q., . . . Wang, W. (2010). An outcome evaluation of the Sources of Strength suicide prevention program delivered by adolescent peer leaders in high schools. American Journal of Public Health, 100, 1653–1661. http://dx.doi.org/10.2105/ AJPH.2009.190025 Yoder, K. A., Whitbeck, L. B., Hoyt, D. R., & LaFromboise, T. (2006). Suicidal ideation among American Indian youths. Archives of Suicide Research, 10, 177–190. http://dx.doi.org/10.1080/13811110600558240

a culturally informed approach     

245

11 DEPRESSION PREVENTION AND TREATMENT INTERVENTIONS: EVOLUTION OF THE SAN FRANCISCO LATINO MENTAL HEALTH RESEARCH PROGRAM ADRIáN AGUILERA, JEANNE MIRANDA, SERGIO AGUILAR-GAXIOLA, KURT C. ORGANISTA, GERARDO M. GONZáLEZ, JOHN McQUAID, LAURA P. KOHN-WOOD, HUYNH-NHU LE, CHANDRA GHOSH-IPPEN, GUIDO G. URIZAR, JOSé SOTO, TAMAR MENDELSON, ALINNE Z. BARRERA, LEANDRO D. TORRES, YAN LEYKIN, STEPHEN SCHUELLER, NANCY LIU, AND RICARDO F. MUñOZ

The intent of this chapter is to illustrate how clinical research training programs based at public-sector hospitals can contribute to the development, evaluation, and dissemination of evidence-based practices for diverse populations. The interventions developed at San Francisco General Hospital (SFGH), a teaching hospital of the University of California, San Francisco (UCSF), are based on Albert Bandura’s (1969, 1997, 2001) social learning/ social cognitive theory. The basic structure of the depression interventions developed at SFGH consists of a focus on behavior, cognitions, and interpersonal factors; that is, activities, thoughts, and people. These stem directly from the joint dissertation conducted by Toni Zeiss, Mary Ann Youngren, and Ricardo Muñoz, under Lewinsohn’s direction (Zeiss, Lewinsohn, & Muñoz, 1979). In brief, the study compared three nonoverlapping depression treatments with each other and with a wait-list control. Treatment was provided

http://dx.doi.org/10.1037/14940-012 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

247

as 12 individual sessions, three sessions per week for 4 weeks. The behavioral treatment was designed by Lewinsohn, the cognitive treatment by Muñoz, and the interpersonal skill training treatment by Zeiss and Youngren. Community adults suffering from clinical depression were randomly assigned to immediate treatment in one of the three conditions or to a 4-week delayed treatment. The results showed that immediate treatment reduced depressive symptoms significantly more than did no treatment, and that there were no differences among the three treatment conditions. The treatment protocols were subsequently combined into a book called Control Your Depression (Lewinsohn, Muñoz, Youngren, & Zeiss, 1978). The protocols became the basis for most of the intervention manuals developed at SFGH. DEPRESSION PREVENTION RESEARCH PROJECT The Depression Prevention Research Project was the first randomized controlled trial (RCT) designed to determine whether the incidence of clinical depressive episodes could be reduced; that is, whether new cases of depression could be prevented (Muñoz, Beardslee, & Leykin, 2012; Muñoz & Ying, 1993; Muñoz et al., 1995). The study found that depressive symptoms were significantly reduced by the preventive intervention, but the incidence rates were too low to provide sufficient statistical power to detect significant differences. When Control Your Depression was in press, Art Ulene, a physician working for the NBC television network, produced a series of ten 4-minute segments illustrating the mood management methods from the book. These segments were shown during the NBC news programs nationally, at noon, 6 p.m., and 11 p.m., Monday through Friday for 2 weeks. When they were shown in San Francisco, Muñoz examined whether depressive symptoms in the general community were affected by the segments. He found that those with initially high depressive symptom scores who watched the segments showed significant reductions in symptoms (Muñoz, Glish, Soo-Hoo, & Robertson, 1982). To study the process more closely, Muñoz obtained a National Institute of Mental Health grant to carry out an RCT to test whether the cognitive– behavioral methods developed by Lewinsohn and colleagues could prevent clinical episodes of depression. The goal was to identify adult primary care patients at SFGH who were not currently clinically depressed and randomly assign consented participants to a preventive intervention (the Depression Prevention Course) or to a control condition. The study was conducted in Spanish and English, with a small pilot study conducted in Chinese by Yu-Wen Ying and Florentius Chan. The final sample of 150 participants consisted of 24% African Americans, 24% Latinos, 36% Whites, 10% Asians, and 5% Native Americans (Muñoz & 248       aguilera et al.

Ying, 1993); 139 were followed up at 1 year. The study found that, using the Beck Depression Inventory (BDI), the intervention group showed significantly lower depression symptoms. Only six of the 139 met criteria for a new episode of major depression during the year of the study: four in the control group and two in the experimental condition (both of whom dropped out early from the intervention). Although results were in the expected direction, the low incidence does not allow sufficient power to test whether the rate of new cases was significantly reduced. Depressive symptoms measured by the BDI (A. T. Beck, Ward, & Mendelson, 1961) were significantly lower after intervention when compared with the control condition. Cognitive and behavioral variables also mediated the reduction in symptoms as predicted (Muñoz et al., 1995). The Depression Prevention Course consisted of eight sessions presented in a group format, with one course instructor. The intervention was a class, not therapy, to reduce stigma and because the participants did not have a clinical diagnosis and thus were not “cases.” The eight sessions included an introductory session, two classes on thoughts, two on activities, two on people, and a final summary and graduation class. SAN FRANCISCO GENERAL HOSPITAL DEPRESSION CLINIC The SFGH Depression Clinic was located in the heart of San Francisco’s Latino Mission District. Between 1987 and 1997, a series of publications came from this work aimed at adapting cognitive–behavioral treatment (CBT) to the social and cultural needs of Mexican and Central American patients. This clinical and research experience helped trainees become culturally and linguistically competent mental health professionals and train others to achieve the same goals in the process. One of our first publications was a naturalistic effectiveness study of CBT for depression in our complex and challenging patient population with results showing clinically significant decreases from severe to moderate levels of clinical depression (Organista, Muñoz, & González, 1994). A study comparing group CBT with and without case management (Miranda, Azocar, Organista, Dwyer, & Areane, 2003) showed significant improvement in Spanish-speaking Latino patients when they were offered case management services. These experiences led to a publication describing how to recruit and retain Latino patients for psychotherapy services and research (Miranda, Azocar, Organista, Muñoz, & Lieberman, 1996) and an overview of the use of CBT with Latinos (Organista & Muñoz, 1996). The manualized treatment for depression that was developed at the SFGH Depression Clinic was later used in a large trial of dissemination of depression prevention and treatment interventions     

249

quality improvement programs for depression in managed primary care. Forty-six primary care clinics in six U.S. managed care organizations were randomized to usual care (mailing of practice guidelines) or the one of two quality improvement (QI) interventions: QI meds, which improved medication treatment; and QI therapy, which provided training and supervision for providers using the manualized treatment. The two QI interventions resulted in improved outcomes for patients at 6 and 12 months (Wells et al., 2000). Those receiving the intervention were less likely to be depressed over time. In a 9-year follow-up, QI therapy improved cumulative outcomes among minorities and reduced disparities (Wells et al., 2007). The manualized treatment was adapted further for use with poor young women. In a large randomized trial of this adapted treatment (the WE Care study), both short-term cognitive–behavioral treatment following the manual and short-term pharmacotherapy were found to be effective in treatment major depression in low-income African American and Latina women (Miranda, Chung, et al., 2003). In a 12-month follow-up, the psychotherapy intervention continued to be effective (Miranda et al., 2006). METHADONE MAINTENANCE STUDY After the initial work in developing and testing CBT interventions for depression, the approach was applied to other health and mental health issues. One pilot study tested a short-term culturally sensitive Spanishlanguage CBT intervention for Spanish-speaking methadone maintenance patients. The intervention was adapted as a 6-week mood management course for Spanish-speaking methadone maintenance patients (Muñoz, González, & Pérez-Arce, 1991). Our study, conducted at the SFGH methadone maintenance clinic, assessed the implementation feasibility (ability to recruit participants, participant willingness to accept the intervention, and participant attendance) and the effect of CBT on drug use and high-risk behaviors related to HIV in Spanish-speaking injection drug users. Initially, 35 patients were recruited for the study and 11 agreed to participate. Five of the participants were HIV-positive. Our study demonstrated that a short-term culturally sensitive Spanish language CBT intervention is feasible to implement with Spanish-speaking methadone maintenance patients. All participants attended at least half of the course sessions. Furthermore, the CBT intervention reduced depression symptom levels and the presence of current depressive episodes in these participants including HIV-positive patients. However, the intervention did not significantly decrease patient drug use or HIV-related highrisk behaviors, such as unsafe sexual practices and the sharing of needles. 250       aguilera et al.

In a subsequent replication with English-speaking methadone maintenance patients (Roehrich, Muñoz, & Sorensen, 1998), 23 patients were randomized either to receive a 10-week, twice-weekly mood management course immediately or to receive the same course after a 5-week delay. All participants were low income, most were minorities, and 87% were HIV-positive. Depressive symptoms were significantly reduced for both groups following the course, whereas no depression reduction was observed during the 5-week delay. In addition, the course yielded a decrease in HIV high-risk behaviors, a trend toward reduced drug use, and increased patient confidence in their ability to avoid drugs. VOICE RECOGNITION DEPRESSION SCREENING The voice recognition depression study derived from three key concerns. First, most individuals with depression who visited a health care provider are seen in a primary care setting, not a mental health setting (Shapiro et al., 1984). Second, at the time of this study, most cases of depression in primary care were not detected (Coyne, Schwenk, & Fechner-Bates, 1995). Third, Latinos in particular underuse mental health services because of barriers such as lack of Spanish-speaking mental health providers and lower literacy rates among economically disadvantaged immigrant populations (Preciado & Henry, 1997). To address these limitations, Muñoz and González initiated a pilot study of a bilingual computerized voice recognition depression screening tool and completed a randomized study examining the reliability, validity, and acceptability of this approach to assessment (Muñoz, McQuaid, González, Dimas, & Rosales, 1999). To test the use of voice recognition technology in a medical setting, we recruited 104 female participants (56 English-speaking and 48 Spanishspeaking) receiving care at the SFGH Women’s Clinic. Participants completed two measures in their language of preference: the Center for Epidemiologic Studies Depression scale (CES-D; Radloff, 1977) to assess depressive symptoms and the Major Depressive Episode (MDE) Screener (Muñoz, 1998) to screen for a diagnosis of major depression. Participants received each measure twice, in one of four randomly assigned orders: face-to-face (FF) twice, FF followed by computer administered (CA), CA followed by FF, or CA twice. In the CA condition, participants listened to the recorded questions for each item administered over a speaker and then responded to each item using a microphone. The program used speaker-dependent speech recognition technology, which in the 1990s required that the participant first train the computer to recognize the participant’s specific answers. Once training was completed, the participant then answered each question with one of the depression prevention and treatment interventions     

251

entrained responses. The study found high rates of agreement between the FF and CA assessments. The correlation on the CES-D was .89, and kappa for diagnosis between the two approaches was .82. Results were comparable for both English and Spanish speakers. Overall, the CA assessment was an effective means for screening for depression. This study was an early example of developing innovative uses of technology to increase services to individuals who otherwise would not be served. When this study was conducted, there were concerns about whether such an assessment would be acceptable to individuals without access to or familiarity with computers, and this was one of the first studies to show that such an approach was acceptable across languages and in a public health setting. AFRICAN AMERICAN WOMEN DEPRESSION INTERVENTION Another adaptation of the CBT work at the SFGH Depression Clinic targeted African American women. For this pilot study we focused on testing cultural adaptations to the existing CBT manual that included changes to the structure and process of treatment as well as change to the didactic content. We adapted the existing manualized protocol to treat African American women. We then compared the outcomes in demographically matched African American women receiving treatment with the adapted protocol versus the nonadapted protocol. Adaptations included limiting our group to African American women, maintaining a closed rather than open enrollment throughout the 16 weeks, adding experiential meditation exercises to open and/or close treatment sessions, using a termination ritual at the end of the 16-week period, and changing some of the language used to describe CBT techniques. Didactic or content adaptations included adding modules on healthy relationships, spirituality, racial identity, and issues related to African American families. We recruited 10 women into the African American Women’s CBT group (AACBT). All were diagnosed with major depressive disorder and had received an Axis IV/ psychosocial stressor diagnosis (typically “economic strain”). Half exhibited psychiatric comorbidity, and the majority also experienced a significant medical condition. The BDI was administered during the first and last week of treatment. We also used exploratory evaluative techniques such as audiotaping and coding group sessions for themes, affect, tone, process, and cohesion. While African American women in both the AACBT group and the CBT groups began treatment with BDI scores in the severe range, following treatment, women in the AACBT group reported a larger average drop in symptom intensity than did women in the nonadapted CBT group (posttreatment BDI 252       aguilera et al.

decrease of 12.6 points vs. 5.9 points, respectively). Predominant themes in the group discussions reflected the contextual, cultural, and psychological realities of the women’s lives, including suicidal ideation, substance abuse of significant others, racist experiences in social service agency interactions, social isolation, and caretaking. Many of these themes were opened up, at least in part, by the culturally adapted content of the new modules (Kohn, Oden, Muñoz, Robinson, & Leavitt, 2002). Our brief report describing our results has been cited several times in the burgeoning literature on cultural adaptations and cultural tailoring and on treatment outcomes for African American populations, including a seminal manuscript on cultural competence and evidence-based practice published in American Psychologist that describes our study as one of “only a couple . . . [that] contain designs that yielded information about the relative importance of cultural adaptations” (Whaley & Davis, 2007, p. 571). Further, we consistently receive requests for the AACBT manual, for our data to be included in meta-analytic reviews on cultural adaptations, and for our experience to be described in book chapters that focus on cultural tailoring. MAMáS Y BEBés/MOTHERS AND BABIES: MOOD AND HEALTH PROJECT The Mamás y Bebés/Mothers and Babies: Mood and Health Project (MB Project) began in 1997 at SFGH. Depression during the perinatal period is a significant public health problem (Horowitz & Goodman, 2005), is highly prevalent in low-income ethnically diverse women (Hobfoll, Ritter, Lavin, Hulsizer, & Cameron, 1995), and is associated with well-documented risk factors and negative consequences for the mother, infant, and the quality of the mother– infant relationship (National Research Council & Institute of Medicine, 2009). The goal of the MB Project was to evaluate the feasibility of preventing the onset of perinatal depression in low-income English- and Spanish-speaking women. Identification of Perinatal Women at Risk for Postpartum Depression Research demonstrating that more than half the cases of postpartum depression (PPD) in the United States go undetected (Guzmán, 2001) and that Latinas have higher rates of PPD than do other ethnic groups (Zayas, Jankowski, & McKee, 2003) highlighted the critical nature of preventive interventions targeted at this population. The benefits of perinatal depression prevention, moreover, are two-fold, protecting the mental health of both mother and child (Downey & Coyne, 1990). The Depression Prevention Research Project (see above) highlighted the importance of identifying a sample at imminent high risk for developing depression prevention and treatment interventions     

253

an MDE to have sufficient statistical power to detect preventive effects (see Muñoz, Cuijpers, Smit, Barrera, & Leykin, 2010, pp. 184–188). Unless incidence (i.e., the number of new cases) in the control group is substantial, statistical power is likely to be insufficient to detect lower incidence in an experimental group, even if the preventive intervention is effective (Muñoz et al., 2010). A pilot project focused on successfully identifying Latinas at risk of developing depression after pregnancy to establish a target group for future preventive interventions. Using a brief mood screener (Le & Muñoz, 1998; Muñoz, 1998) to measure history of MDEs and the CES-D (Radloff, 1977) to measure current depressive symptomatology, we were able to identify groups of pregnant Latinas with different vulnerabilities for developing PPD. Specifically, women with a history of at least one MDE and/or CES-D scores of 16 or greater at study entry were significantly more likely to develop an MDE within 1 year of giving birth than were those without these risk factors (26.7% vs. 2.9%; Le, Muñoz, Soto, Delucchi, & Ghosh Ippen, 2004). With this knowledge in mind, we were ready to evaluate an intervention targeted at preventing depression in expectant Latinas. Mothers and Babies Pilot Randomized Controlled Trial at UCSF/SFGH The MB Course (Muñoz et al., 2004) was adapted from depression prevention and treatment manuals developed with English- and Spanishspeaking public-sector patients at SFGH (Muñoz, 1984; Muñoz, Ghosh Ippen, Rao, Le, & Dwyer, 2000). The 12-week course integrates components for addressing depressive symptomatology using cognitive–behavioral (A. T. Beck, Rush, Shaw, & Emery, 1979; Lewinsohn et al., 1978) and interpersonal psychotherapy treatment approaches (Klerman, Weissman, Rounsaville, & Chevron, 1984) along with components to enhance physiological regulation during pregnancy, to strengthen the parent–child bond during and after pregnancy, and to help prospective mothers transfer the skills they learned during the course to their children. A pilot study conducted at SFGH involved 41 pregnant women, 70% of whom were Spanish-speaking Latinas considered at high risk for depression based on a history of MDE and/or a CES-D score ≥ 16 (Muñoz et al., 2007). Women were randomly assigned to the MB Course or to a comparison group. Results provided support for acceptability of the course with participants rating each class session positively (scores ranged from 4.1 to 4.5 on a scale of 1–5) and more than 90% indicating they would use the techniques in their daily lives. In this pilot study, analyses examining treatment effects did not reach statistical significance even though the MDE incidence rate was 14% for the intervention versus 25% for the comparison group over the postpartum year, which is a small effect size (h = 0.28). At 18 months, the 254       aguilera et al.

incidence in the control condition went up to 35%, with no change in the experimental group, suggesting that it is possible to reduce the proportion of new MDEs by half. Cortisol Results of Mothers and Babies Pilot Intervention Study As part of the Mothers and Babies pilot intervention study, salivary cortisol (a stress hormone) was collected from participants during pregnancy and at 6 and 18 months postpartum in light of recent research demonstrating that elevated stress levels during the pre- and postpartum period are related to poor maternal and infant health outcomes (Austin & Leader, 2000; Gunnar, 1998). The primary objective of this study was to examine whether the MB Course would be effective in regulating salivary cortisol and self-reported stress levels among mothers and their infants at 6 and 18 months postpartum, relative to two control groups. Our sample was composed of predominantly Spanish-speaking, low-income women (80%; mean age = 25 ± 5 years) who were screened for depression during their second trimester of pregnancy (M = 16 ± 5 weeks of gestation). Women at high risk for depression (i.e., having either a past history of major depression or current elevated symptoms of depression [ ≥ 16 on CES-D]) were randomized to either the MB Course (n = 24) or a usual care (UC) group (n = 33), while a low-risk comparison (LRC) group (n = 29) was composed of women not meeting either depression criteria. Results showed that (a) infants of women in the Mothers and Babies Course and LRC groups had significantly lower cortisol levels than did infants of women in the UC group at 6 months postpartum (p < .001) and (b) women in the MB Course had lower cortisol levels than did women in the UC group at 18 months postpartum (p < .01). This study is one of the few to prospectively show that prenatal cognitive–behavioral stress management interventions may be efficacious in regulating biological markers of stress among mothers and their infants, thereby decreasing their risk for later health complications (Urizar & Muñoz, 2011). First Years Study at UCSF/SFGH The First Years Study, was launched in the fall of 2001 as a component of the Mamás y Bebés/Mothers and Babies: Mood and Health Project (Urizar, Sanchez, & Muñoz, 2009). The main objective of this study was to document the cognitive development of children (ages 1–4) of low-income, Spanishand English-speaking mothers, originally identified as being at low or high risk for depression during pregnancy. Two cohorts were followed. Cohort 1 (Risk Identification Study) consisted of 100 pregnant women (at low and high risk for depression) who received usual prenatal care at SFGH in 1998. depression prevention and treatment interventions     

255

Cohort 2 (Mothers and Babies Pilot Intervention Study) consisted of 58 pregnant women (at high risk for depression) who participated in an RCT, examining the effects of a prenatal mood management course on preventing the incidence of MDEs. Data from the First Years Study (consisting of Cohorts 1 and 2) would allow us to prospectively examine whether maternal depressive symptoms (CES-D), assessed during pregnancy and 6 months postpartum, were associated with child cognitive development (Mullen Scales of Early Learning) at 2 to 4 years postpartum. Results demonstrated a significant child gender by prenatal depression interaction (controlling for birth complications and child language) with girls born to mothers with high depressive symptoms during pregnancy (CES-D ≥ 16) having lower overall cognitive development scores at 2 to 4 years postpartum, relative to girls born to mothers with low depressive symptoms (R2 = .10, p < .05). Further, boys born to mothers with high depressive symptoms at 6 months postpartum had lower fine motor skills and visual reception abilities at 2 to 4 years postpartum, relative to boys born to mothers with low depressive symptoms (R2 = .10, p < .05). Adaptation of the MB Project for Central American Women in Washington, DC Since its inception, the MB Project has moved beyond the SFGH boundaries. Given the promising findings in San Francisco with predominantly Mexican women (Muñoz et al., 2007), Huynh-Nhu Le obtained a major grant to conduct a larger RCT (N = 217) evaluating the efficacy of an 8-week version of the MB Course in the Washington, DC, metropolitan area (Le, Perry, & Stuart, 2011). A cultural adaptation five-step iterative process was used to reduce the 12-week MB Course to 8 weeks to further address the needs of the predominantly Central American women living in the Washington, DC region (Le, Zmuda, Perry, & Muñoz, 2010). Results from the RCT indicated that the MB intervention was efficacious in the short term (postintervention) but not at 1-year postpartum (Le et al., 2011). Adaptation of the MB Project for African American Women in Baltimore, MD Home visitation (HV) programs are a promising setting in which to deliver depression prevention programs because they serve large and growing numbers of low-income perinatal women; an estimated 400,000 to 500,000 women in 40 states receive HV services (Johnson, 2009). Moreover, research indicates high levels of depression among HV clients (Tandon, Parillo, Jenkins, & Duggan, 2005) that interfere with women’s ability to benefit from HV program services aimed at improving parenting and home environment 256       aguilera et al.

