Children and the Formal Features of Television: Approaches and Findings of Experimental and Formative Research 9783111641089, 9783598202056

Table of contents :
Tracing research into formal aspects of television: An introduction
Part I. Children and the formal attributes of television: Findings of research, implications for production
The forms of television: Effects on children’s attention, comprehension and social behavior
Children’s attention to television: Implications for production
Emotions depicted on and stimulated by television programs
Children and the formats of television advertising
How children understand television and learn from it: A Swedish perspective
Television in the lives of French children: A review of recent research
Beyond the formats of television: The effects of student preconceptions on the experience of televiewing
Part II. Principles, methods and problems of formative research
The educational use of production variables and formative research in programming
Formative research in the production of television for children
Formative research and evaluation of instructional programs
Methodological aspects of formative research
About the contributors

Citation preview

Communication Research and Broadcasting No. 6 Editor: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen (IZI)

Children and the Formal Features of Television Approaches and Findings of Experimental and Formative Research Edited by Manfred Meyer

KG-Saur München • New York• London • Paris 1983

Editor of the series: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen Editor of this issue: Manfred Meyer Editorial assistant: Rosemarie Hagemeister Address of editor and editorial staff: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen Rundfunkplatz 1, D-8000 München 2 Telephone (089)5900-2140, Telex 52107-0 brm d

CIP-Kurztitelaufnahme der Deutschen Bibliothek Children and the formal features of Television : approaches and findings of experimental and formative research / ed. by Manfred Meyer. — München ; New York ; London ; Paris : Saur, 1983. (Communication research and broadcasting ; No. 6) ISBN 3-598-20205-9 NE: Meyer, Manfred [Hrsg.]; GT

© 1983 by K.G. Saur Verlag KG. München Phototypesetting by Ähren-Verlag W. Brehm oHG, München Printed in the Federal Republic of Germany by grafik + druck GmbH Co., München ISBN 3-598-20205-9

CONTENTS Manfred Meyer

Tracing research into formal aspects of television: An introduction

Part I

Children and the formal attributes of television: Findings of research, implications for production

Mabel Rice, Aletha C. Huston, John C. Wright

The forms of television: Effects on children's attention, comprehension and social behavior

21

Daniel R. Anderson Diane E. Field

Children's attention to television: Implications for production

56

Aimée Dorr, Catherine Doubleday, Peter Kovaric

Emotions depicted on and stimulated by television programs

Ellen Wartella, Linda S. Hunter

Children and the formats of television advertising

144

Ingegerd Rydin

How children understand television and learn from it: A Swedish perspective

166

Television in the lives of French children: A review of recent research

188

Beyond the formats of television: The effects of student preconceptions on the experience of televiewing

209

Pierre Corset Gavriel Salomon

97

Part II

Principles, methods and problems of formative research

Keith W. Mielke

The educational use of production variables and formative research in programming

233

Edward L. Palmer

Formative research in the production of television for children

253

Formative research and evaluation of instructional programs

279

Methodological aspects of formative research

310

Saul Rockman Konrad J. Burdach About the contributors

331 5

Tracing research into formal aspects of television: An introduction One of the principal objectives of the Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen, since its foundation in 1965, has been to provide the producers and editors of television programmes for children and young people with information on theoretical issues and empirical findings of social-science research, to make them acquainted with the psychological phenomena related to children's televiewing and to stimulate informative discussions between them and experts on communication research by organising seminars, workshops and information weeks. One of the channels through which these efforts find their expression is our English-language publication series Communication Research and Broadcasting, the articles of which are also made available to our German-speaking readers through the Institute's other publication series, Schriftenreihe. This volume is intended as a kind of reader comprising information on some new aspects of research into the field of television and children: The first part contains overview articles on the experimental treatment of television's formal features and the effects on children's attention, comprehension and social behaviour, on new approaches to their cognitive and emotional interaction with what is offered to them on the screen, and on the nature and complexity of the televiewing process itself. In the second part we provide — especially for our European readers — some introductory work on the rationale, goals, methods and methodological aspects of formative research, a field of investigation that has — at least in the United States — rapidly developed in connection with television programming for children, and in which the testing of the effectiveness of forms and modes of presentation plays an essential part. The point of departure of this book was the realisation that recent research literature contains some issues which — so we believe — are directly related to the everyday business of programme production. On the one hand, there are researchers and research teams studying television's formal features and their effects. In our context it is a question of how children perceive and understand these formal features and visual codes — such as zooms, camera movements, cuts, montage techniques, ellipses and programme pacing — as well as auditory features — such as music, special sound effects, peculiar voices etc. —, to what extent they arouse and sustain their attention, support comprehension or encourage intended learning processes. On the other hand, it could be noticed that a type of applied research, which initially was referred to as "formative evaluation" but which is now known 7