quality (Duggan, Berlin, Cassidy, Burrell, & Tandon, 2009). Further developments included adapting and implementing the MB program for use with urban African American women served by HV programs. The team conducted a series of 11 focus groups with clients, home visitors, and supervisors at three HV programs in Baltimore City. These qualitative data suggested that women in HV settings struggled with depression but perceived significant barriers to accessing and benefitting from mental health services (Leis, Mendelson, Perry, & Tandon, 2011) and indicated that the women would be interested in participating in skills-based groups aimed at managing stress and mood during pregnancy and postpartum. This research informed adaptations to the MB curriculum content to enhance its cultural appropriateness for an urban African American population. The team subsequently conducted a pilot RCT to assess intervention feasibility and preliminary outcomes in a sample of 78 women at four Baltimore HV programs. Findings indicated that the intervention significantly improved depressive symptoms at posttest, 3-month follow-up, and 6-month follow-up (Tandon, Perry, Mendelson, Kemp, & Leis, 2011). At the 6-month follow-up, 32.4% of women in the usual care control condition met criteria for a depressive episode using the Structured Clinical Interview for DSM Disorders compared with 14.6% of women in the MB Course. These findings lend further support to the possibility of being able to reduce new episodes of major depression by half. Doing so would be a major advance for psychological science and mental health care. Screening for High Risk for Perinatal Depression Although there are identifiable risk factors and effective treatments for perinatal depression, few women, especially low-income ethnic minority women, receive screening or treatment during this critical period. Reliable screening tools for perinatal depression are readily available and recommended for use in public health settings, yet screening remains limited in practice, in part because of the lack of coordination of services available to women seeking care during pregnancy and postpartum. As one of several federal programs that serve women and infants across the perinatal period, the Women, Infants, and Children (WIC) program can bridge this gap in screening and preventive services for their low-income populations. Le is currently conducting a study funded by the Maternal and Child Health Bureau to examine the feasibility of integrating routine screening for perinatal depression in the WIC program in the Washington, DC area. Women who screen at high risk for depression are asked to participate in a shortened 6-week version of the MB Course, which has been adapted for both pregnant women and mothers in the first postpartum year. This study is currently depression prevention and treatment interventions     

257

ongoing and will evaluate the effectiveness of the perinatal MB Course for preventing perinatal depression and improving the health and well-being of women and their infants. It will also evaluate the impact on WIC participation and outcomes up to the first year postpartum. Lessons learned will have implications for implementing perinatal depression services for low-income, ethnic minority WIC populations. MOTHERS AND BABIES INTERNET PROJECT To take advantage of the Internet’s global reach and to contribute toward the reduction of suffering among women and children, we adapted the Mamás y Bebés/Mothers and Babies course (Muñoz et al., 2007) to be delivered and tested online. The adaptation of the intervention occurred in four stages. First, references to the in-person group format and interactive activities that would be difficult to deliver in an automated online method were removed; second, the intervention content was reviewed and approved for conceptual fidelity by Muñoz; third, the intervention was translated into an electronic, web-based format using a data collection software program (e.g., adding short videos, updated photographs); and fourth, the site underwent usability testing with SFGH patients. Each stage resulted in multiple iterations prior to the launching of the site in January 2009. The primary goal of this online study was to examine the efficacy of the web-adapted course (eMB). We hypothesized that participants randomized to the eMB course would demonstrate lower rates of depressive symptoms and a reduced incidence of PPD. The secondary aim was to provide the lead investigator and trainee (Alinne Barrera) the opportunity to develop skills related to the use of innovative intervention techniques targeting depression prevention. The trial was a two-condition, pre-post design with a baseline assessment at study entry and monthly follow-up assessments up to 6 months postpartum. Cohort maintenance procedures (i.e., telephone follow-ups for those who did not respond to the automated e-mail invitations to return to the website) were implemented to reduce attrition. Participants who met high-risk criteria (CES-D score > 16 and/or past MDE history) comprised the cohort maintenance sample. During the first 6 months of recruitment, 2.053 women were eligible to participate, 1,000 consented to participate, and 658 (n = 658) were randomized to the MB Internet Course (n = 335) or the control condition (n = 323), which presented an informational brochure on perinatal depression (e.g., symptoms, prevalence, treatment). Fifty-nine percent (n = 388) of those randomized formed the cohort maintenance subsample. Monthly follow-ups were via automated e-mail invitation to assess for 258       aguilera et al.

pregnancy status, depressive symptoms, presence of PPD (at 3 and 6 months only), and use of the intervention materials. The study demonstrated that pregnant women are interested in online resources focused on methods to reduce their risk of PPD. The site was successful at recruiting, enrolling, and randomizing a global sample of pregnant women. However, it was less successful at retaining these participants throughout the 6-month follow-up period. Approximately one third completed at least one follow-up assessment. PPD assessments at 3 and 6 months were completed by 36 and 49 participants, respectively. Such a low completion rate makes findings difficult or impossible to interpret. This issue of massive dropout rates represents a major obstacle that will need to be addressed by Internet intervention researchers (Eysenbach, 2005). Neither incidence of MDEs nor PPD symptoms (as measured by the Edinburgh Postnatal Depression Scale; Cox, Holden, & Sagovsky, 1987) yielded significant group differences. The Internet adaptation of the MB Course provided an initial structure on how to develop and test web-based interventions. Future studies need to address difficulties with engagement, improve data collection methods, and update the site to reflect recent developments in technology. The 12-week version of the MB Course has been publicly available and free for downloading since the late 1990s at the UCSF website (http:// www.medschool2.ucsf.edu/latino/manuals.aspx). The 8-week and 6-week course versions are also available at the George Washington University website (http://mbp.columbian.gwu.edu/manuals). Various organizations have used the MB Course in their own settings. For example, the Oregon Easter Seals program has provided funding to conduct the MB Course in both English and Spanish since 2000. More recently, in California, the Riverside County Department of Mental Health received state funding to provide the 8-week MB Course for low-income Spanish-speaking Latinas. As more interest has been generated for implementing the MB intervention, training becomes an important issue to consider for dissemination purposes. THE STOP SMOKING AND DEPRESSION INTERNET PROJECT The adaptation of the MB Course to the Internet built on experience gained from the Stop Smoking and Depression Internet Project. We launched our first Internet intervention site in the year 2000, and thus began a series of studies on an Internet adaptation of a print-based self-help smoking cessation intervention that we had previously developed and tested in a randomized trial conducted in Spanish with Latinos in the San Francisco Bay area and administered entirely via surface mail (Muñoz et al., 2006; Muñoz, VanOss Marín, Posner, & Pérez-Stable, 1997). This previous intervention included depression prevention and treatment interventions     

259

a mood management component to address difficulties associated with depression that might arise during the course of a quit attempt. The RCT demonstrated that the mood management intervention significantly increased quit rates at 3 and 6 months and could be successfully delivered through the mail (Muñoz et al., 1997). We therefore decided to test whether these interventions could also be successfully delivered at a population level over the Internet. Our subsequent studies have convinced us that delivery of behavioral interventions has great potential for both smoking and depression. The potential reach of the Internet is undeniable: We have screened more than 130,000 Spanish- and English-speaking visitors to our Stop Smoking site from over 200 countries since the year 2000, with more than 60,000 participants consenting to participate in trials of the intervention. We have consistently found quit rates ranging from 17% to 21% at 12 months (Muñoz et al., 2006, 2009). These quit rates are comparable to those found with other methods (such as the nicotine patch, which yields 14% to 22% quit rates at 6 months, compared with 4% to 8% quit rates for placebo patches), demonstrating the potential of the Internet and modern digital communications technology more generally to deliver effective health interventions to large numbers of individuals across the globe (Barrera, Pérez-Stable, Delucchi, & Muñoz, 2009). We have also found important results related to smoking and depression. For example, despite common misconceptions to the contrary, we found no evidence that abstaining from smoking increases the odds for developing later depressive episodes among individuals with or without a history of past depression (Torres et al., 2010). Indeed, continued smoking was related to slightly increased odds of depressive episodes shortly after the quit attempt. Other secondary analyses have provided key insights. The importance of providing online, evidence-based smoking cessation was underscored in a study of more than 17,500 Spanish- and English-speaking participants from more than 157 countries entering our trials over a 5-year period; the use of effective smoking cessation methods was very low among these participants and they had smoking profiles generally consistent with those of more traditional smoking cessation studies (Barrera et al., 2009). Another study led us to reconsider the use of the CES-D in our ongoing online projects after we found evidence of variations in the factor structure for the CES-D in our large international Internet sample (Leykin, Torres, Aguilera, & Muñoz, 2011). After completing our Internet RCTs, we redesigned the website so that additional participants are able to choose any of the elements of the conditions that were tested in the RCTs. This allows thousands of users throughout the world to have access to whichever parts of our interventions they prefer. Thus, these outcome studies are termed participant preference trials. Our results indicate that these trials yield at least as high quit rates as do the RCTs 260       aguilera et al.

(Muñoz, Aguilera, et al., 2012). Unlike face-to-face interventions, which are generally discontinued after the grant funding the RCT ends, self-help automated Internet interventions can continue to be provided to thousands of people at relatively low cost (for hosting the site and paying a part-time staff person to respond to e-mail questions and make sure the site is up and running). Face-to-face interventions usually show drift away from the way they are administered during the RCT, when close supervision of videotaped sessions is usually a part of the study. Self-help automated Internet interventions show no drift: They are provided exactly as tested. Face-to-face interventions have an average lag of 17 years from the end of the randomized trials showing effectiveness until they are available to the general public (Institute of Medicine, 2001). Our Stop Smoking site was made available to anyone in the world for free 4 hours after the RCT ended. TEXT MESSAGING AS AN ADJUNCT TO CBT The clinical environment at public-sector settings such as SFGH is challenging and has constantly led clinicians and researchers to seek ways to improve care for a disadvantaged population experiencing multiple life stressors. As a result of depression symptoms and difficult life circumstances, patients often have difficulty attending sessions and adhering to homework assignments in our CBT groups. This difficulty led us to use mobile technology as an aid to help increase engagement with CBT content and improve skill rehearsal. We sent automated messages asking people to rate their mood daily as well as indicate the number of daily positive thoughts, pleasant activities, or social interactions (based on the monthly theme covered in the CBT group). We also sent medication and appointment reminders. Initial testing of the text-messaging adjunct to the CBT groups developed at SFGH showed that the technology is well liked and can be feasibly employed with low-resourced populations. An initial testing with 12 participants yielded response rates of 65% with a range of 23% to 99% (Aguilera & Muñoz, 2011). We were initially targeting increases in adherence to CBT homework, but what became evident during the pilot testing was that patients who received text messages commented on the feeling of connectedness with their therapist. Many commented that they felt “cared for” when they received a message asking them about their mood or reminding them to take their medications (Aguilera & Berridge, 2014). Although the text messaging system is automated, patients regularly responded with “Gracias” after being reminded to take medications. During the holidays, some even sent messages saying “Feliz Navidad.” These indicate that although participants were told that messages are automated, they nonetheless feel that someone was on depression prevention and treatment interventions     

261

the other end reading their responses. This fact is very informative to the development of technology adjuncts and stand-alone interventions because patients may be more likely to respond and engage with technology when they believe that someone will be reviewing the data. Given the positive response we received during the feasibility pilot study, we are expanding the intervention and are planning a more thorough test of the efficacy of the adjunct compared with standard treatment. Future applications of technology will look to engage family members by using mobile phones as a vehicle for psychoeducation about depression and contacting close social supports chosen by a patient when he or she is experiencing difficulties. MULTILINGUAL ONLINE DEPRESSION SCREENER The online depression screener—the Mood Screener—was launched to estimate the prevalence of depression among members of the Internet looking for information on depression online. Launched first in English (December 2009), the screener was expanded to Russian, Chinese, and Spanish in the fall of 2011. As of May 15, 2012, more than 175,000 people had visited our four sites and more than 72,000 were screened for depression. In our 2011 report on the first year of the English screener (Leykin, Muñoz, & Contreras, 2012), a high proportion of visitors to our site (66.6% of 24,965 people) met criteria for major depression based on our screener. However, only 25% of those screening positive for depression sought help for their depression. The main recruitment channel for this study is Google AdWords— the placement of sponsored links in Google search results. The advertisement appears in search inquiries for depression, sadness, suicide, and related keywords in the four study languages. Participants arriving on the site are screened for eligibility (18 years of age or older); they are also asked about their country of residence, race, and gender. Eligible participants then complete the MDE Screener (Muñoz, 1998), an 18-item self-report depression screener that assesses for the presence of nine symptom of depression as well as for difficulties in functioning due to these symptoms. The MDE Screener has good agreement with clinician-administered interviews (Muñoz et al., 1999; Vázquez, Muñoz, Blanco, & López, 2008). Participants receive personalized feedback based on their answers to the MDE screener. Those with high symptoms and those indicating suicidality are urged to seek care. All participants are then asked to join a research study, where they will be e-mailed monthly invitations to recheck their mood, for the next 12 months. Those consenting answer further questions about themselves (other demographics, health, and questions about their treatment for depression). They then complete the Lifetime Depression portion of the MDE screener, which asks about previous 262       aguilera et al.

TABLE 11.1 Data From the Mood Screener Language Date launched Countries represented Screened for eligibility Current MDE Suicide attempt, past 2 weeks Past MDE Suicide attempt, past

English

Russian

Chinese

Spanish

12/09/09 174 65,731 67% 8% 79% 18%

09/27/11 138 15,177 63% 7% 70% 15%

11/03/11 103 5,731 49% 20% 64% 23%

11/09/11 85 8,082 71% 10% 77% 14%

Note. MDE = major depressive episode.

episodes of depression, and receive feedback on their responses. Finally, consented participants receive a monthly e-mail invitation to return to the site and rescreen for depression. Table 11.1 presents selected data from the four languages as of May 15, 2012. Individuals increasingly turn to the Internet for answers to their questions about health (Atkinson, Saperstein, & Pleis, 2009; Fox, 2006), with up to 88% of Americans with Internet access doing so (Harris Poll, 2010). People use health information obtained on the Internet to inform their subsequent behavior (Fox, 2006). Many Internet users are reluctant to engage with mental health professionals. Therefore, it is imperative to study their needs and experiences, as well as to offer services, where they already are—on the Internet. The online screener, aside from being an important research tool, provides a service that is available to all adults who are interested in being screened for major depression. By offering participants feedback on their responses, we hope to motivate individuals to seek care from their health care providers. By making the site available in four of the 10 most spoken languages on the Internet, we offer this service to over 60% of all Internet users (InternetWorldStats.com, 2011). To date, we have been able to reach individuals from 198 countries, making this a global health resource. DEPRESSION PREVENTION AND DEPRESSION MANAGEMENT COURSES With our Depression Prevention and Depression Management courses, we hoped to offer individuals—whether they meet criteria for a major depressive disorder or are only at risk for the disorder—an opportunity to take control of their symptoms. In June 2010, we launched the Depression Management Course (DMC), a self-help Internet-based depression management intervention. The depression prevention and treatment interventions     

263

Depression Prevention Course (DPC)—a self-help Internet-based depression prevention intervention—was launched in July 2011, with the Spanish version following in November of the same year. The two studies share a similar architecture, illustrated in Figure 11.1. Briefly, individuals complete a brief demographic and eligibility questionnaire. Those who are eligible sign a consent document and proceed to the baseline questionnaires. Those completing the baseline measures are randomized to a research condition (phone or e-mail-only follow-up, DPC only) or assigned to a lesson order optimized for their symptoms (DMC only). After receiving their unique username and password, participants are permitted to complete the course at their own pace and are able to skip or revisit any lesson. On every site visit, participants indicate their mood as well as other course-specific variables (e.g., number of pleasant activities). Participants receive follow-up e-mails 1, 3, and 6 months from the date of consent. A randomly chosen subsample of the DPC participants not responding to e-mails are contacted by phone. The DPC intervention consists of an adaptation of the Coping With Depression Course, developed by Lewinsohn et al. (1978) at the University of Oregon. The intervention provides modules addressing three aspects of life that affect mood: activities, people, and thoughts. Each module has two parts: the first helps participants identify activities, people, and thoughts that affect their mood; and the second helps participants increase those behaviors that have a positive impact on mood and decrease those with negative impact. The DMC consists of eight lessons based on the classic texts of cognitive therapy for depression (A. T. Beck, Rush, Shaw, & Emery, 1979; J. Beck, 1995) as well as on the clinical experience and interests of Yan Leykin and Ricardo Muñoz. The lessons address automatic thoughts and cognitive restructuring, behavior activation, decision-making strategies, depressogenic and healthy environments, and perfectionism. The site also lists resources for participants. Both DPC and DMC are currently in the pilot-testing phase, and efficacy data are not yet available. To pilot the DMC, 100 people screening positive for an MDE via the online depression screener described in this chapter

Initial DPC/DMC visit Screening

Consent

Baseline surveys

All subsequent DPC/DMC visits Personal log-in

Assessment

Course lessons

Follow-up 1 month 3 month 6 month

Figure 11.1.  Depression Prevention Course (DPC)/Depression Management Course (DMC) structure and participant flow.

264       aguilera et al.

TABLE 11.2 Depression Management and Prevention Course Websites Course Date launched Participants screened Participants consented Total course visits Total lesson visits

Depression Prevention: English

Depression Prevention: Spanish

Depression Management

July 2011 134  90 780 446

November 2011 204  23  62  33

June 2010  147  128 1,489  324

were invited; additional participants came from the National Clinical Trial registry and other sources. DMC accepted all individuals over 18 years of age. The main recruitment channel of the DPC was Google AdWords. Eligible individuals are U.S. residents at risk for future depressive episodes but not currently meeting MDE criteria. The initial data on recruitment flow through the two sites are presented in Table 11.2, which shows that the participants are interested in the courses and return to the websites multiple times. Participants’ comments regarding the course and individual lessons have been largely positive. The overarching goal of this project is to provide accessible, appealing, and effective interventions for the prevention and management of depression, available anywhere in the world. This approach allows the skills and knowledge of SFGH trainees and faculty to benefit not just SFGH patients, but those throughout the world. Too few individuals have access to quality depression care, and many choose not to engage such care if available. These interventions, and those like them, have the potential of reducing health disparities by offering access to treatment options to underserved individuals in need. CONCLUSION In this chapter, we described and gave specific examples of how a clinical research program can be developed within a public-sector service context to address issues relevant to Latino and other underserved populations. The authors continue to be firmly committed to including the neediest populations in cutting-edge clinical research. Rooted in aspects of cognitive and behavioral theories, we have been able to apply principles in multiple settings, for multiple targets, using varied mediums to decrease suffering resulting from depression. Our efforts have shown that public-sector patients will participate in clinical research that is culturally and linguistically relevant and addresses depression prevention and treatment interventions     

265

their concerns. Public-sector patients can benefit from the latest advances in mental health interventions and from technological advances, as long as the way the technology is used is adapted to their needs, including language and simplicity of use. It is also possible (and highly important) to expand service delivery beyond treatment into prevention. We have found ways to identify subgroups at high risk for new episodes of major depression within the next 12 months. By teaching high-risk individuals mood management skills similar to those taught to patients in CBT, we can reduce new cases of MDE by approximately half (Muñoz, Schueller, Barrera, Le, & Torres, 2014). It is possible to provide effective psychological interventions via the Internet and to conduct evaluation using RCT methods with participants from around the globe. REFERENCES Aguilera, A., & Berridge, C. (2014). Qualitative feedback from a text messaging intervention for depression: Benefits, drawbacks, and cultural differences. JMIR mHealth uHealth, 2(4), e46. Aguilera, A., & Muñoz, R. F. (2011). Text messaging as an adjunct to CBT in lowincome populations: A usability and feasibility pilot study. Professional Psychology: Research and Practice, 42, 472–478. http://dx.doi.org/10.1037/a0025499 Atkinson, N. L., Saperstein, S. L., & Pleis, J. (2009). Using the internet for healthrelated activities: Findings from a national probability sample. Journal of Medical Internet Research, 11(1), e4. http://dx.doi.org/10.2196/jmir.1035 Austin, M. P., & Leader, L. (2000). Maternal stress and obstetric and infant outcomes: Epidemiological findings and neuroendocrine mechanisms. Australian and New Zealand Journal of Obstetrics and Gynaecology, 40, 331–337. http:// dx.doi.org/10.1111/j.1479-828X.2000.tb03344.x Bandura, A. (1969). Principles of behavior modification. New York, NY: Holt Rinehart & Winston. Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman. Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1–26. http://dx.doi.org/10.1146/annurev.psych.52.1.1 Barrera, A. Z., Pérez-Stable, E. J., Delucchi, K. L., & Muñoz, R. F. (2009). Global reach of an Internet smoking cessation intervention among Spanish- and Englishspeaking smokers from 157 countries. International Journal of Environmental Research and Public Health, 6, 927–940. http://dx.doi.org/10.3390/ijerph6030927 Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression. New York, NY: Guilford Press. Beck, A. T., Ward, C., & Mendelson, M. (1961). Beck depression inventory (BDI). Archives of General Psychiatry, 4, 561–571. Beck, J. (1995). Cognitive therapy: Basics and beyond. New York, NY: Guilford Press.

266       aguilera et al.