as "formative research", began to emerge from the cooperation of programme producers, script writers and message designers with psychologists and educationalists in the planning and producing of the now world-famous Sesame Street series at the end of the sixties. It is a question of research activities accompanying production which are primarily intended to give help in making decisions for an optimum form of presenting programmes or programme segments with regard to achieving the given programme objectives, such as appeal and attractiveness; gaining, directing and sustaining attention; comprehensibility; knowledge gain etc. Roughly speaking, the initial stages of formative research are concerned with further development and systématisation of, and with the attempt to find a scientific approach to, the pilot testing — familiar to us for decades — of draft programmes, especially those using different presentational elements. One could also mention in this connection the influences of the rapidly developing instructional technology — also in the sixties — and of its systems approach to media production, particularly of the rigorous principle of "test and develop", as has been consistently applied to the production of course material for the Open University in Great Britain. As far as basic research is concerned, the question of television's formal features and their effects was not unimportant in the fifties and sixties when experiments were being made with — then still young — educational television, and one of the questions of interest was whether comparatively better learning results could be achieved with the media and certain modes of presentation (e.g. colour, animation, visual cues). The results of these many varied efforts were analysed and described in the classical study by Chu and Schramm (1968). This line of research on the educational effects of production variables continued in the U.S.A. (C.M. Anderson, 1972; Schramm 1972; Gulliford, 1973); readers interested in further bibliographical information are referred to this Institute's documentation on learning by television (Meyer et al., 1980). There is also an interesting research tradition in Canada, associated with the work of Gary O. Coldevin at the Concordia University, Montreal, and of Arthur Sullivan and Jon Baggaley, Memorial University of Newfoundland, culminating in a series of annual conferences on "Experimental Research in TV Instruction", which started in St. John's, Newfoundland, in 1978 (Baggaley and Sharp, 1979; Baggaley, 1980). Baggaley and his colleagues had formerly conducted a long series of experiments on TV presentation techniques at the Centre for Communication Studies, University of Liverpool, using university students as subjects (Baggaley and Duck, 1976; Baggaley et al., 1980). In only one experiment — 8

a replication study with a set of variables related to a TV speaker's appearance and behaviour, interviewer reactions and variable camera angles — were schoolchildren tested. What is interesting in our connection is the author's conclusion that "...the analysis of children's responses to image variations suggest quite categorically that their capacities to draw inferences from nonverbal images on highly subtle bases are greatly underestimated" (Baggaley et al., 1980, p. 127). The authors recommend that "...individual effects upon children should clearly be subjected to much further scrutiny; for unless we can understand and predict the judgement criteria applied by children in complex perceptual situations, we shall be unable to help them to resolve the conflicts they evidently experience as a result, or to overcome their potential vulnerability to persuasive influences of an abusive kind" (ibid., p. 128). In the Federal Republic of Germany it was Hertha Sturm, former Director of Research of the Institute, who drew attention to the importance of mediaspecific formal features, especially with regard to the emotional impact on children of programme segments marked by ellipses, quick cuts, rapid changes of camera angle and zoom (Sturm, 1975; 1981). She is working on a strictly recipient-oriented approach to the field and pleads for a media dramaturgy that is more favourable to the recipient. Her most recent research work, just published in the international review Fernsehen und Bildung of the Bayerischer Rundfunk (Sturm et al., 1983), goes in the direction of the effects of presentational variables; she and her colleagues have investigated the effects of three different text versions underlying an identical children's programme, using psycho-physiological measures with approx. 9-year-old elementary schoolchildren. The focus of this research is on emotional responses to a variation of speech in relation to a film presentation. One of the first to take a research interest in the formal properties of the television medium per se, in a sense that it presents a symbol system by itself with a unique mixture of verbal and nonverbal representational codes, was Gavriel Salomon in Israel. Parallel to his research into the effects of Sesame Street on Israeli children (Salomon et al., 1972) and taking as a point of departure the hypothesis that the visual codes seen on television can be internalised, thus shaping mental processes (Salomon, 1972), he carried out a number of experiments into the way in which certain production techniques (e.g. zooming in and out, close-ups vs. long shots, varied camera angles) influence the children's cognitive skills (Salomon, 1977; 1979). Salomon indeed provided some empirical evidence that seemed to be in line with one of the central notions of Marshall McLuhan — without of course particularly referring to him — who in the mid-sixties stirred up both the academic community and some pessimistic media philosophers with his book 9