Cox, J. L., Holden, J. M., & Sagovsky, R. (1987). Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. The British Journal of Psychiatry, 150, 782–786. http://dx.doi.org/10.1192/ bjp.150.6.782 Coyne, J. C., Schwenk, T. L., & Fechner-Bates, S. (1995). Nondetection of depression by primary care physicians reconsidered. General Hospital Psychiatry, 17, 3–12. Downey, G., & Coyne, J. C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin, 108, 50–76. http://dx.doi.org/10.1037/ 0033-2909.108.1.50 Duggan, A. K., Berlin, L. J., Cassidy, J., Burrell, L., & Tandon, S. D. (2009). Examining maternal depression and attachment insecurity as moderators of the impacts of home visiting for at-risk mothers and infants. Journal of Consulting and Clinical Psychology, 77, 788–799. http://dx.doi.org/10.1037/a0015709 Eysenbach, G. (2005). The law of attrition. Journal of Medical Internet Research, 7(1), e11. http://dx.doi.org/10.2196/jmir.7.1.e11 Fox, S. (2006, Oct. 29). Online health search 2006: Most Internet users start at a search engine when looking for health information online, very few check the source and date of the information they find. Washington, DC: Pew Internet & American Life Project. Retrieved from http://www.pewinternet.org/files/old-media/Files/ Reports/2006/PIP_Online_Health_2006.pdf.pdf Gunnar, M. R. (1998). Quality of early care and buffering of neuroendocrine stress reactions: Potential effects on the developing human brain. Preventive Medicine, 27, 208–211. http://dx.doi.org/10.1006/pmed.1998.0276 Guzmán, B. (2001, May). The Hispanic population: 2000 (Census 2000 Brief No. C2KBR/01-3). Washington, DC: U.S. Census Bureau. Retrieved from http:// www.census.gov/prod/2001pubs/c2kbr01-3.pdf Harris Poll. (2010). “Cyberchondriacs” on the rise? Retrieved from http://www. prnewswire.com/news-releases/cyberchondriacs-on-the-rise-99925874.html Hobfoll, S. E., Ritter, C., Lavin, J., Hulsizer, M. R., & Cameron, R. P. (1995). Depression prevalence and incidence among inner-city pregnant and postpartum women. Journal of Consulting and Clinical Psychology, 63, 445–453. http:// dx.doi.org/10.1037/0022-006X.63.3.445 Horowitz, J. A., & Goodman, J. H. (2005). Identifying and treating postpartum depression. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 34, 264–273. http://dx.doi.org/10.1177/0884217505274583 Institute of Medicine. (2001). Crossing the quality chasm: A new healthcare system for the 21st century. Washington, DC: National Academy Press. InternetWorldStats.com. (2011). The Internet big picture: World Internet users and population stats. Retrieved from http://www.internetworldstats.com/stats.htm Johnson, K. (2009). State-based home visiting: Strengthening programs through state leadership. Washington, DC: National Center for Children in Poverty. Klerman, G. L., Weissman, M. M., Rounsaville, B. J., & Chevron, E. S. (1984). Interpersonal psychotherapy of depression. New York, NY: Basic Books. depression prevention and treatment interventions     

267

Kohn, L. P., Oden, T., Muñoz, R. F., Robinson, A., & Leavitt, D. (2002). Adapted cognitive behavioral group therapy for depressed low-income African American women. Community Mental Health Journal, 38, 497–504. http://dx.doi.org/ 10.1023/A:1020884202677 Le, H.-N., & Muñoz, R. F. (1998). The Maternal Mood Screener. [Unpublished raw data]. Retrieved from http://www.medschool2.ucsf.edu/latino/manuals.aspx Le, H.-N., Muñoz, R., Soto, J. A., Delucchi, K., & Ghosh Ippen, C. (2004). Identification of risk for onset of major depressive episodes during pregnancy and postpartum. Hispanic Journal of Behavioral Sciences, 26, 463–482. http:// dx.doi.org/10.1177/0739986304269165 Le, H.-N., Perry, D. F., & Stuart, E. A. (2011). Randomized controlled trial of a preventive intervention for perinatal depression in high-risk Latinas. Journal of Consulting and Clinical Psychology, 79, 135–141. http://dx.doi.org/10.1037/ a0022492 Le, H.-N., Zmuda, J., Perry, D. F., & Muñoz, R. F. (2010). Transforming an evidencebased intervention to prevent perinatal depression for low-income Latina immigrants. American Journal of Orthopsychiatry, 80, 34–45. http://dx.doi.org/ 10.1111/j.1939-0025.2010.01005.x Leis, J. A., Mendelson, T., Perry, D. F., & Tandon, S. D. (2011). Perceptions of mental health services among low-income, perinatal African-American women. Women’s Health Issues, 21, 314–319. http://dx.doi.org/10.1016/j.whi.2011.03.005 Lewinsohn, P. M., Muñoz, R. F., Youngren, M. A., & Zeiss, A. M. (1978). Control your depression. Englewood Cliffs, NJ: Prentice-Hall. Leykin, Y., Muñoz, R. F., & Contreras, O. (2012). Are consumers of Internet health information “cyberchondriacs”? Characteristics of 24,965 users of a depression screening site. Advance online publication. Depression and Anxiety, 29, 71–77. Leykin, Y., Torres, L. D., Aguilera, A., & Muñoz, R. F. (2011). Factor structure of the CES-D in a sample of Spanish- and English-speaking smokers on the Internet. Psychiatry Research, 185(1–2), 269–274. http://dx.doi.org/10.1016/ j.psychres.2010.04.056 Miranda, J., Azocar, F., Organista, K. C., Dwyer, E., & Areane, P. (2003). Treatment of depression among impoverished primary care patients from ethnic minority groups. Psychiatric Services, 54, 219–225. http://dx.doi.org/10.1176/ appi.ps.54.2.219 Miranda, J., Azocar, F., Organista, K. C., Muñoz, R. F., & Lieberman, A. (1996). Recruiting and retaining low-income Latinos in psychotherapy research. Journal of Consulting and Clinical Psychology, 64, 868–874. http://dx.doi.org/10.1037/ 0022-006X.64.5.868 Miranda, J., Chung, J. Y., Green, B. L., Krupnick, J., Siddique, J., Revicki, D. A., & Belin, T. (2003). Treating depression in predominantly low-income young minority women: A randomized controlled trial. JAMA, 290, 57–65. http:// dx.doi.org/10.1001/jama.290.1.57

268       aguilera et al.

Miranda, J., Green, B. L., Krupnick, J. L., Chung, J., Siddique, J., Belin, T., & Revicki, D. (2006). One-year outcomes of a randomized clinical trial treating depression in low-income minority women. Journal of Consulting and Clinical Psychology, 74, 99–111. Muñoz, R. F. (1984). The Depression Prevention Course. Retrieved from http://www. medschool2.ucsf.edu/latino/manuals.aspx Muñoz, R. F. (1998). The Major Depressive Episode (MDE) Screener. Retrieved from http://www.medschool2.ucsf.edu/latino/manuals.aspx Muñoz, R. F., Aguilera, A., Schueller, S. M., Leykin, Y., & Pérez-Stable, E. J. (2012). From online randomized controlled trials to participant preference studies: Morphing the San Francisco Stop Smoking site into a worldwide smoking cessation resource. Journal of Medical Internet Research, 14(3), e64. http:// dx.doi.org/10.2196/jmir.1852 Muñoz, R. F., Barrera, A. Z., Delucchi, K., Penilla, C., Torres, L. D., & Pérez-Stable, E. J. (2009). International Spanish/English Internet smoking cessation trial yields 20% abstinence rates at 1 year. Nicotine & Tobacco Research, 11, 1025–1034. http://dx.doi.org/10.1093/ntr/ntp090 Muñoz, R. F., Beardslee, W. R., & Leykin, Y. (2012). Major depression can be prevented. American Psychologist, 67, 285–295. http://dx.doi.org/10.1037/a0027666 Muñoz, R. F., Cuijpers, P., Smit, F., Barrera, A. Z., & Leykin, Y. (2010). Prevention of major depression. Annual Review of Clinical Psychology, 6, 181–212. http:// dx.doi.org/10.1146/annurev-clinpsy-033109-132040 Muñoz, R. F., Ghosh Ippen, C., Rao, S., Le, H. L., & Dwyer, E. V. (2000). Manual de terapia de grupo para el tratamiento cognitivo-conductual de la depresión: Aprendiendo a manejar su realidad personal [Group cognitive behavioral therapy for depression manual: Learning to mange your personal reality]. Retrieved from http://www.medschool2.ucsf.edu/latino/manuals.aspx Muñoz, R. F., Glish, M., Soo-Hoo, T., & Robertson, J. (1982). The San Francisco Mood Survey Project: Preliminary work toward the prevention of depression. American Journal of Community Psychology, 10, 317–329. http://dx.doi.org/ 10.1007/BF00896498 Muñoz, R. F., González, G., & Pérez-Arce, P. (1991). Depression prevention course for Spanish-speaking methadone maintenance patients. Unpublished manuscript, University of California, San Francisco. Muñoz, R. F., Le, H.-N., Ghosh-Ippen, C., Diaz, M. A., Urizar, G., & Lieberman, A. F. (2004). Mothers and Babies Course—Instructor’s manual. Unpublished manuscript, University of California, San Francisco/San Francisco General Hospital. Muñoz, R. F., Le, H.-N., Ghosh Ippen, C., Diaz, M. A., Urizar, G. G., Jr., Soto, J., . . . Lieberman, A. F. (2007). Prevention of postpartum depression in low-income women: Development of the Mamás y Bebés/Mothers and Babies course. Cognitive and Behavioral Practice, 14, 70–83. http://dx.doi.org/10.1016/j.cbpra. 2006.04.021 depression prevention and treatment interventions     

269

Muñoz, R. F., Lenert, L. L., Delucchi, K., Stoddard, J., Perez, J. E., Penilla, C., & Pérez-Stable, E. J. (2006). Toward evidence-based Internet interventions: A Spanish/English website for international smoking cessation trials. Nicotine & Tobacco Research, 8, 77–87. http://dx.doi.org/10.1080/14622200500431940 Muñoz, R. F., McQuaid, J. R., González, G. M., Dimas, J., & Rosales, V. A. (1999). Depression screening in a women’s clinic: Using automated Spanish- and English-language voice recognition. Journal of Consulting and Clinical Psychology, 67, 502–510. http://dx.doi.org/10.1037/0022-006X.67.4.502 Muñoz, R. F., Schueller, S. M., Barrera, A. Z., Le, H.-N., & Torres, L. (2014). Major depression can be prevented: Implications for research and practice. In I. H. Gotlib & C. L. Hammen (Eds.), Handbook of depression (3rd ed., pp. 471–491). New York, NY: Guilford Press. Muñoz, R. F., VanOss Marín, B., Posner, S. F., & Pérez-Stable, E. J. (1997). Mood management mail intervention increases abstinence rates for Spanish-speaking Latino smokers. American Journal of Community Psychology, 25, 325–343. http:// dx.doi.org/10.1023/A:1024676626955 Muñoz, R. F., & Ying, Y. W. (1993). The prevention of depression: Research and practice. Baltimore, MD: Johns Hopkins University Press. Muñoz, R. F., Ying, Y. W., Bernal, G., Pérez-Stable, E. J., Sorensen, J. L., Hargreaves, W. A., . . . Miller, L. S. (1995). Prevention of depression with primary care patients: A randomized controlled trial. American Journal of Community Psychology, 23, 199–222. http://dx.doi.org/10.1007/BF02506936 National Research Council, & Institute of Medicine. (2009). Depression in parents, parenting, and children: Opportunities to improve identification, treatment, and prevention. Washington, DC: The National Academies Press. Organista, K. C., & Muñoz, R. F. (1996). Cognitive behavioral therapy with Latinos. Cognitive and Behavioral Practice, 3, 255–270. http://dx.doi.org/10.1016/ S1077-7229(96)80017-4 Organista, K. C., Muñoz, R. F., & González, G. (1994). Cognitive behavioral therapy for depression in low-income and minority medical outpatients: Description of a program and exploratory analyses. Cognitive Therapy and Research, 18, 241–259. http://dx.doi.org/10.1007/BF02357778 Preciado, J., & Henry, M. (1997). Linguistic barriers in health education and services. In J. G. Garcia & M. C. Zea (Eds.), Psychological interventions and research with Latino populations (pp. 235–254). Boston, MA: Allyn & Bacon. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. http:// dx.doi.org/10.1177/014662167700100306 Roehrich, L., Muñoz, R. F., & Sorensen, J. L. (1998, January). Mood management for injection drug users. Poster presented at the 8th International Congress on the Treatment of Addictive Behavior, Santa Fe, NM. Shapiro, S., Skinner, E. A., Kessler, L. G., Von Korff, M., German, P. S., Tischler, G. L., . . . Regier, D. A. (1984). Utilization of health and mental health services:

270       aguilera et al.

Three Epidemiologic Catchment Area sites. Archives of General Psychiatry, 41, 971–978. http://dx.doi.org/10.1001/archpsyc.1984.01790210053007 Tandon, S. D., Parillo, K. M., Jenkins, C., & Duggan, A. K. (2005). Formative evaluation of home visitors’ role in addressing poor mental health, domestic violence, and substance abuse among low-income pregnant and parenting women. Maternal and Child Health Journal, 9, 273–278. http://dx.doi.org/10.1007/s10995-0050012-8 Tandon, S. D., Perry, D. F., Mendelson, T., Kemp, K., & Leis, J. A. (2011). Preventing perinatal depression in low-income home visiting clients: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 79, 707–712. http:// dx.doi.org/10.1037/a0024895 Torres, L. D., Barrera, A. Z., Delucchi, K., Penilla, C., Pérez-Stable, E. J., & Muñoz, R. F. (2010). Quitting smoking does not increase the risk of major depressive episodes among users of Internet smoking cessation interventions. Psychological Medicine, 40, 441–449. http://dx.doi.org/10.1017/S0033291709990560 Urizar, G. G., Jr., & Muñoz, R. F. (2011). Impact of a prenatal cognitive-behavioral stress management intervention on salivary cortisol levels in low-income mothers and their infants. Psychoneuroendocrinology, 36, 1480–1494. http://dx.doi. org/10.1016/j.psyneuen.2011.04.002 Urizar, G. G., Jr., Sanchez, J. A., & Muñoz, R. F. (2009). Pre- and post-natal depression and its effects on later cognitive development among low-income, multiethnic children. Annals of Behavioral Medicine, 37, S226. Vázquez, F. L., Muñoz, R. F., Blanco, V., & López, M. (2008). Validation of Muñoz’s Mood Screener in a nonclinical Spanish population. European Journal of Psychological Assessment, 24, 57–64. http://dx.doi.org/10.1027/1015-5759.24.1.57 Wells, K. B., Sherbourne, C. D., Miranda, J., Tang, L., Benjamin, B., & Duan, N. (2007). The cumulative effects of quality improvement for depression on outcome disparities over 9 years: Results from a randomized, controlled group-level trial. Medical Care, 45, 1052–1059. http://dx.doi.org/10.1097/ MLR.0b013e31813797e5 Wells, K. B., Sherbourne, C. D., Schoenbaum, M., Duan, N., Meredith, L., Unützer, J., . . . Rubenstein, L. V. (2000). Impact of disseminating quality improvement programs for depression in managed primary care: A randomized controlled trial. JAMA, 283, 212–220. http://dx.doi.org/10.1001/jama.283.2.212 Whaley, A. L., & Davis, K. E. (2007). Cultural competence and evidence-based practice in mental health services: A complementary perspective. American Psychologist, 62, 563–574. http://dx.doi.org/10.1037/0003-066X.62.6.563 Zayas, L. H., Jankowski, K. R. B., & McKee, M. D. (2003). Prenatal and postpartum depression among low-income, Dominican and Puerto Rican women. Hispanic Journal of Behavioral Sciences, 25, 370–385. http://dx.doi.org/10.1177/ 0739986303256914 Zeiss, A. M., Lewinsohn, P. M., & Muñoz, R. F. (1979). Nonspecific improvement effects in depression using interpersonal skills training, pleasant activity schedules, or cognitive training. Journal of Consulting and Clinical Psychology, 47, 427–439. depression prevention and treatment interventions     

271

12 IMPROVING THE PARTICIPATION OF FAMILIES OF COLOR IN EVIDENCE-BASED INTERVENTIONS: CHALLENGES AND LESSONS LEARNED NORWEETA G. MILBURN AND MARGUERITA LIGHTFOOT

African American and Latina/o adolescents are at high risk for longterm negative, HIV/AIDS-related outcomes (Anderson & Smith, 2005; Centers for Disease Control and Prevention, 2005; Flores et al., 2002), which are intensified for those who are also delinquent (Benda, Flynn Corwin, & Toombs, 2001; Morris et al., 1995; Romero et al., 2007). Delinquent adolescents are young people under the age of 18 years who commit a crime, are apprehended, and may be adjudicated through the juvenile justice system. Many delinquent adolescents are African American and Latina/o (Teplin, Abram, McClelland, Dulcan, & Mericle, 2002). Many adolescents are at risk for contracting HIV/AIDS because they engage in unprotected sexual behavior and their sexual activity is often associated with substance use. Those adolescents most at risk, including those who are delinquent, are more likely to be sexually active, initiate sexual activity at younger ages, and have more sexual

http://dx.doi.org/10.1037/14940-013 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

273

partners—behaviors that reinforce the need for HIV/AIDS interventions with this population (DiClemente, Lanier, Horan, & Lodico, 1991; Morris et al., 1995; Nader, Wexler, Patterson, McKusick, & Coates, 1989). Adolescents most at risk also exhibit higher levels of substance use and mental health problems (e.g., internalizing behaviors, depression) and are more likely to engage in high-risk behaviors (e.g., have more partners, not use condoms, have sex while using substances, particularly if depressed; Tolou-Shams, Stewart, Fasciano, & Brown, 2010). Interventions that have been successful in preventing substance abuse and mental health problems are a useful starting point for at-risk adolescents (Tolou-Shams et al., 2010). However, prevention efforts that had little effect when targeting those most at risk, delinquent adolescents, on an individual level point to the difficulty in intervening with this population and the need to develop interventions that impact the context in which these adolescents live (Gillmore et al., 1997; Slonim-Nevo, Auslander, Ozawa, & Jung, 1996). Recent efforts that have achieved success suggest that individually focused prevention efforts need to be embedded within broader social contexts, such as the family system, to sustain behavior change (Kumpfer, Alvarado, Tait, & Turner, 2002; Lightfoot & Milburn, 2009, 2012; Pequegnat et al., 2001; Rotheram-Borus, Flannery, Rice, & Lester, 2005). Behavioral family-based interventions are an effective way to intervene to prevent negative developmental outcomes for children and youth (Taylor & Biglan, 1998). Family-based interventions have been used for delinquent (Henggeler, Schoenwald, Borduin, Rowland, & Cunningham, 1998) and other high-risk adolescent populations (Bauman et al., 2002; Catalano, Haggerty, Fleming, Brewer, & Gainey, 2002; Dishion, Kavanagh, Schneiger, Nelson, & Kaufman, 2002; Ennett et al., 2001; Hawkins & Fitzgibbon, 1993; Kumpfer et al., 2002; Rotheram-Borus, 1991; Rotheram-Borus, Goldstein, & Elkavich, 2002; Rotheram-Borus et al., 2001; Spoth, Redmond, Trudeau, & Shin, 2002; Taylor & Biglan, 1998). However, family-based interventions with delinquent adolescents are a “relatively unexplored area in the HIV prevention literature” (Tolou-Shams et al., 2010, p. 258). Families of delinquent youth often have relationships that are characterized by conflict, inappropriate problem solving, and poor communication (Patterson & Stouthamer-Loeber, 1984; Rankin & Kern, 1994; Wasserman, Miller, Pinner, & Jaramillo, 1996). Nevertheless, positive family relationships and positive social connections to family have been linked to reduced risk for HIV/AIDS, including decreased drug use and safer sex practices (Rotheram-Borus, Flannery, Rice, & Lester, 2005) and reduced HIV/AIDS risk behaviors and recidivism (Dodge, Dishion, & Lansford, 2006). Paramount to the success of family-based interventions for behavior change is the active participation of adolescents and their families at every level of the intervention, particularly in enrollment, engagement, and retention. Prior studies have found getting families to participate 274       milburn and lightfoot

to be very challenging. For example, in the aforementioned studies, intervention completion rates have been as low as 25%. Therefore, appropriate and tailored strategies and techniques that engage families need to be identified and employed. This chapter examines the challenges of having families of color1 participate in evidence-based interventions, strategies to address these challenges, lessons that we learned in implementing some of these strategies, and the implications for evidence-based practice and further research. As an exemplar, we focus on families of at-risk adolescents—in this instance, delinquent adolescents—and address the challenges, lessons learned, and implications within the context of evidence-based behavioral interventions for HIV/AIDS risk behaviors including substance use and unsafe sexual practices and behaviors. CHALLENGES ASSOCIATED WITH THE PARTICIPATION OF FAMILIES OF COLOR AND STRATEGIES TO ADDRESS THESE CHALLENGES Evidence-based family interventions seem to be efficacious yet the dilemma of how to get these interventions out into communities to reach families who can benefit from them persists. Although some recent work has emphasized the importance of addressing how to best recruit, engage, and retain participants in interventions to reach more families and have more widespread uptake and dissemination of interventions (Coatsworth, Duncan, Pantin, & Szapocznik, 2006a; Gorman-Smith et al., 2002; Gross, Julion, & Fogg, 2001), overall very little research has been done in this area (Spoth, 2008; Spoth, Redmond, Trudeau, & Shin, 2002). Prevention science has examined the three dimensions of participation: recruitment, engagement, and retention. To date, research usually retrospectively examines only one or two of these dimensions (Coatsworth, Duncan, Pantin, & Szapocznik, 2006b; Gorman-Smith et al., 2002; Gross et al., 2001; Prado, Pantin, Schwartz, Lupei, & Szapocznik, 2006; Spoth et al., 2002; Winslow, Bonds, Wolchik, Sandler, & Braver, 2009). Efforts to increase the participation of families have often focused on implementing family interventions during life transitions when receptivity to change behaviors may be greatest. One expects that families will be most receptive to participating in interventions that enable them to more smoothly move through normal life transitions. However, participation rates are often low even in interventions that focus on normative life transitions for parents and their children such as the birth of a child, starting elementary school, and Families of color refers to families whose members are not White, reflecting a common experience of racism.

1

improving the participation     

275

transitioning from elementary school to middle school (Coatsworth et al., 2006b). In addition, participation is particularly low in family interventions for older children (Spoth et al., 2002). In light of these challenges, greater care must be given to how to enhance participation in family-based interventions with adolescents, but particularly delinquent adolescents because these families are not going through typical, normative life transitions. Characteristics that relate to higher levels of recruitment, retention, and engagement fall into several broad categories: demographics (e.g., race/ ethnicity), child characteristics (e.g., severity of problem behaviors), family characteristics (e.g., stress), and the intervention itself (e.g., project goals, ethnic/racial match of project staff and participants, and incentives; Coatsworth et al., 2006b; Dumas, Begle, French, & Pearl, 2010; Gross et al., 2001). Demographic characteristics that are linked to higher retention rates are higher income and education levels (Coatsworth et al., 2006a). A child characteristic that has been linked to higher retention is higher GPA (Carpentier et al., 2007). Family characteristics that are linked to higher retention rates are having fewer children with problems, having more family stress, having a large family with family members who can babysit, having both parents or the most influential parent consent to participate, being organized, the family’s ability to manage project and work and home schedules, and reporting fewer barriers (e.g., transportation, work schedule; Coatsworth et al., 2006b; Gross et al., 2001; Winslow et al., 2009). Intervention characteristics that are linked to higher retention rates are ethnic/racial match of program staff to families, trust, and opportunity for parental group process (Coatsworth et al., 2006b; Prado et al., 2006). Intervention characteristics are also linked to higher rates of recruitment and retention, specifically, having project goals that are consistent with the goals of the parent or adolescent, or both (Gross et al., 2001; Winslow et al., 2009). The findings on characteristics that are important for linking families to interventions are promising but they have not yet been translated a priori into strategies for enhancing participation in family-based interventions, particularly because many of the characteristics are not malleable. Various strategies for enhancing participation have rarely been examined prospectively even though strategies for recruiting and retaining families in therapy and mental health services, such as motivational interviewing and other techniques, have been studied (Coatsworth et al., 2006a; Dumas et al., 2010; McKay, McCadam, & Gonzales, 1996; McKay, Stoewe, McCadam, & Gonzales, 1998; Spoth et al., 2002). Improving recruitment, engagement, and retention also requires the development of strategies that take cultural variation into consideration. Previous research indicates that racial/ethnic differences in factors are linked to retention, particularly for those factors identified as important for families 276       milburn and lightfoot

in general (e.g., barriers, organization, trust; Coatsworth et al., 2006a). Among Latina/o families, higher educational level and income, being married, having fewer people in the household, having more than one family member in the intervention, reporting fewer barriers (e.g., transportation, work schedule), and being more organized were linked to higher retention. Among African American families, reporting fewer barriers, being organized, and having intentions and motivation to participate in the intervention were linked to higher retention. Therefore, strategies that enable parents to overcome barriers, become more organized, and increase motivation can enhance retention (McKay et al., 1996). For example, addressing logistical barriers (e.g., provide child care, food, or transportation costs; use convenient location) can also increase enrollment, especially for low-income parents (Winslow et al., 2009). This strategy is not new. Acculturation is a factor that is linked to the participation of Latina/o families, but the relationship is nuanced by constructs such as language preference and employment. Carpentier and colleagues (2007) found Mexican American parents who elected Spanish language intervention sessions had higher enrollment and retention in a school-based family intervention than did their counterparts who elected English language intervention sessions. Moreover, parents who elected English language intervention sessions with the most acculturated children were the least likely to enroll and be retained. English language preference parents who worked were more likely to enroll and be retained but the opposite was true for Spanish language preference parents who worked. Trust is often raised as a barrier in the recruitment of African American families, specifically that African Americans mistrust research because of the unethical Tuskegee Study (White, 2005). Empirical findings, however, on the impact of the Tuskegee Study on the participation of African Americans in research are equivocal at best (Crawley, 2001; White, 2005). African Americans trust research when communication between participants and the research team is clear, compelling, and sensitive. This type of communication can improve recruitment (Crawley, 2001; Gwadz et al., 2010). Some argue that race and ethnicity do not matter as much as risk in the recruitment of African American and Latina/o families. For example, Connell and Dishion (2008) found parents whose children were the most at risk were the most likely to enroll in a family-based intervention. However, the intervention and assessment teams in this study were matched to participants on race/ethnicity. The literature on the relationship between the identified characteristics, including cultural variation, and participation suggests strategies that target improving each of the three dimensions of participation: recruitment, engagement, and retention. These strategies address many of the contextual improving the participation     

277

barriers that parents struggle with. Because many of the demographic, child, and family characteristics that are associated with recruitment, engagement, and retention cannot be addressed or changed, targeting the characteristics of the intervention suggests strategies for improving participation. To date, few of these strategies have been refined and tried at the beginning of interventions so we do not know what will work and at what cost (e.g., is the investment worth the outcome?). To date, the only strategy that has been tested is the use of incentives. Dumas and colleagues (2010) found incentives helped with recruitment but not retention for low-income and less educated African American parents. Those who were told they would receive an incentive to attend sessions were more likely to enroll in the intervention than were their counterparts. However, the incentive did not increase the likelihood of actually attending sessions. Guyll, Spoth, and Redmond (2003) found an incentive ($100) increased enrollment in a research project for parents with less education and who reported they were less willing to participate in research. These findings were less robust for enrollment in the actual intervention component of the research project; incentives improved enrollment in the assessments more than enrollment in the intervention. Providing incentives improves enrollment but not retention and may not be feasible or sustainable for broad dissemination of family-based interventions. PILOT OF A STRATEGY FOR INCREASING PARTICIPATION Family interventions can have a major public health impact on reducing substance abuse and conduct problems in high-risk adolescent populations such as delinquent adolescents (Spoth, 2008). However, as noted, parent involvement in interventions is costly and can be difficult to achieve. Participation rates of 50% seem to be common across prevention interventions (Winslow et al., 2009); rates for family interventions are even lower (e.g., 25%; Stormshak, Connell, & Dishion, 2009). Clearly, better participation is needed: “To increase the public health significance of prevention parenting programs, prevention scientists need to identify predictors of participation and develop and test theory-based strategies to increase participation” (Winslow et al., 2009, p. 151). Improving recruitment, engagement, and retention can improve the outcomes of family-based interventions (Gorman-Smith et al., 2002). We developed and piloted a strategy to target each participation domain—recruitment, engagement, and retention in the implementation and delivery of an evidence-based family-based intervention with a sample of delinquent African American adolescents and their parents or guardians (R21 DA024955; principal investigator: Milburn; coinvestigator: Lightfoot). 278       milburn and lightfoot

The evidence-based intervention was “Families That Care: Guiding Good Choices” (GGC, also known as “Preparing for the Drug Free Years”; Haggerty, Kosterman, Catalano, & Hawkins, 1999), a brief five-session, family-based intervention that uses a skill-building approach to improving the family’s capacity that does not blame family members for having problems. GGC addresses family conflict, problem solving, and communication to increase family connectedness and decrease drug use (Gorman-Smith et al., 2002; Gross et al., 2001). The developed strategy aimed to increase rates of recruitment, levels of, and retention rates in both the intervention sessions and the assessments. The strategy included the use of a motivational DVD to recruit parents or guardians, a motivational recruitment session, and a presession before the intervention for families to share who they are and articulate their family goals. The strategy was in addition to established procedures used in previous research to increase engagement and retention, such as providing transportation and scheduling sessions to fit the family’s schedules. In light of previous findings with African American families, the developed strategy sought to increase intentions and motivations to participate in the intervention as well as maximize the intervention characteristics found to be important (i.e., racial/ethnic match of program staff and families, creating opportunities for the parents to process, and identifying how the intervention’s goals were consistent with the family’s goals). The inspirational, motivational DVD focused on increasing families’ feelings of competence to address their adolescents’ problem behaviors, normalizing family difficulties, encouraging families to identify opportunities for growth and recognize that the intervention could be helpful, reinforcing the families’ commitment to the ultimate goal of seeing their youth be successful (e.g., graduate from high school), and highlighting the value of persistence and perseverance. The content of the DVD evolved from qualitative analysis of key informant interviews and focus groups with school staff (e.g., principals, teachers, probation officers), parents, and adolescents. We also conducted formative marketing research to tailor GGC to maximize engagement in the intervention by the families. The qualitative analysis of key informant interviews and focus groups suggested some cultural tailoring of GGC was necessary. For example, we learned many of these youth reside in high-risk neighborhoods and are exposed to drug use. Many parents knew about the myriad of problems associated with drug use (e.g., gang affiliations, juvenile justice involvement). Some parents had a history of substance abuse problems themselves. Consequently, some changes were made in the first session. Sections on the prevalence of drug use and community, school, family, and peer and individual risk factors were shortened in light of the experiences of these adolescents and their parents. Also, an alternative model was used to improving the participation     

279

explain the social development strategy that underlies GGC to parents (e.g., used image of child enveloped in protective layers). The training for GGC led us to further modify the initial strategy. We did not need the presession for families to tell their stories (e.g., feel heard) and articulate their current family goals because the content of this presession was built into the first session of GGC, which provided time for parents to share their hopes and dreams for their adolescent child and describe their lives, where they were now and their family goals, and how GGC’s goals could match their family goals. Preliminary results from the pilot test of the strategy (N = 27) suggested the strategy improved recruitment (100% completed baseline in strategy vs. 71% in nonstrategy) and retention (100% 3 months follow-up completion rate in the strategy vs. 50% in nonstrategy). The strategy did not improve engagement, defined as attendance of at least one session (38% vs. 40%) and the average number of sessions attended (3 vs. 4) compared with the nonstrategy. LESSONS LEARNED We learned a number of things from this preliminary work. One, the family-based intervention should have as few face-to-face sessions as possible. As a result, we integrated the presession content into the first intervention’s session, eliminating the need for a face-to-face presession for families to tell their stories. The first session of GGC was expanded to allow parents to share their hopes and dreams for their youth and describe their lives, where they were now and their family goals, and how GGC’s goals could match their family goals. This modification, however, reflected only the parent perspective. The perspective of the adolescent was not included in the intervention because the adolescent does not attend the first session of GGC. However, adolescents and parents expressed an interest in hearing the adolescent’s perspective at the beginning of the intervention. Clearly, adolescent perspectives need to be included but how remains to be determined. Second, we targeted African American youth who were attending alternative education schools, known locally as probation schools. However, schools were diverse, with increasing numbers of Latina/o students. There is a critical need to also engage, recruit, and retain Latina/o adolescents and their parents in family-based interventions. The schools were gracious to allow us to just target African American students because of a limited project budget but it was very clear as the pilot work evolved that Latina/o students were interested in being involved (and have been involved in our work in this area; e.g., University of California at Los Angeles Center for HIV Identification, Prevention, and Treatment Services Development Pilot 2009; principal investigator: Milburn). Third, 280       milburn and lightfoot

we found that GGC should be delivered at the school sites (not in homes as we had originally planned, to make the intervention more accessible to families) so that parents could participate in a group for mutual support to learn and practice the parenting techniques (e.g., family meeting). Finally, we also found that although parents selected a time that was best for the intervention sessions, the parents’ schedules required flexibility by the research staff with multiple times being needed on a weekly basis (e.g., morning, during lunch hours, evening, and Saturday morning). IMPLICATIONS FOR EVIDENCE-BASED PRACTICE WITH FAMILIES OF COLOR A number of family interventions for delinquent youth have been empirically tested, including Multisystemic Therapy, Functional Family Therapy, Family Integrated Transitions (Underwood, Phillips, von Dresner, & Knight, 2006), and Brief Strategic Family Therapy (Szapocznik, Hervis, & Schwartz, 2003). The primary goal of these approaches is to improve parenting skills; however, they also encourage disassociation from negative peers and help the family form a neighborhood support network of friends, neighbors, and family members (Aos, 2004; Robbins & Szapocznik, 2000; Sexton & Alexander, 2000). The impact of these interventions is limited by their ability to recruit, engage, and retain families into the interventions. High recruitment, engagement, and retention are necessary to garner a sufficient sample size to detect intervention effects, increase intervention dosage, and increase statistical power (Spoth, 2008). Therefore, high recruitment and retention are needed to adequately evaluate an intervention. High engagement is likely to increase study retention and the family’s integration of the intervention’s concept and skills. Consequently, the effect of engagement is also evident in obtaining the desired outcomes. Being highly engaged will lead to the intervention having a greater impact and better outcomes (Gorman-Smith et al., 2002). It is clear that research on enhancing participation in family-based interventions is needed to advance prevention science. To date, research usually retrospectively examines one or two dimensions of participation (Hurd, Valerio, Garcia, & Scott, 2010; Kumpfer & Tait, 2000), but not all three: recruitment, engagement, and retention. These studies typically examine the relationship between the characteristics of the families (e.g., family size, scheduling barriers) and one dimension of participation. Though informative, these studies typically identify family, parent, or child characteristics that cannot be modified for the intervention (e.g., parent education level). Furthermore, intervention studies typically incorporate protocols to address many of the logistical barriers to a family’s participant that improving the participation     

281

are malleable (e.g., transportation). To advance the field, movement beyond simply providing transportation and child care is required. Overcoming contextual barriers to participation must be addressed. Our findings suggest that a focus on incorporating contextual barriers into the intervention has promise. Our participation strategy sought to increase intentions and motivations to participate in the intervention study as well as maximize the intervention characteristics found to be important (i.e., racial/ethnic match of program staff and families, creating opportunities for the parents to process, and identifying how the intervention’s goals were consistent with the family’s goals). By incorporating these important characteristics directly into the intervention, recruitment and retention were improved beyond what was found for families who received standard logistical support. These findings suggest that interventions need to not only include skill-building exercises but also allow the family time and space to process the benefits of the intervention and how it meets the goals of the family. The developed strategy did not significantly improve engagement in the intervention. Though we conducted formative research to culturally tailor the intervention, the changes were insufficient to facilitate engagement. It is important to note that the developed strategy was adjunctive to an existing evidence-based intervention, GGC. As such, these findings offer promise for improving not only the development of new interventions but also participation in other existing evidence-based interventions. Additional research is needed to further identify and test the important intervention characteristics that could improve the recruitment, engagement, and retention of families into interventions. REFERENCES Anderson, R. N., & Smith, B. L. (2005). Deaths: Leading causes for 2002. National Vital Statistics Reports, 53(17), 1–89. Aos, S. (2004). Washington State’s Family Integrated Transitions Program for Juvenile Offenders: Outcome evaluation and benefit–cost analysis (Report ID: 04-12-1201). Olympia: Washington State Institute for Public Policy. Bauman, K. E., Ennett, S. T., Foshee, V. A., Pemberton, M., King, T. S., & Koch, G. G. (2002). Influence of a family program on adolescent smoking and drinking prevalence. Prevention Science, 3, 35–42. http://dx.doi.org/ 10.1023/A:1014619325968 Benda, B. B., Flynn Corwyn, R., & Toombs, N. J. (2001). Recidivism among adolescent serious offenders: Prediction of entry into the correctional system for adults. Criminal Justice and Behavior, 28, 588–613. http://dx.doi.org/10.1177/ 009385480102800503

282       milburn and lightfoot

Carpentier, F. R., Mauricio, A. M., Gonzales, N. A., Millsap, R. E., Meza, C. M., Dumka, L. E., . . . Genalo, M. T. (2007). Engaging Mexican origin families in a school-based preventive intervention. The Journal of Primary Prevention, 28, 521–546. http://dx.doi.org/10.1007/s10935-007-0110-z Catalano, R. F., Haggerty, K. P., Fleming, C. B., Brewer, D. D., & Gainey, R. R. (2002). Children of substance abusing parents: Current findings from the Focus on Families project. In R. J. McMahon & R. D. V. Peters (Eds.), The effects of parental dysfunction on children (pp. 179–204). http://dx.doi. org/10.1007/978-1-4615-1739-9_9 Centers for Disease Control and Prevention. (2005). HIV/AIDS Surveillance Report, 2004 (Vol. 16, pp. 1–46). Atlanta, GA: U.S. Department of Health and Human Services. Coatsworth, J. D., Duncan, L. G., Pantin, H., & Szapocznik, J. (2006a). Differential predictors of African American and Hispanic parent retention in a familyfocused preventive intervention. Family Relations, 55, 240–251. http://dx.doi. org/10.1111/j.1741-3729.2006.00373.x Coatsworth, J. D., Duncan, L. G., Pantin, H., & Szapocznik, J. (2006b). Patterns of retention in a preventive intervention with ethnic minority families. The Journal of Primary Prevention, 27, 171–193. http://dx.doi.org/10.1007/s10935-005-0028-2 Connell, A. M., & Dishion, T. J. (2008). Reducing depression among at-risk early adolescents: Three-year effects of a family-centered intervention embedded within schools. Journal of Family Psychology, 22, 574–585. http://dx.doi.org/10.1037/ 0893-3200.22.3.574 Crawley, L. M. (2001). African-American participation in clinical trials: Situating trust and trustworthiness. JAMA, 93(12, Suppl.), 14S–17S. DiClemente, R. J., Lanier, M. M., Horan, P. F., & Lodico, M. (1991). Comparison of AIDS knowledge, attitudes, and behaviors among incarcerated adolescents and a public school sample in San Francisco. American Journal of Public Health, 81, 628–630. http://dx.doi.org/10.2105/AJPH.81.5.628 Dishion, T. J., Kavanagh, K., Schneiger, A., Nelson, S., & Kaufman, N. K. (2002). Preventing early adolescent substance use: A family-centered strategy for the public middle school. Prevention Science, 3, 191–201. http://dx.doi. org/10.1023/A:1019994500301 Dodge, K. A., Dishion, T. J., & Lansford, J. E. (2006). Deviant peer influences in intervention and public policy for youth. Social Policy Report, XX(I), 1–19. Dumas, J. E., Begle, A. M., French, B., & Pearl, A. (2010). Effects of monetary incentives on engagement in the PACE parenting program. Journal of Clinical Child and Adolescent Psychology, 39, 302–313. http://dx.doi.org/10.1080/ 15374411003691792 Ennett, S. T., Bauman, K. E., Pemberton, M., Foshee, V. A., Chuang, Y. C., King, T. S., & Koch, G. G. (2001). Mediation in a family-directed program for prevention of adolescent tobacco and alcohol use. Preventive Medicine, 33, 333–346. http://dx.doi.org/10.1006/pmed.2001.0892 improving the participation     

283

Flores, G., Fuentes-Afflick, E., Barbot, O., Carter-Pokras, O., Claudio, L., Lara, M., . . . Weitzman, M. (2002). The health of Latino children: Urgent priorities, unanswered questions, and a research agenda. JAMA, 288, 82–90. http://dx.doi.org/10.1001/ jama.288.1.82 Gillmore, M. R., Morrison, D. M., Richey, C. A., Balassone, M. L., Gutierrez, L., & Farris, M. (1997). Effects of a skill-based intervention to encourage condom use among high risk heterosexually active adolescents. AIDS Education and Prevention, 9(Suppl.), 22–43. Gorman-Smith, D., Tolan, P. H., Henry, D. B., Leventhal, A., Schoeny, M., Lutovsky, K., & Quintana, E. (2002). Predictors of participation in a family-focused preventive intervention for substance use. Psychology of Addictive Behaviors, 16(Suppl.), S55–S64. http://dx.doi.org/10.1037/0893-164X.16.4S.S55 Gross, D., Julion, W., & Fogg, L. (2001). What motivates participation and dropout among low-income urban families of color in a prevention intervention? Family Relations, 50, 246–254. http://dx.doi.org/10.1111/j.1741-3729.2001.00246.x Guyll, M., Spoth, R., & Redmond, C. (2003). The effects of incentives and research requirements on participation rates for a community-based preventive intervention research study. The Journal of Primary Prevention, 24, 25–41. http://dx.doi.org/ 10.1023/A:1025023600517 Gwadz, M. V., Colon, P., Ritchie, A. S., Leonard, N. R., Cleland, C. M., Riedel, M., . . . Mildvan, D. (2010). Increasing and supporting the participation of persons of color living with HIV/AIDS in AIDS clinical trials. Current HIV/ AIDS Report, 7, 194–200. Haggerty, K., Kosterman, R., Catalano, R. F., & Hawkins, J. D. (1999). Preparing for the drug free years. Retrieved from http://www.ncjrs.gov/pdffiles1/173408.pdf Hawkins, J. D., & Fitzgibbon, J. J. (1993). Risk factors and risk behaviors in prevention of adolescent substance abuse. Adolescent Medicine, 4, 249–262. Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland, M. D., & Cunningham, P. B. (1998). Multisystemic therapy for antisocial behavior in children and adolescents (2nd ed.). New York, NY: Guilford Press. Hurd, N. M., Valerio, M. A., Garcia, N. M., & Scott, A. A. (2010). Adapting an HIV prevention intervention for high-risk, incarcerated adolescents. Health Education & Behavior, 37, 37–50. http://dx.doi.org/10.1177/1090198109335655 Kumpfer, K. L., Alvarado, R., Tait, C., & Turner, C. (2002). Effectiveness of schoolbased family and children’s skills training for substance abuse prevention among 6–8-year-old rural children. Psychology of Addictive Behaviors, 16(Suppl.), S65–S71. http://dx.doi.org/10.1037/0893-164X.16.4S.S65 Kumpfer, K. L., & Tait, C. M. (2000, April). Family skills training for parents and children. OJJDP Juvenile Justice Bulletin (NCJ 180140). Retrieved from https:// www.ncjrs.gov/pdffiles1/ojjdp/180140.pdf Lightfoot, M., & Milburn, N. G. (2009). HIV prevention and African American youth: Examination of individual-level behavior is not the only answer. Culture, Health and Sexuality, 11, 731–742.

284       milburn and lightfoot

Lightfoot, M., & Milburn, N. G. (2012). Common elements of family-based HIV interventions for adolescents. Couple and Family Psychology: Research and Practice, 1, 120–133. http://dx.doi.org/10.1037/a0028832 McKay, M. M., McCadam, K., & Gonzales, J. J. (1996). Addressing the barriers to mental health services for inner city children and their caretakers. Community Mental Health Journal, 32, 353–361. http://dx.doi.org/10.1007/BF02249453 McKay, M. M., Stoewe, J., McCadam, K., & Gonzales, J. (1998). Increasing access to child mental health services for urban children and their caregivers. Health & Social Work, 23, 9–15. http://dx.doi.org/10.1093/hsw/23.1.9 Morris, R. E., Harrison, E. A., Knox, G. W., Tromanhauser, E., Marquis, D. K., & Watts, L. L. (1995). Health risk behavioral survey from 39 juvenile correctional facilities in the United States. Journal of Adolescent Health, 17, 334–344. http:// dx.doi.org/10.1016/1054-139X(95)00098-D Nader, P. R., Wexler, D. B., Patterson, T. L., McKusick, L., & Coates, T. (1989). Comparison of beliefs about AIDS among urban, suburban, incarcerated, and gay adolescents. Journal of Adolescent Health Care, 10, 413–418. http://dx.doi. org/10.1016/0197-0070(89)90221-0 Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of family management practices and delinquency. Child Development, 55, 1299–1307. http:// dx.doi.org/10.2307/1129999 Pequegnat, W., Bauman, L. J., Bray, J. H., DiClemente, R., Dilorio, C., Hoppe, S. K., . . . Szapocznik, J. (2001). Measures relevant to the role of families in prevention and adaptation to HIV/AIDS. AIDS and Behavior, 5, 1–19. http://dx.doi.org/ 10.1023/A:1009557103659 Prado, G., Pantin, H., Schwartz, S. J., Lupei, N. S., & Szapocznik, J. (2006). Predictors of engagement and retention into a parent-centered, ecodevelopmental HIV preventive intervention for Hispanic adolescents and their families. Journal of Pediatric Psychology, 31, 874–890. http://dx.doi.org/10.1093/jpepsy/jsj046 Rankin, J., & Kern, R. (1994). Parental attachments and delinquency. Criminology, 32, 495–515. http://dx.doi.org/10.1111/j.1745-9125.1994.tb01163.x Robbins, M. S., & Szapocznik, J. (2000, April). Brief Strategic Family Therapy. Juvenile Justice Bulletin (NCJ 179285). Retrieved from https://www.ncjrs.gov/pdffiles1/ ojjdp/179285.pdf Romero, E. G., Teplin, L. A., McClelland, G. M., Abram, K. M., Welty, L. J., & Washburn, J. J. (2007). A longitudinal study of the prevalence, development, and persistence of HIV/sexually transmitted infection risk behaviors in delinquent youth: Implications for health care in the community. Pediatrics, 119, e1126–e1141. http://dx.doi.org/10.1542/peds.2006-0128 Rotheram-Borus, M. J. (1991). Serving runaway and homeless youths. Family & Community Health, 14, 23–32. Rotheram-Borus, M. J., Flannery, D., Rice, E., & Lester, P. (2005). Families living with HIV. AIDS Care, 17, 978–987. http://dx.doi.org/10.1080/09540120500101690 improving the participation     

285

Rotheram-Borus, M. J., Goldstein, A. M., & Elkavich, A. S. (2002). Treatment of suicidality: A family intervention for adolescent suicide attempters. In S. G. Hofmann & M. C. Tompson (Eds.), Treating chronic and severe mental disorders: A handbook of empirically supported interventions (pp. 191–212). New York, NY: Guilford Press. Rotheram-Borus, M. J., Lee, M. B., Murphy, D. A., Futterman, D., Duan, N., Birnbaum, J. M., . . . Teens Linked to Care Consortium. (2001). Efficacy of a preventive intervention for youths living with HIV. American Journal of Public Health, 91, 400–405. http://dx.doi.org/10.2105/AJPH.91.3.400 Sexton, T. L., & Alexander, J. F. (2000, December). Functional family therapy. OJJDP Juvenile Justice Bulletin (NCJ 184743). Retrieved from https://www.ncjrs. gov/pdffiles1/ojjdp/184743.pdf Slonim-Nevo, V., Auslander, W. F., Ozawa, M. N., & Jung, K. G. (1996). The longterm impact of AIDS-preventive interventions for delinquent and abused adolescents. Adolescence, 31(122), 409–421. Spoth, R. L. (2008). Translating family-focused prevention science into effective practice: Toward a translational impact paradigm. Current Directions in Psychological Science, 17, 415–421. http://dx.doi.org/10.1111/j.1467-8721.2008.00617.x Spoth, R. L., Redmond, C., Trudeau, L., & Shin, C. (2002). Longitudinal substance initiation outcomes for a universal preventive intervention combining family and school programs. Psychology of Addictive Behaviors, 16, 129–134. http:// dx.doi.org/10.1037/0893-164X.16.2.129 Stormshak, E. A., Connell, A., & Dishion, T. J. (2009). An adaptive approach to familycentered intervention in schools: Linking intervention engagement to academic outcomes in middle and high school. Prevention Science, 10, 221–235. http:// dx.doi.org/10.1007/s11121-009-0131-3 Szapocznik, J., Hervis, O., & Schwartz, S. (2003). Brief strategic family therapy for adolescent drug abuse (NIH Publication No. 03-4751). Bethesda, MD: National Institute on Drug Abuse. http://dx.doi.org/10.1037/e598162007-001 Taylor, T. K., & Biglan, A. (1998). Behavioral family interventions for improving child-rearing: A review of the literature for clinicians and policy makers. Clinical Child and Family Psychology Review, 1, 41–60. http://dx.doi.org/10.1023/ A:1021848315541 Teplin, L. A., Abram, K. M., McClelland, G. M., Dulcan, M. K., & Mericle, A. A. (2002). Psychiatric disorders in youth in juvenile detention. Archives of General Psychiatry, 59, 1133–1143. http://dx.doi.org/10.1001/archpsyc.59.12.1133 Tolou-Shams, M., Stewart, A., Fasciano, J., & Brown, L. K. (2010). A review of HIV prevention interventions for juvenile offenders. Journal of Pediatric Psychology, 35, 250–261. http://dx.doi.org/10.1093/jpepsy/jsp069 Underwood, L. A., Phillips, A., von Dresner, K., & Knight, P. D. (2006). Critical factors in mental health programming for juveniles in corrections facilities. International Journal of Behavioral Consultation and Therapy, 2, 107–140. http:// dx.doi.org/10.1037/h0100771

286       milburn and lightfoot

Wasserman, G. A., Miller, L. S., Pinner, E., & Jaramillo, B. (1996). Parenting predictors of early conduct problems in urban, high-risk boys. Journal of the American Academy of Child & Adolescent Psychiatry, 35, 1227–1236. http://dx.doi. org/10.1097/00004583-199609000-00020 White, R. (2005). Misinformation and misbeliefs in the Tuskegee Study of Untreated Syphilis fuel mistrust in the healthcare system. JAMA, 97, 1567–1573. Winslow, E. B., Bonds, D., Wolchik, S., Sandler, I., & Braver, S. (2009). Predictors of enrollment and retention in a preventive parenting intervention for divorced families. The Journal of Primary Prevention, 30, 151–172. http://dx.doi. org/10.1007/s10935-009-0170-3

improving the participation     

287

13 CULTURALLY ADAPTING EVIDENCE-BASED PRACTICES FOR ETHNIC MINORITY AND IMMIGRANT FAMILIES WEI-CHIN HWANG

Historically, psychotherapy was developed from a primarily Western and European perspective. Because great diversity exists internationally as well as within the United States, finding a way to culturally adapt psychotherapy to improve treatment engagement and outcomes is of critical importance (Hwang, 2006b). Research demonstrates that ethnic minorities are less likely to receive quality health services and evidence worse treatment outcomes when compared with European Americans (Institute of Medicine, 1999; U.S. Department of Health and Human Services [USDHHS], 2001). Moreover, relatively little is known about the efficacy of evidence-based psychological practices for ethnic minorities and immigrants, as opposed to European Americans, regarding whom considerable progress has been made in establishing and defining effective and possibly efficacious treatments (Bernal & Scharró-del-Río, 2001; Hall, 2001; USDHHS, 2001).

http://dx.doi.org/10.1037/14940-014 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

289

Unfortunately, few models have been developed to help clinical researchers and practitioners systematically adapt psychological treatments for immigrants and other diverse clientele. The field needs to make a shift from what Resnicow, Baranowski, Ahluwalia, and Braithwaite (1999) called surface structure adaptations (e.g., providing ethnically matched therapists, conducting therapy in a client’s native language, designing clinics to be culturally aesthetic, locating clinics in neighborhoods that are easily accessible) to deep structure adaptations (e.g., incorporating the ideas, beliefs, and values of the culture into the treatment). Deep structure adaptations are much more difficult to understand, conceptualize, and implement. Nevertheless, deep structure adaptations have more potential for improving outcomes and tailoring treatment to match the client’s background. Currently, few frameworks that guide deep structural adaptations exist. One of the first approaches was developed by Bernal, Bonilla, and Bellido (1995) and focused on eight dimensions in which adaptation can take place: language, persons, metaphors, content, file concepts, goals, methods, and context. More recently, Hwang (2006b) developed the top-down and theorydriven psychotherapy adaptation and modification framework (PAMF) and later integrated it with the community-based and bottom-up formative method for adapting psychotherapy (FMAP; Hwang, 2009). The purpose of this chapter is to discuss how the integrative PAMF and FMAP approaches can be used to culturally adapt psychotherapy for diverse populations. USING AN INTEGRATIVE FRAMEWORK FOR CULTURALLY ADAPTING PSYCHOTHERAPY The FMAP framework is a bottom-up, participatory approach for culturally adapting psychotherapy (Hwang, 2009). It can be used to modify evidencebased practices as well as empirically supported treatments. This framework was used to create a culturally adapted cognitive–behavioral therapy (CBT) manual for depressed Chinese Americans, a project funded by a National Institute of Mental Health R34 treatment development and clinical trial grant (principal investigator: Hwang). Although the treatment was developed for individual psychotherapy for depression, many family issues arose (e.g., family and couples conflict, reductions in support networks and family support, intergenerational family problems, family expectations and pressures, acculturative family issues) and were addressed in the intervention. The culturally adapted CBT manual has been tested against a non–culturally adapted CBT manual at two Asian American–focused community mental health clinics (Asian Pacific Family Center in Los Angeles and Richmond Area Multi-Services in San Francisco). Results are currently being analyzed and will be discussed in forthcoming papers. 290       wei-chin hwang

The FMAP approach consists of five phases: (a) generating knowledge and collaborating with stakeholders, (b) integrating generated information with theory and empirical and clinical knowledge, (c) reviewing the initial culturally adapted clinical intervention with stakeholders and revising the culturally adapted intervention, (d) testing the culturally adapted intervention, and (e) finalizing the culturally adapted intervention. This iterative approach provides the opportunity to revise and adjust to feedback provided by consumers (e.g., therapist and clients). In Phase I, an emphasis is placed on generating knowledge and collaborating with consumers or stakeholders (e.g., clients, administrators, therapists). In this particular project, five types of stakeholders were involved: (a) Asian-focused community mental health agencies (i.e., clinics that self-designate as specializing in serving predominantly Asian American populations), (b) mental health providers (psychiatrists, psychologists, social workers, marital family therapists), (c) traditional Chinese medicine (TCM) practitioners, (d) Buddhist monks and nuns, and (e) both spiritual and religious Taoist masters. Collaborating with community experts in the field as well as indigenous healers can provide valuable feedback for enhancing treatment. Although seeking client feedback is an important part of treatment development, an informed decision was made to do this in later phases (IV and V) because many of the clients had little to no exposure to mental health services, were unlikely to be able to differentiate different types of psychological treatment, and could potentially lose confidence in the treatment if project staff asked for their advice. Collaborations were made with mental health clinics to ensure that the intervention developed would be ecologically valid and could be feasibly implemented in real-world settings (e.g., the frequency of sessions, staffing and assignment of caseloads, hours of operation, billing and financial limitations). Agencies involved (all in California) included Asian Americans for Community Involvement in San Jose, Asian Community Mental Health Services in Oakland, Asian Pacific Counseling and Treatment Center in Los Angeles, Asian Pacific Mental Health Services in Gardena, and Chinatown North Beach Service Center in San Francisco. Several clinics were involved to ensure that clinic biases and a range of clinic beliefs, assumptions, and notions of best practice would be included. Seven Asian-focused clinics, two of which served as primary clinical trial sites and five as focus-group collaborators, were incorporated. Fourteen focus groups were conducted at community mental health clinics located in California, with multiple focus groups conducted at larger clinics. Focus groups helped reduce clinician- and agency-specific biases. Mental health care providers working in the agencies were asked to participate because they were experts in the field and had insights and expertise in working with depressed Asian American clients. Focus groups were not held culturally adapting evidence-based practices     

291

at clinical trial sites to ensure that the treatment conditions for a subsequent clinical trial at these clinics would not be compromised. Another advantage of having practitioners participate in the development of the treatment was to facilitate buy-in to the treatment because they would be the ones to potentially use it once it was developed. Each focus group consisted of four to six mental health practitioners with a range of clinical experiences, which helped facilitate both breadth and depth of discussions. Two sets of focus groups were conducted. The first 4 hours involved general discussions of cultural adaptation, their experiences working with Asian American clients, and review of an evidence-based treatment manual (Miranda et al., 2006). The second set of focus groups was conducted in Phase III and is described further below. Because indigenous medicines and religions have strongly influenced Chinese culture for thousands of years, interviews were also conducted with several Buddhist monks and nuns, spiritual and religious Taoist masters, and TCM practitioners. Collaborating with traditional healers helped ensure that cultural adaptations were grounded in client belief systems and helped generate ideas of how Western mental health treatment and traditional medicines can work together to optimize outcomes. In Phase II, information generated from Phase I was integrated with extant empirical and clinical knowledge consolidated in the PAMF (Hwang, 2006b). An iterative top-down and bottom-up approach to treatment development can greatly improve the validity and effectiveness of adaptations created. The PAMF is a three-tiered approach to therapy adaptations (i.e., domains, principles, and rationales) and was developed to help identify areas that can be targeted for adaptation while providing rationales for increasing the credibility of said modifications. It was developed to help practitioners make the shift from the more abstract notion of being culturally competent to develop concrete and specific skills to more effectively work with diverse clientele. Domains of adaptation include (a) understanding dynamic issues and cultural complexities; (b) orienting clients to psychotherapy and increasing mental health awareness; (c) understanding cultural beliefs about mental illness, its causes, and what constitutes appropriate treatment; (d) improving the client–therapist relationship; (e) understanding cultural differences in the expression and communication of distress; and (f) addressing cultural issues specific to the population. Principles are specific recommendations for adapting therapy for specific groups, and rationales are corresponding explanations for why these adaptations may be effective. Specific principles and rationales are detailed more fully elsewhere (Hwang, 2006b; Hwang, Wood, Lin, & Cheung, 2006). The third phase of the FMAP involves having consumers and stakeholders review the intervention for further revision. In this particular case, the new treatment manual was taken back to the same therapists who 292       wei-chin hwang

participated in the focus groups and another set of 4-hour focus groups was conducted (Hwang, 2008b, 2008c). Discussions focused on initial impressions of the new intervention and identifying areas for improvement. Therapists were very excited about the new intervention and believed that the manual would be effective in treating depressed Chinese Americans. This phase was largely confirmatory and the discussion focused primarily on wording changes to help ensure ideas, words, and concepts would translate properly into different Chinese languages. To help strengthen language equivalence, the manual was written with translations in Chinese in mind. The intervention was translated and back-translated by a team of four master’s-level therapists, one postdoctoral fellow, and the author. In addition, feedback from 15 undergraduate students, three master’s-level therapists, one postdoctoral fellow, and four graduate students was elicited. Because there are regional differences in written Chinese, translated materials were reviewed by people from different Chinese regions (e.g., Mainland China, Taiwan, and Hong Kong) to ensure cross-region comprehensibility. The fourth phase of the FMAP involves testing the intervention to make sure that it is ecologically effective. The intervention was a 12-session depression treatment program for Chinese Americans (Hwang, 2008b, 2008c). A randomized controlled trial was used to test its effectiveness and assessments were conducted at baseline and sessions 4, 8, and 12 of treatment, as well as 3 months posttreatment. In addition, various clinical outcome measures from clients, therapists, and independent assessors were used to assess symptom reduction, treatment satisfaction, premature dropout, working alliance formation, and attitudes toward psychotherapy. Therapy sessions were recorded so that they could be coded for treatment fidelity (adherence, receipt, and enactment of treatment), and therapists and clients were also asked to self-report on treatment adherence. Weekly group supervision was provided to the therapist in both treatment conditions by the author of the new manual, because the author also had extensive experience with the standard CBT manual. Although having one supervisor for both conditions could potentially have led to allegiance biases, the use of different supervisors could also have led to a supervisor effect that would have been impossible to disaggregate. A choice was made to have one supervisor because supervision information gathered from supervising both conditions can provide valuable information to further refine the treatment. The fifth phase of the FMAP involves gathering additional information after the clinical trial is over (e.g., conducting interviews with clients and therapists about their experiences participating in the treatment program). Information gathering is an important aspect of consumer-based participatory and iterative approach to treatment refinement. The culmination of knowledge acquired through the different phases can be used to further refine and culturally adapting evidence-based practices     

293

improve the culturally adapted treatment. Examples of cultural adaptations are highlighted below and organized according to the domains of the PAMF. Moreover, it is important to emphasize that although this project focused on adapting psychotherapy for depressed Chinese Americans, the information generated can improve the field’s understanding of psychological science and clinical practice. In other words, just as basic psychological science informs clinical science and practice, cultural psychology as a science strengthens clinical science and can potentially lead to improved care for all. Understanding Diversity Among Ethnic Minority and Immigrant Families The first domain of the PAMF highlights the importance of what S. Sue (1998) called dynamic sizing (e.g., understanding stereotypes and how to use this information to individualize the treatment for the client) and what Hays (2001) called cultural complexities (e.g., focusing on more than just a person’s ethnicity and recognizing different aspects of diversity, such as gender, age, immigration status, and sexual orientation). These two concepts are particularly important when working with ethnic minority and immigrant families. For example, those who have been here for several generations experience very different stressors than do families of new immigrants. In addition, country of origin (e.g., how modernized the country is and whether English was spoken in that country) and circumstances of migration (e.g., refugee status, working migrants, educated or working-class professionals, and parachute children experiences) can also greatly change the difficulty of and desire to achieve acculturation or assimilation. The ethnic density and location where one immigrates can also affect the acculturation, assimilation, or enculturation process. For example, the pressure to assimilate is much greater for families who move to less diverse locations such as Salt Lake City versus areas with expansive ethnic neighborhoods such as Los Angeles. In addition, though in the past there was an unspoken rule that immigrants would eventually assimilate into mainstream America, a large body of research suggests that acculturation is deleterious to various aspects of health (e.g., depression, anxiety, substance abuse, various physical health problems; Escobar & Vega, 2000; Kessler et al., 1994). Unfortunately, some research also suggests that immigrant families experience greater intergenerational family conflict because of acculturation and family issues, which exacerbates the risk for mental illness (Hwang, 2006a; Hwang & Wood, 2009; Hwang, Wood, & Fujimoto, 2010). Rather than pushing immigrants to assimilate and give up their cultural values, which tends to be a point of contention among immigrant parents and their children, a better emphasis may be to help families become biculturally competent to reduce 294       wei-chin hwang

conflict and retain culturally protective factors (LaFromboise, Coleman, & Gerton, 1993). People from various backgrounds have different stereotypes toward, understanding of, and willingness to seek mental health treatment. Those who are less acculturated have greater misconceptions of mental illness and perceive those as mentally ill as being “crazy” (Hwang, 2006b). As a result, there is greater stigma among immigrants and help-seeking rates for psychiatric issues tend to be lower (Bui & Takeuchi, 1992; Flaskerud & Liu, 1991; Hu, Snowden, Jerrell, & Nguyen, 1991). A greater emphasis placed on individual and community stigma reduction may help demystify mental illness and its treatment (Yang, Phelan, & Link, 2008). At a therapeutic level, therapists often struggle with how to make use of stereotypical information that they hear about or read in scholarly articles. For example, what do therapists do with the knowledge that family is very important to Asian Americans and Latinos who come from collectivistic cultural backgrounds? When working with people from different backgrounds, therapists tend to feel an implicit pressure to try to emotionally connect with their clients by showing that they know something about their clients’ culture and understand what is important to them or by making statements that highlight their commonality with the client (e.g., “I have many friends from your same ethnic background” or “I have the same ancestral origin as you”). As a result, many therapists make the mistake of making statements such as “Education is very important to Asian Americans,” which at face value does not seem problematic because sometimes it can strengthen relations. Unfortunately, it is difficult to know which clients may take offense to stereotypical statements, which is often influenced by client–therapist ethnic match or mismatch and the ethnic-identity stages of the individuals involved. Therapists should make use of stereotypical cultural knowledge as contextual information but continue individualizing therapy for their clients. For example, rather than saying, “Family is really important to Chinese people,” they can say, “I can tell that you really care a lot about your family and that family is really important to you.” Making individualized statements helps clients feel validated rather than stereotyped. In addition, instead of acting on stereotypical assumptions such as “Thought records don’t work with Asian Americans,” it is best to specify even further if one is to stereotype at all, such as “Thought records work better with more educated and more acculturated Asian Americans.” My recommendation would be to use stereotypical statements only when trying to reduce stigma and normalize the therapy process for a client (e.g., “I understand that there is a lot of stigma about mental illness and its treatment in the Chinese American community”) or when integrating cultural strengths and healing practices (e.g., cultural symbols, philosophies, and culturally adapting evidence-based practices     

295

teachings—“Chinese people have a long history of strength and resilience; are there any cultural metaphors or stories that can help you find a solution to this problem?”). Integrating cultural symbols can also enhance healing practices. For example, the tai qi diagram (often referred to as the yin and yang diagram) has a religious association with Taoism but also has cultural meaning because it has been an integral part of Chinese culture and traditional Chinese medicine for thousands of years. Helping clients understand about the need to have balance in energy, or qi (dark and light sides of the picture), about the principle of impermanence (as represented by the curvy line in the middle, which symbolizes the ups and downs of life), and about polarities (e.g., no matter how bad things are, something good is likely to happen and vice versa—as represented by the small dark and light dots in the light and dark regions of the diagram) can be an extremely effective way of integrating cultural strength into psychotherapy. Mental health practitioners must strive to individualize psychotherapy for clients. Unfortunately, there is a greater tendency to make stereotypical statements with ethnic minorities than with White populations (e.g., people don’t often say that family is very important for White people), with the default being that therapy is individualized for White populations. To truly individualize treatments for all groups, therapists need to do a great deal of research on the different cultures that they interact with, just as they do a great deal of research the first time they see a client with a diagnosis they have less experience with. Therapists must use contextual and cultural knowledge in conceptualizing and framing treatment but also need to individualize their understanding for each client and make individualized statements to them. Therapy Orientation as a Way to Reduce Stigma Because many ethnic minorities and immigrants are less familiar with mental health treatment and Western conceptualizations of mental illness (Hwang, 2006b), it is important to orient clients to therapy and the treatment process. Properly orienting clients to therapy is the second domain of the PAMF. A comprehensive therapy orientation program was developed and included important issues such as why and how the program was developed, structural issues in therapy (e.g., meeting length and time), facts and fallacies about therapy, roles and responsibilities of therapists and clients, course of therapy, preventing premature dropout, setting goals for therapy, and addressing emergency issues. Therapy orientation should also include discussions about symptomatology and any diagnoses. Focus groups in the first and third stages of the FMAP revealed that discussing diagnoses with clients can be a controversial issue among mental health practitioners (Hwang, 2008a, 2009). For example, some therapists felt 296       wei-chin hwang

that telling clients their diagnosis would exacerbate feelings of stigma and lead clients to drop out of treatment. In contrast, other therapists felt that telling clients their diagnosis is an ethical and professional responsibility. The PAMF approach takes into account all of the aforementioned issues and recommends making adjustments in terms of timing and method of presentation. For example, rather than simply introducing the nine Diagnostic and Statistical Manual of Mental Disorders (5th ed.; American Psychiatric Association, 2013) symptoms of major depression, therapists can list physical symptoms and mental symptoms in separate columns. Doing so helps individuals and family members understand that depression is not just a mental illness but also involves many physical symptoms. Moreover, asking clients whether they think they have a diagnosis rather than telling them they have a diagnosis uses self-recognition methods for acceptance, which should also be followed by the therapist’s professional opinion. These culturally competent treatment approaches to psychoeducation helped destigmatize diagnoses, increase understanding, and normalize experiences. Principles of active rather than passive learning were also used. A checkbox was placed next to each symptom, and clients were asked to check off the boxes that they felt were representative of the problems that they were having. A discussion about stress and genetic vulnerability (i.e., given enough stress, anybody can become depressed) also helped reduce stigma and reinforce the idea that stress management is very important and will be an integral part of treatment. No clients in the program reacted adversely to talking about their diagnosis, which is contrary to stereotypes that Chinese Americans would not be able to tolerate having a psychiatric diagnosis. In fact, nearly all of the clients were able to identify and accept their diagnosis (see Hwang, 2008a, for a more detailed review of culture and diagnostic issues). These are examples of how studying culture can help strengthen clinical practice and improve clinician understanding of psychological science. Specifically, White American clients may also benefit from understanding that psychiatric illness contains a large somatic component. Active-learning approaches to educating clients about stress management can also help reduce the incidence of depression as well as its recurrence. Such approaches can be especially important for ethnic minorities who are less familiar with psychotherapy and have higher stigma toward mental illness but can also be beneficial for White Americans. Using Cultural Knowledge to Enhance the Treatment Process For therapy to be effective, it must be clearly aligned with the belief system of the client. The third domain of the PAMF is to understand the client’s cultural beliefs about mental illness and its causes. Cultural bridging, culturally adapting evidence-based practices     

297

or the bridging and integration of cultural beliefs with therapeutic concepts and practices, can help engage clients and decrease resistance (Ham, 1989; Hong, 1993; Hwang et al., 2006). One way to accomplish this is to use cultural metaphors, teachings, and stories in a therapeutic manner (Bernal et al., 1995; Costantino, Malgady, & Rogler, 1986). Chinese sayings known as chengyu (i.e., metaphorical sayings that can help teach ethics, highlight morals, provide inspiration, and influence behaviors) can be used to help clients reframe their thinking and engage in healthier activities (Hwang et al., 2006). Although many metaphors may not have been historically created for use in psychotherapy, they can be integrated with and bridged with therapeutic concepts. Most therapists reported that they had never thought of using chengyu in psychotherapy, with a smaller portion reporting that they naturally did so. After discussions, all therapists agreed that integrating metaphors would be beneficial and effective. After researching thousands of chengyu, one was chosen to represent the spirit of the culturally adapted therapy manual and was placed on the cover page: 山不轉路轉;路不轉人轉;人不轉心轉. The translated meaning is: “If a mountain is obstructing your way, then find a road around it. If there is no road around it, then you need to find or make a path of your own path. If you can’t find a way around it or create a path, then you need to change the way you think and feel about the problem.” This metaphor can be bridged with problem-solving therapeutic concepts: Try your best to solve the problems that life throws your way; be flexible in thinking and tackle problems from many different angles; and sometimes you can’t change the objective stressor, but you can subjectively reframe and change the way you think and feel about it to decrease stress and make the most of the situation. Similarly, the nonadapted manual had a related metaphor that discussed how when it rains, people can have very different reactions (e.g., “I hate the rain and traffic is going to be horrible” vs. “It’s great that it’s raining because it will help clean the air and makes me feel refreshed”). A person can’t change the fact that it is raining, but he or she can change the way he or she thinks about it. After discussing the meaning of a metaphor, therapists can apply it to various situations and problems that their clients might face. For example, the chengyu that is the theme of the manual can be further adapted for families. The therapist could say, “It seems like your family is facing some difficult obstacles right now. Let’s think of some ways to solve this problem.” These ways might include one or more of the following: understand the impact that immigration has on families, figure out how to communicate with your children in more effective ways, do some short-term and long-term financial planning for the family, and learn conflict resolution and relaxation techniques to deescalate fights. This solution-focused approach aligns with the problem-solving nature of Chinese culture. After trying out various strategies and evaluating 298       wei-chin hwang

the advantages and disadvantages of different solutions, the therapist can help the family with acceptance and reframing. The therapist could say, Sometimes life is hard and although we should do our best to try to change the situation, sometimes that doesn’t work and the most effective approach is to do our best to think and feel about it in a different way so that our family can make the most of the situation. At this point, it seems like we have tried many different ways to solve the problem, but it won’t be easily resolved in a short time frame. We can continue to be proactive in problem resolution, but in the meantime, we need to think about different strategies to conserve our emotional and mental resources and not be drained by stress, anxiety, anger, and frustration. This way, you’ll be better able to cope with the problem and help your family find the most effective solution.

In my experience, Chinese American clients can be very reactive to CBT if a problem-solving focus is not integrated prior to cognitive reframing. They can become defensive when told that there is something wrong with the way they think or that their thoughts are distorted or irrational. A process-oriented approach that takes into account timing and ordering of interventions can make a significant difference in acceptance and willingness to change. Another method of cultural bridging and adapting psychotherapy is to make a play on words. For example, the title of the culturally adapted treatment manual is Improving Your Mood: A Culturally Responsive and Holistic Approach to Treating Depression in Chinese Americans. A bracket was placed around the fifth character of the Chinese title, 提升您 心「晴」指數: 反映 文化與綜觀整合的華裔美國人憂鬱治療. Putting brackets around the fifth character created a double meaning for the fourth and fifth character combination from “mood” to “brightening one’s day or clearing up the darkness.” Clients were told that the goal of this program is to help them improve their mood as well as to help them solve their problems and brighten their day. Another way to bridge cultural beliefs with therapeutic concepts is to integrate cultural symbols, concepts of traditional medicines, and spiritual beliefs. For example, the notion of qi is an integral part of Chinese medicine and culture (Hwang, 2006b). If the client believes in qi, then reference can be made to psychotherapy as helping to balance one’s qi (e.g., emotional, physical, and spiritual centeredness). Clients can be told that cognitive reframing can help balance one’s qi and cultivate intra- and interpersonal health. Such bridging can help align therapeutic goals with cultural beliefs and gain clients’ buy-in to psychotherapy, thus resulting in greater treatment adherence and reduced dropout. Clients in this clinical trial responded well to the cultural bridging of therapeutic concepts to Chinese cultural notions of qi. This is another example of how studying culture can enhance clinical practice and contribute to psychological science. Many people who are not of culturally adapting evidence-based practices     

299

Chinese descent also believe in energy and balance. A holistic approach may help improve therapeutic outcomes for all people, although the metaphor used may be similar or different (e.g., chakras in traditional Asian Indian medicine, pendulums for balance, or a seesaw in child therapy). It can also help engage clients in therapeutic life changes, such as eating and exercise, which improve both physical and mental health (Walsh, 2011). Methods for Improving the Client–Therapist Relationship The fourth domain of the PAMF is to figure out ways to improve the client–therapist relationship. A strong therapeutic relationship is essential to positive therapeutic outcomes (Norcross, 2011). S. Sue and Zane (1987) recommended helping clients feel immediate direct benefit from treatment or what they called a gift (e.g., normalization, cognitive clarity, reassurance, hope, goal setting, skills acquisition, anxiety reduction, and depression relief). Obviously, this is easier said than done and engaging resistant clients in the initial sessions of treatment can be very different from helping clients who are familiar with and believe in psychotherapy. The second domain of the PAMF (i.e., orienting clients to psychotherapy) can play a pivotal role in client engagement and adherence to treatment. Moreover, using cultural knowledge about the importance of family and asking clients whether they want their family to be part of the treatment process can also increase therapist credibility. This strategy can be particularly important in the community mental health setting where many clients are severely ill and rely on family members to drive them to therapy. It is also important to note that each clinic has a different culture and personality. Some of the Asian-focused clinics recommended engaging family members as part of the standard treatment process. The majority of clinics recommended asking about family engagement, discussing the advantages and disadvantages of doing so, and also providing the option of confidentiality regarding family members. Normalizing client experiences and destigmatizing mental illness and its treatment can also be a very powerful gift. In addition, reassuring clients about confidentiality can allay immediate fears concerning privacy and judgment by others. The treatment manual takes a holistic mind–body approach to improving mental well-being. One essential method for improving the client–therapist relationship was to destigmatize psychotherapy and highlight the relationship between physical and mental health treatment. A critical metaphor was made between psychotherapy and physical therapy: When people tear a muscle or ligament, a physical therapist helps them understand the nature of the injury through education awareness, rehab and exercise of the injured part of the body, and discussion of causal and prevention issues. Similarly, when people experience trauma or negative life events, they can 300       wei-chin hwang

be emotionally or psychologically injured and need to exercise their brain to think about things in different ways and develop effective coping strategies. Reframing of CBT words such as homework to exercise and practice helped reinforce the holistic and strengthening focus of the program and also addressed the issue that there is already too much academic pressure in the Chinese community and nobody wants to do additional homework. Clients responded positively to these modifications—yet another example of how studying culture can enhance clinical practice for other groups and push the field of psychology forward. Another method of strengthening the working alliance is to provide clients with concrete tasks that facilitate skills development (e.g., cognitive reframing and communication effectiveness). In the nonadapted CBT treatment manual (Miranda et al., 2006), an exercise called chaining was used to help clients avoid all-or-nothing thinking and think in more positive ways. Chaining is an active-learning exercise, in which clients write down the problem or situation they are experiencing and work with the therapist to think in more positive and beneficial ways (i.e., the client writes down the problematic situation in the middle, leaving vertical spaces below and above where he or she can write down thoughts that lead to either improved or worse emotions). Although this exercise is very powerful and effective, the focus on changing one’s cognitions may be too much for those coming from cultures where the initial major focus is on problem solving, not changing the way one thinks (many Asian philosophies and religions focus on not thinking as a way to de–stress and as a way to reach enlightenment; Hwang, 2011b). This cognitive focus is similar to many other Western therapeutic homework assignments (e.g., Beck’s [1976] cognitive therapy and Greenberger & Padesky’s [1995] thought records). However, given the solution- and goal-oriented nature of Chinese culture (Hwang et al., 2006), modifying cognitive homework and exercises to be more solution focused was an important cultural adaptation. In Chinese culture, the emphasis on cause and effect is represented by the term yinguo (yin = cause; guo = effect; Hwang et al., 2006). There is also a great emphasis on nature and mountains in Chinese culture (via the five great mountains in Taoism and four sacred mountains in Buddhism). In fact, many Chinese paintings focus on mountain scenes rather than portraits of people, which tend to be a greater emphasis in Western paintings. Climbing mountains is believed to be beneficial for a person’s mind, body, and spirit. Doing so is also believed to help people with perspective-taking on life, as one reaches different vantage points in one’s journey. To address these issues, the principal investigator created the climbingthe-mountain exercise. Similar to chaining, there is a spot to write down the situation. However, the spaces above and below the situation are not vertically culturally adapting evidence-based practices     

301

oriented. Instead, they are represented by progressive steps ascending toward one’s goals and descending toward worst outcomes—an emphasis on yinguo or cause and effect. The greatest emphasis is on achieving one’s goals and avoiding detrimental outcomes, which aligns well with the solution-focused nature of Chinese culture. This exercise or practice was repeated multiple times throughout the treatment manual, with different sessions focusing on different aspects of climbing the mountain (e.g., sessions focused on problem solving vs. detrimental behaviors first, followed by effective vs. ineffective thinking and healthy vs. unhealthy communication styles). Because many Chinese Americans are not used to therapeutic worksheets, the repetition of the same worksheet along various areas of intervention (e.g., thoughts, communications, and actions) helped consolidate skillbuilding and reduce confusion. To adapt the climbing-the-mountain exercise to address family issues, clients were asked what they could do to help improve their family situations, how they might think about their situation in different ways, and how they could communicate with family members more effectively. This is also another example of how studying culture can help improve clinical practice. Many non-Chinese Americans may also benefit from a staged solutionfocused approach that focuses on solving one’s problems first, specifically by having clients write down their goals and worst outcomes so they understand why they are filling out this worksheet, followed by guiding them to change the way they think and feel if they cannot change the actual situation. Understanding and Improving Family Communication The fifth domain of the PAMF focuses on understanding cultural differences in the expression and communication of distress and strengthening individual and family member communication. In doing so, one must understand the usage of verbal and nonverbal communication and direct and indirect methods of social engagement, as well as communication styles (e.g., aggressive, assertive, passive, and passive aggressive). Many collectivistic cultures place a higher emphasis on nonverbal as well as indirect communication as compared with individualistic cultures (D. W. Sue, 1990). Because many practitioners in the United States live in a comparatively verbal and direct communication cultural environment, great care needs to be taken not to overemphasize this method of the communication and to be more vigilant in helping clients be effective in the nonverbal and indirect as well. For example, helping more acculturated youth understand and develop skills in this arena can greatly reduce conflict and misinterpretations (e.g., their parents may not say that they love them, but they express it nonverbally by making food and asking if they remembered to bring their jacket). Helping youth cognitively reframe or use what I call a cultural–linguistic translation 302       wei-chin hwang

box can also be helpful (Hwang, 2011a). Using worksheets to represent what is coming in and how to reframe and reinterpret what comes out can help consolidate skills development and provide a skills-based and concrete visual understanding of how acculturative processes can create communication barriers for families (see the worksheet provided in Hwang, 2011a). In addition, therapists can help youth understand that linguistically, Chinese does not allow for stating “I love you” or wo ai ni in a family context, because it is used only in romantic situations; otherwise, it might sound odd or incestuous. Helping youth reframe why their parents don’t say that they love them verbally and understand other expressions of care is an important focus of therapy. Although there may be a specific cultural–linguistic reason for why immigrant parents may not verbalize caring, parents also need to be trained to be biculturally competent and to engage in language switching (Hwang, 2011a). After the initial resistance is worked through, this therapeutic intervention can be particularly effective. The laughter and understanding created can help strengthen the family’s emotional bond. For example, youth and parents often both laugh and smile when the therapist pushes the parents to language and culture switch and verbally say, “I love you” (Hwang, 2011a). Emphasizing bicultural communication effectiveness training can be a very powerful way to improve family relations as well as provide immigrants with concrete “gifts,” which further enhances therapist credibility and client belief in psychotherapy. Having a strong understanding of communication styles and nuances can greatly facilitate the therapeutic process as well as provide concrete skills for individuals and families to use in resolving problems. Strengthening various aspects of communication also adds to the universal effectiveness of psychotherapy and improves clinical practice for all people, regardless of race. Focusing on Problems Facing the Community Many ethnic minorities and immigrant families face many issues that the mainstream population does not (e.g., immigration stressors, language barriers, experiences with racial discrimination, pressure to assimilate and conform, separation from family, citizenship stresses) as well as adversities that some groups face at higher rates (e.g., divorce, single-parent families). Understanding and properly addressing these issues, the sixth domain, is an important part of individualizing treatment for diverse populations and makes a critical difference in client engagement, perceptions of therapist credibility, and effective outcomes (Hwang, 2006b). Many immigrants also face acculturation-related family problems that are exacerbated by linguistic and cultural gaps between the parents and children. For example, acculturative family distancing (AFD) was covered in one culturally adapting evidence-based practices     

303

section of the treatment manual (Hwang, 2006b; Hwang & Wood, 2009). AFD is defined as the distancing that occurs between parents and youth as a result of communication difficulties and cultural value incongruence that are a consequence of different rates of acculturation among parents and children. AFD increases family conflict and as a result leads to depression and other psychological problems among parents and youth (Hwang, 2011a; Hwang & Wood, 2009; Hwang et al., 2010). Other issues that minorities and immigrants face include racial microaggressions (D. W. Sue et al., 2007) and internalized racism (Jones, 2000). Although Asian Americans are often stereotyped as a model minority, they too face many direct as well as indirect racial transgressions. Probably the most insidious is the immigrant community’s battle with internalized racism. For example, the phrase “fresh off the boat” or “FOB” is used in many ways, ranging from describing those who are recent immigrants to making fun of people who “talk funny, dress weird, and eat strange things.” Just as White Americans historically made fun of ethnic minorities in a similar manner, immigrant communities have begun adopting this method of separating themselves from the outgroup and trying to conform with the ingroup. Specifically, more acculturated immigrants make fun of more recent immigrants and describe them as “FOBs” or “FOBBY.” Although many people try to write it off by stating that these are just descriptive terms, they have a particularly insidious impact on the mental health of immigrant families that many people are not aware of. When acculturated immigrants make fun of FOBs, they are essentially and unknowingly making fun of their own parents, which is an unnatural process that does not normally occur in human development and is an example of internalized racism. This destructive and often invisible force damages family relations because children disidentify parents as role models and disempower them, which results in youth trying to gain their self-esteem from other sources and being indirectly pressured to conform to mainstream White culture. Many classic social psychological experiments showed that pressure to conform can be a deleterious force, negatively influencing people’s lives and leading them to do horrific things (Milgram, 1963; Zimbardo, 2008). Internalized racism manifested in this way is not isolated to Asian Americans and also occurs among other ethnic minority and immigrant groups. CONCLUSION This chapter has provided an integrated top-down and bottom-up approach to culturally adapting psychotherapy. Because of the vast cultural differences among the world’s population, advances need to improve 304       wei-chin hwang

the palatability and effectiveness of psychotherapy for all people. An as-is approach to psychotherapy may be insufficient to meeting the needs of the community, and developing evidence-based and culturally adapted treatments will be critical to improving mental health care for ethnic minorities and immigrants. The study of culture, which should be seen as an integral part of advancing the field of psychology, can lead to many benefits in improving clinical practice and contributing to psychological science. REFERENCES American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Beck, A. T. (1976). Cognitive therapy and the emotional disorders. Oxford, England: International Universities Press. Bernal, G., Bonilla, J., & Bellido, C. (1995). Ecological validity and cultural sensitivity for outcome research: Issues for the cultural adaptation and development of psychosocial treatments with Hispanics. Journal of Abnormal Child Psychology, 23, 67–82. http://dx.doi.org/10.1007/BF01447045 Bernal, G., & Scharró-del-Río, M. R. (2001). Are empirically supported treatments valid for ethnic minorities? Toward an alternative approach for treatment research. Cultural Diversity & Ethnic Minority Psychology, 7, 328–342. http:// dx.doi.org/10.1037/1099-9809.7.4.328 Bui, K. V., & Takeuchi, D. T. (1992). Ethnic minority adolescents and the use of community mental health care services. American Journal of Community Psychology, 20, 403–417. http://dx.doi.org/10.1007/BF00937752 Costantino, G., Malgady, R. G., & Rogler, L. H. (1986). Cuento therapy: A culturally sensitive modality for Puerto Rican children. Journal of Consulting and Clinical Psychology, 54, 639–645. http://dx.doi.org/10.1037/0022-006X.54.5.639 Escobar, J. I., & Vega, W. A. (2000). Mental health and immigration’s AAA: Where are we and where do we go from here? Journal of Nervous and Mental Disease, 188, 736–740. http://dx.doi.org/10.1097/00005053-200011000-00003 Flaskerud, J. H., & Liu, P. Y. (1991). Effects of an Asian client-therapist language, ethnicity and gender match on utilization and outcome of therapy. Community Mental Health Journal, 27, 31–42. http://dx.doi.org/10.1007/BF00752713 Greenberger, D., & Padesky, C. (1995). Mind over mood: Change how you feel by changing the way you think. New York, NY: Guilford Press. Hall, G. C. N. (2001). Psychotherapy research with ethnic minorities: Empirical, ethical, and conceptual issues. Journal of Consulting and Clinical Psychology, 69, 502–510. http://dx.doi.org/10.1037/0022-006X.69.3.502 Ham, M. D. (Ed.). (1989). Immigrant families and family therapy [Special issue]. Journal of Strategic & Systemic Therapies, 8(2), 36–40. culturally adapting evidence-based practices     

305

Hays, P. (2001). Addressing cultural complexities in practice: A framework for clinicians and counselors. http://dx.doi.org/10.1037/10411-000 Hong, G. K. (1993). Synthesizing Eastern and Western psychotherapeutic approaches; Contextual factors in psychotherapy with Asian Americans. In J. L. Chin, J. L. Liem, M. D. Ham, & G. K. Hong (Eds.), Transference and empathy in Asian American psychotherapy: Cultural values and treatment needs (pp. 77–90). Westport, CT: Praeger. Hu, T. W., Snowden, L. R., Jerrell, J. M., & Nguyen, T. D. (1991). Ethnic populations in public mental health: Services choice and level of use. American Journal of Public Health, 81, 1429–1434. http://dx.doi.org/10.2105/AJPH.81.11.1429 Hwang, W.-C. (2006a). Acculturative family distancing: Theory, research, and clinical practice. Psychotherapy: Theory, Research, Practice, Training, 43, 397–409. http:// dx.doi.org/10.1037/0033-3204.43.4.397 Hwang, W.-C. (2006b). The psychotherapy adaptation and modification framework: Application to Asian Americans. American Psychologist, 61, 702–715. http:// dx.doi.org/10.1037/0003-066X.61.7.702 Hwang, W.-C. (2008a). Diagnostic nondisclosure of schizophrenia to Chinese American patients: Are we being culturally sensitive or feeding into cultural misconceptions? Asian Journal of Counselling, 15(1), 1–32. Hwang, W.-C. (2008b). Improving your mood: A culturally responsive and holistic approach to treating depression in Chinese Americans. Unpublished client manual. Hwang, W.-C. (2008c). Improving your mood: A culturally responsive and holistic approach to treating depression in Chinese Americans. Unpublished therapist manual. Hwang, W.-C. (2009). The Formative Method for Adapting Psychotherapy (FMAP): A community-based developmental approach to culturally adapting therapy. Professional Psychology: Research and Practice, 40, 369–377. http://dx.doi. org/10.1037/a0016240 Hwang, W.-C. (2011a). Acculturative Family Distancing (AFD): Cultural–linguistic understanding and skills development. In F. T. L. Leong, L. Juang, D. B. Qin, & H. E. Fitzgerald (Eds.), Asian American and Pacific Islander children and mental health: Vol. 1. Development and context (pp. 47–70). Santa Barbara, CA: Praeger Press. Hwang, W.-C. (2011b). Cultural adaptations: A complex interplay between clinical and cultural issues. Clinical Psychology: Science and Practice, 18, 238–241. http:// dx.doi.org/10.1111/j.1468-2850.2011.01255.x Hwang, W.-C., & Wood, J. J. (2009). Acculturative family distancing: Links with self-reported symptomatology among Asian Americans and Latinos. Child Psychiatry and Human Development, 40, 123–138. http://dx.doi.org/10.1007/ s10578-008-0115-8 Hwang, W.-C., Wood, J. J., & Fujimoto, K. (2010). Acculturative family distancing (AFD) and depression in Chinese American families. Journal of Consulting and Clinical Psychology, 78, 655–667. http://dx.doi.org/10.1037/a0020542

306       wei-chin hwang

Hwang, W.-C., Wood, J., Lin, K., & Cheung, F. (2006). Cognitive-behavioral therapy with Chinese Americans: Research, theory, and clinical practice. Cognitive and Behavioral Practice, 13, 293–303. http://dx.doi.org/10.1016/ j.cbpra.2006.04.010 Institute of Medicine. (1999). Unequal treatment: Confronting racial and ethnic disparities in health care. Washington, DC: National Academy of Sciences Press. Jones, C. P. (2000). Levels of racism: A theoretic framework and a gardener’s tale. American Journal of Public Health, 90, 1212–1215. http://dx.doi.org/10.2105/ AJPH.90.8.1212 Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M., Eshleman, S., . . . Kendler, K. S. (1994). Lifetime and 12-month prevalence of DSM–III–R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Archives of General Psychiatry, 51, 8–19. http://dx.doi.org/10.1001/ archpsyc.1994.03950010008002 LaFromboise, T., Coleman, H. L. K., & Gerton, J. (1993). Psychological impact of biculturalism: Evidence and theory. Psychological Bulletin, 114, 395–412. http:// dx.doi.org/10.1037/0033-2909.114.3.395 Milgram, S. (1963). Behavioral study of obedience. The Journal of Abnormal and Social Psychology, 67, 371–378. http://dx.doi.org/10.1037/h0040525 Miranda, J., Woo, S., Lagomasino, I., Hepner, K. A., Wiseman, S., & Muñoz, R. (2006). Group cognitive behavioral therapy for depression—Thoughts, activities, people and your mood. San Francisco, CA: San Francisco General Hospital. Norcross, J. C. (2011). Psychotherapy relationships that work: Evidence-based responsiveness. http://dx.doi.org/10.1093/acprof:oso/9780199737208.001.0001 Resnicow, K., Baranowski, T., Ahluwalia, J. S., & Braithwaite, R. L. (1999). Cultural sensitivity in public health: Defined and demystified. Ethnicity & Disease, 9, 10–21. Sue, D. W. (1990). Culture-specific strategies in counseling: A conceptual framework. Professional Psychology: Research and Practice, 21, 424–433. http://dx.doi. org/10.1037/0735-7028.21.6.424 Sue, D. W., Capodilupo, C. M., Torino, G. C., Bucceri, J. M., Holder, A. M. B., Nadal, K. L., & Esquilin, M. (2007). Racial microaggressions in everyday life: Implications for clinical practice. American Psychologist, 62, 271–286. http:// dx.doi.org/10.1037/0003-066X.62.4.271 Sue, S. (1998). In search of cultural competence in psychotherapy and counseling. American Psychologist, 53, 440–448. http://dx.doi.org/10.1037/ 0003-066X.53.4.440 Sue, S., & Zane, N. (1987). The role of culture and cultural techniques in psychotherapy. A critique and reformulation. American Psychologist, 42, 37–45. http:// dx.doi.org/10.1037/0003-066X.42.1.37 U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity, a supplement to Mental Health: A Report of the Surgeon General. Washington, DC: Author. culturally adapting evidence-based practices     

307

Walsh, R. (2011). Lifestyle and mental health. American Psychologist, 66, 579–592. http://dx.doi.org/10.1037/a0021769 Yang, L. H., Phelan, J. C., & Link, B. G. (2008). Stigma and beliefs of efficacy towards traditional Chinese medicine and Western psychiatric treatment among Chinese-Americans. Cultural Diversity and Ethnic Minority Psychology, 14, 10–18. http://dx.doi.org/10.1037/1099-9809.14.1.10 Zimbardo, P. G. (2008). The Lucifer effect: Understanding how good people turn evil. New York, NY: Random House.

308       wei-chin hwang

AFTERWORD: SOME CULTURALLY INFORMED CONCLUSIONS NOLAN ZANE, GUILLERMO BERNAL, AND FREDERICK T. L. LEONG

Collectively, the chapters in this volume address many of the significant challenges as well as the major advances made to develop culturally informed evidence-based treatments (EBTs) for ethnic minority and other culturally diverse individuals. However, we researchers need to seriously address the fact that even the most current treatment research programs are not producing best available evidence on minority populations or ethnocultural issues (Mak, Law, Alvidrez, & Pérez-Stable, 2007). To advance the field, these insightful and informative contributions have converged on a number of noteworthy themes. We have focused on those themes that represent what we consider to be best practices in research on culturally informed interventions. One major theme centers on practical and efficient strategies for designing clinical studies that yield substantive information as to how cultural factors affect treatment outcomes and processes. Several contributors have

http://dx.doi.org/10.1037/14940-015 Evidence-Based Psychological Practice With Ethnic Minorities: Culturally Informed Research and Clinical Strategies, N. Zane, G. Bernal, and F. T. L. Leong (Editors) Copyright © 2016 by the American Psychological Association. All rights reserved.

309

noted the need for more research on mechanisms of change in treatment. Essentially, this research unpacks cultural and ethnic group status in terms of specific psychological variations that can account for discrete problems (e.g., low utilization, poor engagement, low retention, poor response, partial response) that result in disparities in quality of care among ethnic minority clientele. For example, understanding how specific aspects of culture affect treatment processes and outcomes provides an empirical basis for developing more effective cultural adaptations to orthodox EBTs. The cultural adaptations constitute hypotheses concerning the mechanisms of action of the adaptation that are purported to enhance treatment processes (e.g., selfdisclosure, client–therapist relationship) or outcomes (e.g., symptom reduction, improved functioning) or both. Researchers have options and some flexibility in addressing these mechanisms. The unpacking can occur prior to designing an intervention or adaptation by identifying cultural factors that are empirically related to some type of critical process in treatment or that are directly related to a treatment outcome. In this way, adaptations that are selected and used or applied have a greater likelihood of actually contributing to the effectiveness of the treatment. Alternatively, the unpacking can occur after an adaptation has been found to be effective. Dismantling studies can isolate which aspects of the cultural adaptation contributed to better outcomes. Regardless of the approach, mixed-method designs can be especially useful for this deconstructing process by identifying constructs that adequately capture the meanings and lived experiences of a certain people relevant to the context of treatment. Mixed-method studies allow for the rigorous documenting and unpacking of cultural variation and a detailed understanding of the meanings and practices underpinning this variation. Numerous contributors addressed the critical issue of cultural validity in measurement or in the intervention procedures. Cultural validity in measurement refers to the effectiveness of a measure or the accuracy of a clinical diagnosis to address the existence and importance of essential cultural factors. One of the core issues of cultural validity involves determining if the measures used are equivalent for different cultural groups. Efforts to address this issue often have involved the translation and/or adaptation of assessment procedures. Unfortunately, very few studies have provided evidence of measurement equivalence even when these steps have been taken. Moreover, the usual confirmatory factor analytic approach to examine equivalence has proven both unwieldy and impractical when testing diverse samples. As research becomes more global in nature, involving more diverse cultural groups, the invariance task becomes almost impossible. A relatively new strategy has been proposed to resolve this problem (Asparouhov & Muthén, 2014). The alignment approach uses algorithms to reduce the need for restrictions that require exact measurement invariance. This method enables tests for invariance 310       zane, bernal, and leong

across many groups and also provides a detailed account of invariance for every model parameter in each group (see Chapter 6, this volume). Another theme that resonated from several contributors involved the need to account for the contextual nature of cultural effects and influences by using a multilevel approach in both data analysis and study designs. Because individuals are nested within certain cultural environments and social groupings, the data by their very nature are hierarchically structured. In terms of data analysis, tests of equivalence must be conducted at both the group and individual levels to determine if the measures are functioning equivalently across both levels. There is a clear need for researchers to account for this multilevel structure of the culturally clustered data by becoming conversant with the analytical procedures that can test multilevel equivalence. In terms of study design, it is highly possible that individuals within a certain cultural group with specific characteristics may respond differently to interventions. A test of the effectiveness of an EBT or adapted treatment should consider the possibility of evaluating latent classes within ethnic minority groups to better understand the specific psychological influences associated with a culture that may account for a disparity or reduce it. A number of contributors noted that researchers must address the continuing challenge of effectively using community-based participatory research (CBPR) to generate and develop culturally informed interventions. CBPR constitutes a proactive, community-driven approach that actively engages community members in intervention design, evaluation, and implementation. The challenges of using a CBPR approach include the substantial time and effort needed to foster authentic and trusting working relationships among researchers and the community; contrasting priorities between research and community time frames concerning institutional review board and research approval, funding, and application of the research findings (especially when the intervention produces positive effects in functioning and distress reduction); the relative emphasis on individual versus systemic change; and the very practical realities of conducting research in the field (e.g., extensive travel, time for the intervention that does not conflict with the 8- to 12-hour regular workday, compensation to families for costs involved for participating in the research such as child care). CBPR can also extend beyond the development of culturally adapted treatments to the mutual development of culturally informed or culturally centered interventions with community partners. The adaptation efforts and clinical dissemination projects discussed in this book serve as operational exemplars for effective, truly collaborative alliances with communities of color to produce the research needed in this area. There also is the need to provide alternative means for ethnic minority clients to access and use culturally informed mental health interventions. One promising approach involves the use of digital and information technologies afterword     

311

to facilitate access to the highest quality mental health services available for ethnic minority communities. These technologies can increase the reach of existing interventions via the Internet and mobile devices. Specifically, the field would greatly benefit from more research on how newer technology, such as telehealth approaches and smartphone applications, can be used to recruit, engage, and retain clients and families in community settings. The authors also align on a couple of noteworthy issues that need to be considered to generate more progress in the development of culturally informed interventions. First, culturally informed treatments do exist, but there clearly is a need for more systematic research that either designs culturally syntonic interventions or determines their effectiveness relative to conventional treatments. In this way, many of the conceptual, methodological, and statistical challenges identified in this volume can be addressed in a more consistent and methodical way, which should result in more efficient and rigorous efforts. For example, a research program on cultural adaptations that routinely uses a battery of outcome measures that have been tested with respect to their cultural validity and measurement equivalence can minimize challenges in interpreting the data. Second, the last few chapters underscore the fact that a fair number of culturally adapted treatments exist, so the question arises as to why these have not been more widely disseminated and applied. Finally, the need to determine the effect of an adapted treatment relative to a conventional treatment or the effect of conventional treatment with ethnic minority clientele reflects a somewhat conservative process that tends to anchor treatment research on interventions originally developed for mainstream populations. Because there is much shared variance in the psychology of individuals from different cultural groups, it is not surprising that conventional treatments can be at least somewhat effective with ethnic minority clientele. However, the question remains whether these interventions are optimally effective and whether other interventions can provide mental health care more efficiently. Decentering some treatment research away from mainstream approaches may offer a significant benefit, and mixedmethod designs embedded in CBPR may provide certain means for doing this. This codevelopment process also could focus on the critical treatment processes and change mechanisms in treatment research. Such efforts can examine how these processes and changes occur naturally in community or cultural contexts and then develop treatments to optimize these processes and changes. In this way, cultural fit is optimized (Gallimore, Goldenberg, & Weisner, 1993) as the intervention would be similar in form and function to a daily or important activity of that client’s culture. This approach also systematically capitalizes on EBT research because the focus of the intervention would involve a critical change process empirically linked to improvement in treatment. 312       zane, bernal, and leong

REFERENCES Asparouhov, T., & Muthén, B. (2014). Multiple-group factor analysis alignment. Structural Equation Modeling, 21, 495–508. http://dx.doi.org/10.1080/ 10705511.2014.919210 Gallimore, R., Goldenberg, C. N., & Weisner, T. S. (1993). The social construction and subjective reality of activity settings: Implications for community psychology. American Journal of Community Psychology, 21, 537–560. http://dx.doi. org/10.1007/BF00942159 Mak, W. W. S., Law, R. W., Alvidrez, J., & Pérez-Stable, E. J. (2007). Gender and ethnic diversity in NIMH-funded clinical trials: Review of a decade of published research. Administration and Policy in Mental Health and Mental Health Services Research, 34, 497–503. http://dx.doi.org/10.1007/s10488-007-0133-z

afterword     

313

INDEX culturally adapted evidence-based treatments for, 170, 181 evidence-based treatments for, 35 and grouping in RCTs, 90 historical trauma of, 25–26 in San Francisco Latino Mental Health Research Program, 248 suicide prevention with. See Suicide prevention with American Indian/Alaska Native youth underrepresentation of, 6 American Psychological Association (APA), 4, 15–19, 178 American Psychological Association Practice Organization (APAPO), 17 American Psychologist, 252 Analysis of variance (ANOVA), 112 Anthropology, 161 Antisocial behavior, 227 Anxiety disorders, 35, 60, 65, 182 Anxiety symptoms, 227 APA. See American Psychological Association APAPO (American Psychological Association Practice Organization), 17 A priori procedures (method bias), 111 Areán, P., 183 Ariely, G., 110 Ashing-Giwa, K. T., 157, 158 Asian Americans and credibility of therapeutic approach, 189 culturally adapted evidence-based treatments for, 181–183, 290–303 cultural validity in diagnosis of, 57, 59–69 face concerns of, 188–190 and grouping in RCTs, 90 and growth of U.S. population, 3–4 heterogeneity within, 85 psychotherapy outcome research with, 35, 36, 39–40 research on breast cancer experiences of, 158, 159, 162

Academic problems, 227 Acculturative family distancing (AFD), 303–304 Adapted evidence-based treatments (A-EBTs), 32. See also Culturally adapted evidence-based treatments Administration bias, 138 Adolescent Coping with Depression Course, 233 African Americans culturally adapted evidence-based treatments for, 171, 181, 183 depression in, 150, 160 in family evidence-based interventions, 273, 277–279 and grouping in RCTs, 90 and growth of U.S. population, 4 misdiagnosis of, 64 psychotherapy outcome research with, 34–35, 38, 40 SAAF program for, 44 in San Francisco Latino Mental Health Research Program, 248, 250, 252–253, 256–257 self-disclosure by, 185 African Canadians, 152 Ahuluwalia, J. S., 290 AIDS. See HIV/AIDS AILS (American Indian Life Skills), 224, 235–237 AI-SUPERPFP Study, 228–229 Alaska Natives. See American Indians/ Alaska Natives Alcohol use, 10, 154, 185, 227–228. See also Substance use disorders Alegría, M., 154–155 Alignment approach, 310–311 Allwood, C., 151 Altschul, D. B., 228 American Indian Life Skills (AILS), 224, 235–237 American Indians/Alaska Natives community-based participatory research with, 200–201, 210. See also Family Listening Program

315

Asian Americans, continued research’s underrepresentation of, 6 in San Francisco Latino Mental Health Research Program, 248 somatic symptoms of, 190 testing instrument equivalence with, 138–139 Assessment. See also Diagnosis cognitive, 137 cultural differences in, 63 and cultural validity in clinical diagnosis, 66–67 equivalence of measures used for. See Testing instrument equivalence importance of, 57 Assimilation, 294 Ataque de nervios, 41 Attachment, 24 Baby boomers, 4 Bandura, Alfred, 247 Baranowski, T., 290 Bauermeister, J. J., 39 BDI. See Beck Depression Inventory Beagan, B. L., 152, 162 Beals, J., 228, 229 Beck Depression Inventory (BDI), 86, 137, 249, 252 Bellido, C., 42, 171, 290 Benish, S. G., 42, 171 Bergman, M. M., 153 Bernal, G., 31, 37, 39, 42, 45, 171 Bernard, W. T., 152 “Best available research evidence,” 18–21 Bias administration, 138 in clinical judgment, 63–64 construct, 19–20, 106–111 instrument, 137 in intent to treat, 82–85 item, 113–119, 138–139 method, 111–113, 136–138 monocultural, 24 response, 137–138 sample, 136 Bii-Zin-Da-De-Dah program. See Family Listening Program Bipolar disorder, 228 Blase, S. L., 214

316       index

Blozis, S. A., 190 Bonilla, J., 42, 171 Borg, I., 109 Borowsky, I. W., 229 Braithwate, R. L., 290 Branch, M., 159–160 Brave Heart, M. Y., 228 Breast cancer, 158, 159, 162 Bridges to High School Program, 44 Brief Strategic Family Therapy, 35, 281 Brooke, R., 22 Brown-Cheatham, M., 68 Bryman, A., 164 Bulgarian populations, 137 Burnout, 130–133 Byrne, B. M., 85, 129n3, 137 CACs (community advisory councils), 203. See also Family Listening Program CA-ITT (contamination-adjusted intent-to-treat) method, 84 Cambodian American refugees, 40–41 Camilli, G., 118 Campbell, D. T., 59, 112, 113 Campbell, T. L., 137 Canadian populations, 137 Capacity building, 200 CAPS (community action projects), 207 Caracelli, V. J., 156 Carpentier, F. R., 277 Case formulation, 21, 22 CAST (Coping and Support Training), 231 Castro, F. G., 154, 158, 159, 162–164 CBPR (community-based participatory research), 199–201, 311. See also Family Listening Program CBT. See Cognitive–behavioral therapy CDI (Children’s Depression Inventory), 81 Center for Epidemiologic Studies in Depression Scale (CES-D), 251, 252, 254, 256, 260 CFA. See Confirmatory factor analysis CFI (comparative fit index), 110 Chadwick, B. A., 226 Chaining, 301 Chambless, D. L., 34

Chan, Florentius, 248 Chandler, M. J., 230 Chase, J., 228 Chavez, G., 38 Chen, M. S., 6 Chengyu, 298 Cheung, F. M., 66–67 Cheung, M.-K., 64 Chickasaw Nation, 157, 161–163 Children’s Depression Inventory (CDI), 81 Chin, M. H., 214 Christopher, S., 202 Chu, J. P., 183 Client–therapist relationship, 300–302. See also Therapeutic alliance Climbing-the-mountain exercise, 301–302 Clinical decision making, 21 Clinical diagnosis, cultural validity in. See Cultural validity in clinical diagnosis Clinical expertise, 21–22 Clinical utility, 20 Coalition building, 200 Coe, K., 154, 158, 159, 162–164 Cognitive assessment, 137 Cognitive–behavioral therapy (CBT) culturally adapted versions of, 182, 190, 290 in psychotherapy outcome research with ethnic minorities, 34–37, 40–41 in San Francisco Latino Mental Health Research Program. See San Francisco Latino Mental Health Research Program Cognitive development, 24 Cole, J. O., 63 Colearning, 200 Collectivist cultures, 295 Common therapeutic factors, 179 Community action projects (CAPs), 207 Community advisory councils (CACs), 203. See also Family Listening Program Community-based participatory research (CBPR), 199–201, 311. See also Family Listening Program

Community ownership, 200 Community problems, 303–304 Community violence, 227 Comorbidity, 23–24 Comparative fit index (CFI), 110 Complementarity interpersonal, 155, 158–159 as rationale for mixed methods research, 156–158 Conduct problems, 35 Configural equivalence, 106 Configural model, 128 Confirmatory factor analysis (CFA), 109, 127, 130–133, 138. See also Multiple group confirmatory factor analysis Connell, A. M., 277 CONSORT guidelines, 82–83 Construct bias, 19–20, 106–111 Constructivism (research paradigm), 156 Contamination-adjusted intent-to-treat (CA-ITT) method, 84 Continual self-reflection, 21 Control Your Depression (Art Ulene), 248 Convenience samples, 19 Convergent validity, 113 Cook, S. C., 214 Cooley-Strickland, M. R., 40 Coping and Support Training (CAST), 231 Coping Power program, 35 Correlated trait–correlated method (CTCM) model, 113 Correlated trait–correlated uniqueness (CTCU) model, 113 Correlations, 112 Credibility, 189, 303 Creswell, J. W., 155, 163 Critical process approach (culturally adapted evidence-based treatments), 186–190 Cross-cultural research, 125. See also Test adaptations in cross-cultural research; Testing instrument equivalence Crystal, D., 67–68 CSTs (culturally sensitive therapies), 32, 43–45

index     

317

CTCM (correlated trait–correlated method) model, 113 CTCU (correlated trait–correlated uniqueness) model, 113 Cuento therapy, 43–44 Cultural bridging, 297–298 Cultural centeredness, 205 Cultural competence defined, 4 need for, 5, 7 Cultural complexities, 294 Cultural differences clinical expertise on, 21 effects of, 24–25 Cultural knowledge, 297–300 Culturally adapted evidence-based treatments, 169–190, 309–312 approaches to, 170–172 content analysis study of, 172–186 critical process approach to, 186–190 efficacy of, 37–43, 45–46 formative method for. See Formative method for adapting psychotherapy integrative framework for. See Integrative framework for culturally adapting psychotherapy Culturally sensitive therapies (CSTs), 32, 43–45 Cultural psychology, 149 Cultural significance, 187 Cultural symbols, 299–300 Cultural validity in clinical diagnosis, 57–69 and language capability, 64–66 overview, 58 and pathoplasticity of disorders, 59–61 and symptom expression, 62–63 and testing, 66–67 and therapist bias, 63–64 threats to, 59 Cumsille, P. E., 92 Dang, J. H. T., 6 Darney, D., 40 Data, missing, 91–95 Data snooping, 87–88 Davidov, E., 110

318       index

Davis, C. S., 118 Davis, K. E., 27 Deacon, Z., 157, 160–163 DeBruyn, L. M., 230 Deep structure cultural adaptations, 179, 183–186, 290 Delta plot, 115–116 Demjaha, A., 79 Dependency, 68 Depersonalization, 130–132 Depression cognitive–behavioral treatments for, 182, 190 and community violence, 227 cultural differences in, 62–63, 65, 150 diagnosis of, 81 prevention and treatment interventions for. See San Francisco Latino Mental Health Research Program and suicide prevention with American Indian/Alaska Native youth, 228 treatments for, 34–37 Depression Management Course (DMC), 263–265 Depression Prevention Course (DPC), 263–265 Depression Prevention Research Project, 253–254 Deregowski, J. B., 137 Deters, P. B., 229 Development (mixed methods research), 156, 158 Diagnosis, 21, 79–80. See also Cultural validity in clinical diagnosis Diagnostic and Statistical Manual of Mental Disorders, 60, 63, 68, 79, 297 Dickerson, D. L., 226 Differential item functioning (DIF), 114–118 Directive therapeutic styles, 189 Disciplinary practices, 38 Discriminant validity, 113 Dishion, T. J., 277 Diversity, 3–4. See also specific headings DMC (Depression Management Course), 263–265 Domenech Rodríguez, M. M., 45

Donoghue, J. R., 118 Dorans, N. J., 116 DPC (Depression Prevention Course), 263–265 Dumas, J. E., 278 Durvasola, R., 60 Dynamic sizing, 294 EBI (evidence-based intervention), 223 EBPP. See Evidence-based practice in psychology EBTs. See Evidence-based treatments ECGs (ethnocultural groups), 75. See also Randomized controlled trials with ethnic minorities Eclectic approaches, 21–22 Ecological validity model (EVM), 171 Eddy, J. M., 40 EFA (exploratory factor analysis), 108–109 Efficacy, 37–43, 45–47 Eid, M., 113 Elias, B., 226 Elkins, J., 228 Emotional development, 24 Emotional exhaustion, 130–132 Empirically supported treatments (ESTs). See also Evidence-based treatments as best practices, 169 criteria for, 6 Enculturation, 229, 294 Enright, J. B., 62 Error uniqueness, 129 Ethnic and racial diversity, 3–4. See also specific headings Ethnic minorities. See specific headings Ethnic validity model, 58 Ethnocentricity, 63 Ethnocultural groups (ECGs), 75. See also Randomized controlled trials with ethnic minorities Etowa, J., 152 Evans-Campbell, T., 226 Even the Rat Was White (R. V. Guthrie), 19 Evidence-based intervention (EBI), 223 Evidence-based practice in psychology (EBPP), 15–27

APA policy on, 4, 15–19 best available research evidence in, 18–21 clinical expertise in, 21–22 components of, 4–5 definitions of, 16–17 other research concepts vs., 18 overview, 5–7 patient characteristics, values, and context in, 22–27 Evidence-based treatments (EBTs) adapted, 32 culturally adapted. See Culturally adapted evidence-based treatments evidence-based practice in psychology vs., 18 gap between culturally sensitive treatment and, 25 with people of color, 19. See also Psychotherapy outcome research with ethnic minorities support for adoption of, 31 EVM (ecological validity model), 171 Expansion (mixed methods research), 156, 158–159 Exploratory factor analysis (EFA), 108–109 Exposure therapies, 35, 182 External validity, 10, 58, 200 Face concerns, 186, 188–190 Families of color in evidence-based interventions, 273–282 challenges with, 275–278 and existing interventions and research, 281–282 overview, 273–275 pilot work on increased participation of, 278–281 Family communication, 302–303 Family Integrated Transitions, 281 Family interaction theory, 206 Family Listening Program (FLP), 201–214 challenges and lessons learned with, 210–212 community-based participatory research processes in, 207–210

index     

319

Family Listening Program, continued guiding principles in development of, 201–202 and history of University of New Mexico-tribal partnerships, 203–204 implications of, 212–214 overview, 204–207 Family support, 229 Ferri, L., 150, 158 Fickenscher, A., 229 Filial piety, 68, 134–135 First Nation communities, 230. See also American Indians/Alaska Natives Fisher, C. B., 202 Fiske, D. W., 112, 113 Fitzmaurice, G. M., 89 FLP. See Family Listening Program “FOB” (“Fresh off the boat”) immigrants, 304 Folktales, 43–44 Formative method for adapting psychotherapy (FMAP), 290–294, 296–297. See also Integrative framework for culturally adapting psychotherapy Freire, P., 207 “Fresh off the boat” (“FOB”) immigrants, 304 Fujino, D. C., 64 Functional Family Therapy, 281 Gallimore, R., 187 GANA (guiando a ninos activos), 38–39 Garcia, N., 150, 158 Garland, A. F., 38 Garrett Lee Smith Act of 2004, 223 GBG (Good Behavior Game), 231 Gender, 24, 189 Generalizability, 32, 34–37, 111–112, 148 GGC (Guiding Good Choices) Intervention, 279–282 Gillespie, B. M., 83 Globalization, 103 Goffman, E., 188 Goldenberg, C. N., 187 Gone, J. P., 25 Gonzales, N. A., 44

320       index

Gonzalez Smith, V., 184 Good Behavior Game (GBG), 231 Goodheart, C. D., 27 Google, 262 Gorman, J. M., 34 Gotuaco, Catrina, 179 Graham, J. W., 92, 95 Graham, W. F., 156 Greene, J. C., 154, 156, 157, 159 Grencavage, L. M., 179 Griffin, R. S., 40 Grima, A., 118 Griner, D., 42 Groenen, P. J. F., 109 Group cognitive behavioral therapy, 249 Gudeman, H., 62 Guiando a ninos activos (GANA), 38–39 Guiding Good Choices (GGC) Intervention, 279–282 Guthrie, R. V., 19 Guyll, M., 278 Hall, G. C. N., 43 Hallett, D., 230 Halperin, L., 154–155 Hambleton, R. K., 105, 114 Handbook of Multicultural Psychology, 85 Harris, A., 229 Harvey, S. M., 159–160 Hawe, P., 209, 210 Hayward, R. A., 82, 84 Healing practices, cultural, 295–296 Hinson, J. D., 157 Hinson, W. R., 157 Hinton, D. E., 40, 41 Hispanic Americans. See Latino and Hispanic Americans Historical trauma, 25–26, 202, 225–226 HIV/AIDS, 250–251, 273–275. See also Families of color in evidencebased interventions Ho, D. Y. E., 134, 188 Hofmann, S. G., 41 Hollon, S. D., 34 Home visitation programs, 256–257 Homework, 301 Hopelessness, 229 Hoyt, D. R., 228 Hu, H. C., 188

Huang, E. S., 214 Huey, S. J., Jr., 34, 36, 39, 42, 172 Hunsberger, P. H., 22 Hurt, L., 230 Huynh, L., 183 Hwabyung, 62 Hwang, W.-C., 171, 290 ICBT (Internet-delivered CBT), 39–40 Imada, H., 65 Independence, 134–135 Indian Health Service (IHS), 223, 230 Indigenous knowledge theory, 205–206 Individual differences, 21, 23–24 Initiation (mixed methods research), 156, 158 Instrument bias, 137 Instrument equivalence across cultural groups. See Testing instrument equivalence Integrative approaches, 21–22 Integrative framework for culturally adapting psychotherapy, 290–305 client–therapist relationship in, 300–302 community problems in, 303–304 cultural knowledge in, 297–300 diversity among minorities and immigrant families in, 294–296 family communication in, 302–303 overview, 290–294 stigma reduction through therapy orientation in, 296–297 Intelligence, 106 Intent-to-treat (ITT), 76, 82–85 Intergenerational trauma, 200 Internalized racism, 304 Internal validity, 10 International Test Commission Guidelines for Translating and Adapting Tests, 104, 105 Internet-delivered CBT (iCBT), 39–40 Interpersonal complementarity, 155, 158–159 Interpersonal psychotherapy, 254 Interpersonal skills, 21 Interpersonal therapy (IPT), 35–37 Item bias, 113–119, 138–139

Item impact, 114 Item response theory (IRT), 114, 127–128 ITT (intent-to-treat), 76, 82–85 Jackson, J. S., 61 Jacobs, E. A., 151, 162 Jaeckle, W. R., 62 Jemez tribe, research with. See Family Listening Program Johnson, C. L., 226 Johnson, K. D., 226, 228 Johnson, R., 68 Jöreskog, K. G., 113, 128 Jurado, M., 39 Kadzin, A. E., 214 Kagawa-Singer, M., 157, 158 Karasz, A., 150, 158–160 Katz, M. M., 62, 63 Kelly, K., 6 Kenny, D., 113 Kim, J. E., 179, 189, 190 Kim, K., 66 Kleinman, A., 60 Ko, J., 40 Koch, G. G., 118 Kohn, L. P., 40 Kopelowicz, A., 184 Kraemer, H. C., 81, 96 Krener, P. G., 63 Ku, H., 189 Kulick, E., 116 Kuo, H., 66 LaFromboise, T., 229, 232, 234 Laird, N. M., 89 Lalonde, C. E., 230 Lance, C. E., 126 Language, 64–66, 171 Lara, P. N., 6 Lardon, C., 229 Last observation carried forward (LOCF), 93–94 Latino and Hispanic Americans collectivist cultures of, 295 culturally adapted evidence-based treatments for, 171, 181–185 depression in, 150, 160

index     

321

Latino and Hispanic Americans, continued and diabetes, 151 in evidence-based interventions, 273, 277 and grouping in RCTs, 90 and growth of U.S. population, 3, 4 heterogeneity within, 85 misdiagnosis of, 64 perinatal experiences of, 154, 253–258 psychotherapy outcome research with, 34–41, 44–45 research’s underrepresentation of, 6 response bias with, 137 in San Francisco Latino Mental Health Research Program, 248–251, 253–256 sexual decision making by, 159–160 Lau, A. S., 37, 38 Leaf, P. J., 66 Leavitt, D., 40 Lee, C., 229 Lee, C. S., 185 Leech, N., 161, 163 Leong, F. T. L., 64, 66–67 Lester, D., 227 Leung, K., 67, 111 Levant, Ronald F., 15–16, 27 Lewinsohn, P. M., 247 Li, M., 112 Liberman, R. P., 184 Lifespan development, 24 Li-Repac, D., 64 LOCF (last observation carried forward), 93–94 Lord’s chi square, 114 Lowery, H. A., 63 MacCallum, R. C., 80 Mantel–Haenszel (MH) procedure, 118–119 Manualized treatment, 37 MAR (missing at random), 91–95 Marsh, H. W., 135 Martinez, C. R., Jr., 40 Maslach Burnout Inventory (MBI), 130–133 Matos, M., 39 Matsumoto, D., 105

322       index

Maximum likelihood (ML), 95 May, P. A., 227, 230 Mazzeo, J., 118 MBI (Maslach Burnout Inventory), 130–133 MCAR (missing completely at random), 91–94 McCabe, K. M., 38 McDougall, C. M., 226 McKinney, H., 60 MDS (multidimensional scaling), 108–109 Measurement, 19–20, 105–107. See also Testing instrument equivalence Medicare, 17 Medoff, L., 229 Mendenhall, E., 151, 162 Mental disorders defined, 79–80 pathoplasticity of, 59–61 Meredith, W., 128, 129 Metaphors, 171, 298, 300 Methadone maintenance, 250–251 Method bias, 111–113, 136–138 Metric equivalence, 107 MGCFA (multiple group confirmatory factor analysis), 108–111 MH (Mantel–Haenszel) procedure, 118–119 MI (multiple imputation), 95 Microaggressions, 304 Miller, A. B., 154–155 Mind–body approaches, 300–301 Minnesota Multiphasic Personality Inventory (MMPI), 60 Mintz, J., 184 Miranda, J., 32, 34, 36, 78 Missing at random (MAR), 91–95 Missing completely at random (MCAR), 91–94 Missing data, 91–95 Missing not at random (MNAR), 91, 93, 94 Mixed-methods research (MMR), 147–165 basis for, 149 defined, 150–153 and dominance of quantitative methods, 148

and generalizability of qualitative research, 148 philosophical underpinnings of, 153–156 rationales for, 156–161 typologies of, 161–164 utility of, 310 ML (maximum likelihood), 95 MMPI (Minnesota Multiphasic Personality Inventory), 60 MMR. See Mixed-methods research MNAR. See Missing not at random Mohatt, G. V., 229 Monitoring the Future annual survey, 228 Monte Carlo simulations, 90 Mood disorders, 59–60. See also Bipolar disorder; Depression Mood Screener, 262–263 Moral development, 24 Moran, J. R., 205 Morisky, D. E., 157, 158 Morrisseau, A. D., 226 Mothers and Babies Internet Project, 258–259 Motivational interviewing, 185, 276 MTMM (multitrait–multimethod) test validation, 111–113 Mullany, B., 225 Multicultural psychologists, 24 Multicultural research, 125. See also Test adaptations in cross-cultural research; Testing instrument equivalence Multidimensional family therapy, 35 Multidimensional scaling (MDS), 108–109 Multimodal Treatment of ADHD study, 36 Multiple group confirmatory factor analysis (MGCFA), 108–111 Multiple imputation (MI), 95 Multiplicity of testing, 85–88 Multistage social learning model, 206 Multisystemic Therapy, 35, 281 Multitrait–multimethod (MTMM) test validation, 111–113 Muñiz, J., 114, 117 Muñoz, R. F., 40, 247 Myers, H. F., 183

Nakash, O., 154–155 Nathan, P. E., 34 National Cancer Institute, 6 National Institutes of Health (NIH), 6, 76–79, 204, 212 Native American Research Centers for Health (NARCH III), 205, 207 Native Americans. See American Indians/Alaska Natives Naturalism (research paradigm), 154 Nervios, 41 NIH. See National Institutes of Health Nondirective therapeutic styles, 189 Norcross, J. C., 179 Norwegian Psychological Association, 17 Novins, D. K., 229 Oakley, L., 159–160 Oden, T., 40 Okazaki, S., 66 One-session treatment (OST), 39 Onwuegbuzie, A., 161, 163 Oregon Easter Seals, 259 Orr, E., 229 OST (one-session treatment), 39 Otte, K., 40 Otto, M. W., 41 Outcome research with ethnic minorities. See Psychotherapy outcome research with ethnic minorities PA (personal accomplishment), 130–132 PAMF framework. See Psychotherapy adaptation and modification framework Pan, D., 39 Pang, K. Y., 62 Panic, 41 Parent–child interaction therapy (PCIT), 38–39 Parent management training (PMT), 40 Parent training, 35, 182, 183 Park, S., 189 Partial invariance, 107 Participant preference trials, 260–261 Paterniti, D. A., 6 Path analytic modeling, 127 Pathoplasticity of mental disorders, 59–61

index     

323

Patterns of response, 137 PCIT (parent–child interaction therapy), 38–39 Pendley, J., 157 Perinatal experiences, 154, 253–258 Personal accomplishment (PA), 130–132 Personalismo, 184 Personality, 24 Personality disorders, 228 Personality tests, 66–67 Phobias, 39 Pluralism, 156 PMT (parent management training), 40 POJ (Pueblo of Jemez). See Family Listening Program Polit, D. F., 83 Pollack, M. H., 41 Polo, A. J., 34, 36, 42, 172 Poortinga, Ype, 104 Postpartum depression, 253–254, 258–259. See also Perinatal experiences Posttraumatic stress disorder (PTSD), 41, 182, 227, 228–229 Power, statistical, 88–91 Pragmatism, 155–156 Progress monitoring, 21 Projecto Puentos a la Secundária (Bridges to High School Program), 44–45 Psychologization of depression, 63 Psychopathology research, 37, 148 Psychotherapy adaptation and modification framework (PAMF). See also Integrative framework for culturally adapting psychotherapy domains of, 294, 296–298, 300, 302 overview, 171, 290, 292 Psychotherapy for ethnic minorities. See Culturally adapted evidencebased treatments Psychotherapy outcome research with ethnic minorities, 31–47 and efficacy of culturally adapted EBTs, 37–43 future directions for, 45–47 generalizability of, 32, 34–37 and innovation, 43–45 overview, 31–33

324      index

Psychotic disorders, 184 PTSD. See Posttraumatic stress disorder Pueblo of Jemez (POJ). See Family Listening Program Puerto Rican culture, 43–44 Qi, 299 Qualitative research, 148. See also Mixed-methods research Quality improvement (QI) interventions, 250 Quantitative research, 148. See also Mixed-methods research Quintana, S., 171 Racial diversity, 3–4. See also specific headings Racial microaggressions, 304 Racism, internalized, 304 Raju’s area measure, 114 Randomized controlled trials (RCTs), 4 manualized treatment in, 37 and other research evidence, 18 preference for, 15 in San Francisco Latino Mental Health Research Program. See San Francisco Latino Mental Health Research Program Randomized controlled trials with ethnic minorities, 75–96 choosing primary outcomes with, 78–82 and intent-to-treat, 76, 82–85 missing data in, 91–95 multiplicity of testing with, 85–88 NIH policy issues with, 76–79 in psychotherapy outcome research, 35 statistical power considerations with, 88–91 RCTs. See Randomized controlled trials Realism (research paradigm), 156 Reaman, J. A., 205 Redmond, C., 278 Research on ethnic minority issues. See also specific headings criteria for, 5–6 need for, 5–7 Resnicow, K., 290

Response bias, 137–138 Response styles, 137 Ribera, E., 41 Ridley, C. R., 185 Riley, T., 209 Riverside County Department of Mental Health, 259 Robin, F., 117 Robinson, A., 40 Rodríguez, I., 39 Rodríguez, M. D., 42, 171 Rogler, Lloyd, 90 Rosen, D. C., 154–155, 158 Rosselló, J., 37 Royston, P., 80 Russell, G. L., 64 Ryder, A. G., 62 SAAF (Strong African American Families) programs, 44 Sabin, C., 63 Sabin, J. E., 65 Sample bias, 136 Sanborn, K. O., 62 San Francisco Latino Mental Health Research Program, 247–266 Depression Clinic studies, 249–250, 252–253 Depression Prevention and Depression Management courses, 263–265 Depression Prevention Research Project, 248–249 Methadone Maintenance Study, 250–251 Mothers and Babies Projects, 253–259 Multilingual Online Depression Screener, 262–263 Stop Smoking and Depression Internet Project, 259–261 Text Messaging in CBT Project, 261–262 Voice Recognition Depression Study, 251–252 Santiago, R., 39 Scalar equivalence, 107 Schafer, J. L., 92, 95 Scharrón-del-Río, M. R., 31

Schizophrenia cultural differences in, 59–60, 62 and suicide prevention with American Indian/Alaska Native youth, 228 School-based suicide prevention with American Indian/Alaska Native youth, 230–232 Scientist–practitioner model, 21, 148 SCL (Symptom Checklist), 66 SCL-36 (Symptoms Checklist 36), 86 Secondary outcomes, 87 Self-construals, 186 Self Description Questionnaire, 135, 138 Self-disclosure, 138, 184–185, 189 Self-efficacy, 229 Self-esteem, 138–139, 229 SEM (structural equation modeling), 127–130 Sequential constraint imposition, 110 Serna, P., 230 Serpell, R., 137 Sexual activity, 273 Sexual decision making, 159–160 Sheiner, L. B., 84 Shepard, L. A., 118 Shetgiri, R., 35 Shevock, A. E., 92 Shiell, A., 209 Shuy, R. W., 65 Sim, J., 83 Simpatia, 184 Singelis, T. M., 159, 160 Smith, T. B., 42, 171 Smoking cessation, 259–261 Snowden, L. R., 64 Social anxiety, 182 Social-cognitive development, 24 Social cognitive theory, 247 Social development model, 206 Socialization, 24 Social learning theory, 247 Social psychology, 148 Social self-concept, 135–136 Social withdrawal, 227 Solano-Flores, G., 112 Somatic symptoms, 60, 190 Sources of Strength program, 231 Spirituality, 152

index     

325

Spoth, R., 278 Standardization index, 116–118 Stanley, J. C., 59 Statistics. See specific headings Stereotypes, 295–297 Stigma, 186, 190, 295–297 Stigma reduction, 296–297 Stokes, M. E., 118 Stop Smoking and Depression Internet Project, 259–261 Strauss, J. H., 226 Strict equivalence, 129 Strong African American Families (SAAF) programs, 44 Structural equation modeling (SEM), 127–130 Structural equivalence/invariance, 126–127, 129–130. See also Testing instrument equivalence Subgroups, 87 Substance use disorders in American Indian/Alaska Native populations, 200, 204, 227–228 culturally adapted versions of, 182 and HIV/AIDS interventions, 273–274 Sue, D. W., 60, 68 Sue, S., 19, 32, 60, 61, 64, 68 Suicide prevention with American Indian/Alaska Native youth, 200, 223–238 case example, 232–237 culturally unique protective factors in, 229–230 culturally unique risk factors in, 225–228 history of, 223–225 and psychological functioning, 228–229 school-based, 230–232 Surface structure cultural adaptations, 179, 184, 185, 290 Sussewell, D. R., 58 Sussman, J. B., 82, 84 Swedish populations, 137–138 Symptom Checklist (SCL), 66 Symptom expression, 62–63 Symptoms Checklist 36 (SCL-36), 86

326      index

Tai qi diagram, 296 Tak, J., 66 Takeuchi, D. T., 66 Tanzer, N. K., 111, 138 Tashkkori, A., 163 Task Force on Promotion and Dissemination of Psychological Procedures, 31 Taylor, C., 149 Test adaptations in cross-cultural research, 103–119 and construct bias, 106–111 and item bias, 113–119 and method bias, 111–113 overview, 103–104 terminology in, 104–106 Testing. See Assessment Testing instrument equivalence, 125–139 analytic strategies in, 127–133 basic concepts of, 126–127 common complexities in, 133–139 overview, 125–126 Test translation, 105 Text messaging, 261–262 Thayer, D. T., 118 Therapeutic alliance, 68, 184–185 Torres, R., 39 Traditional Chinese medicine, 291, 292 Trauma historical, 25–26, 202, 225–226 intergenerational, 200 and mind–body approaches, 300–301 Trauma-focused cognitive behavior therapy, 35 Treatment alliance, 68, 184–185 Treatment contamination, 82 Treatment efficacy, 20 Treatment implementation, 21 Treatment planning, 21 Triangulation (mixed methods research), 156, 157 Tribal participatory research, 200 Trimble, J. E., 202 Trust, 277 Tsai, T.-I., 157–159, 162 t tests, 112 Tuskegee Study, 277 Tyler, F. B., 58 Type 2 diabetes, 151

Type I errors, 86 Type II errors, 86, 89 Ulene, Art, 248 University of California, San Francisco (UCSF), 247. See also San Francisco Latino Mental Health Research Program University of Nebraska (UN), 204 University of New Mexico, 201, 203–204 U.S. Surgeon General, 20 Validity convergent, 113 cultural. See Cultural validity in clinical diagnosis discriminant, 113 external, 10, 58, 200 internal, 10 Vandenberg, R. J., 126 van de Vijver, F. J. R., 66–67, 105, 111, 129n3, 138 Van Winkle, N. W., 227 Velásquez, R. J., 68 Verbal self-concept, 136–137 Walls, M. L., 226, 228 Walters, A. E., 214 Walters, K. L., 226 Wampold, B. E., 27, 171

Ware, J. H., 89 Weisner, T. S., 187 Westernmeyer, J., 59 Whaley, A. L., 27 Whitbeck, L. B., 226, 228 White privilege, 25 Williams-McCoy, J., 58 Wolsko, C., 229 Women, Infants, and Children (WIC) program, 257 Wright, C. C., 83 Wundt, Wilhelm, 149 Xing, D., 114 Yakushko, O., 65 Yeh, M., 38, 188 Ying, Yu-Wen, 248 Yoder, K. A., 229 Yu, M., 228 Yup’ik people, 229 Zane, N. W., 19, 32, 179, 188, 189, 190 Zarate, R., 184 Zeiss, A. M., 247 Zenisky, A., 114 Zukowski, A. P., 159–160, 162 Zuni Life Skills Development Curriculum (ZLS), 232–237 Zwick, R., 118

index     

327

ABOUT THE EDITORS

Nolan Zane, PhD, is a professor of psychology and Asian American studies at the University of California–Davis and directs the Asian American Center on Disparities Research. His research focuses on face concern and related issues in client and care provider interactions, culturally informed sociobehavioral interventions, and determinants of addictive behaviors among Asian Americans. He is a fellow of the American Psychological Association (APA) and received the Distinguished Career Contribution to Research Award from APA’s Division 45 (Society for the Psychological Study of Culture, Ethnicity & Race), the Samuel Turner Mentor Award, and the President’s Award for Distinguished Contributions to Mental Health Disparities Science from APA’s Division 12 (Society of Clinical Psychology). He also was appointed to the APA’s Presidential Task Force on Evidence-Based Psychological Practice. Guillermo Bernal, PhD, is a professor of psychology at the University of Puerto Rico and director of the Institute for Psychological Research. His work has focused on research, training, and the development of mental health services for ethnocultural groups. He is an early contributor to the dialogue on cultural adaptations of evidence-based treatments. Since 1992, he and his colleagues have generated evidence on the efficacy of culturally adapted evidence-based 329

treatments, carried out translations and development of instruments, and published on factors associated with vulnerability of depression. Dr. Bernal received his doctorate from the University of Massachusetts Amherst in 1978. He is a fellow of APA Divisions 45 (Society for the Psychological Study of Culture, Ethnicity & Race), 12 (Society of Clinical Psychology), and 27 (Society for Community Research and Action: Division of Community Psychology), and a member of 29 (Society for the Advancement of Psychotherapy) and 43 (Society for Couple and Family Psychology). He is vice president of the Caribbean Alliance of National Psychological Associations and editor of the Puerto Rican Journal of Psychology. He has received numerous awards for his research, the most recent of which is the Stanley Sue Award for distinguished contributions to diversity from the Society of Clinical Psychology (2015). Frederick T. L. Leong, PhD, is a professor of psychology and psychiatry at Michigan State University and serves as the director of the Consortium for Multicultural Psychology Research. He has authored or coauthored over 240 journal articles and book chapters, and edited or coedited 14 books. Dr. Leong is a fellow of the APA and the Association for Psychological Science. He is the founding editor of the Asian American Journal of Psychology and associate editor of American Psychologist. His major clinical research interest centers on culture and mental health and cross-cultural psychotherapy, and his industrial–organizational research is focused on cultural and personality factors related to career choice, work adjustment, and occupational stress.

330       about the editors