Understanding Media: The Extensions of Man (McLuhan, 1964). One of his theses was that it is especially the formal attributes of the new, electronic media that profoundly influence our way of thinking and cognitive processing, just as it has been affected in a different way for centuries by the linearity of the print media. Thus the "television generation" grows up with and learns a new "language" that is made up of the representational codes of film and television. Logically, the urge to investigate — and finally the teaching of — visual literacy or media literacy grew out of this line of McLuhan's thinking (cf. Huston-Stein and Wright, 1979). Actually, one of the major research interests in our field is now the question of how the developing child understands — in a wider sense — its media environment, how it understands and perceives television, the televiewing experience itself, the "language" of this medium as its formal property. The possible answers to these questions constitute the overall objective, the common denominator, as it were, for the articles that follow in Part I of this book. The formal features of television, its audiovisual and linguistic representational codes, and their effects on children's attention, comprehension and social behaviour are the central theme of a series of studies carried out by Aletha C. Huston and John C. Wright with their doctoral candidates and colleagues at the Center for Research on the Influences of Television on Children (CRITC), founded in 1978. What is probably the most comprehensive report on these research activities and at the same time an excellent overview of the rationale and current state of this direction in research was written by Huston and Wright in collaboration with Mabel Rice (Rice, Huston, and Wright, 1982). This work was commissioned by the United States' National Institute of Mental Health (NIMH) and published as part of an illuminating inventory of current research entitled Television and Behavior: Ten Years of Scientific Progress and Implications for the Eighties (Pearl, Bouthilet, and Lazar, 1982; Vol. II), which in its turn was intended as an update and elaboration of the mammoth research programme of the Surgeon General's Scientific Advisory Committee on Television and Social Behavior, completed in 1972 (Surgeon General, 1972; with Technical Reports in 5 Volumes). The NIMH kindly permitted us to reprint this paper. We are very grateful to the authors for making an updated, revised and supplemented version available to us which incorporates the findings of the latest experiments, and to which, especially in the concluding remarks, some thoughts have been added on the practical implications for the production of children's programmes, as we had requested. 10

Daniel Anderson and Diane Fields were most consistent in complying with our wish to point out the practical relevance of empirical social-science research. They commence their article by proposing a number of principles of children's attention to television, go on to review the child- and television-related attention research and then "restate the principles and suggest how the producer of children's programmes can use them in programme design" (p. 56). They relate their implications to the production of educational programmes for children, but many of them give rise to at least conclusions or ideas for entertainment programmes. It should be stressed in this connection that it is by no means a question of listing manipulative techniques to arouse and sustain attention, such as formative research tried to do at the beginning of the production of Sesame Street. Anderson belongs to the group of researchers who no longer see in the child the passive viewer confronted with an irritating bombardment of incomprehensive stimuli, but who, like Huston and Wright, understand turning to the television screen as a complex active attentional strategy: "...we have proposed that young children's visual attention to television is actively and strategically guided by their attempts to comprehend ongoing content.... We suggest that children not only take into account many of the demands, distractions and information provided by the viewing context, but also learn the marker functions of formal features and other informative cues" (Anderson and Smith, in press; cf. also Anderson and Lorch, 1983). Aimée Dorr and her colleagues Catherine Doubleday and Peter Kovaric, in their review of research into "Emotions depicted on and stimulated by TV programs", take into account that producers of children's programmes might be interested, on the one hand, in the varying levels of the emotional reactions and needs of their target groups and, on the other, in the ability of children of different age groups to understand and to recognise emotional events on the screen as such. This is also a question of the emotional attitude to the televiewing experience in general, for example as a leisure-time pursuit, and to the gratifications resulting from it. Undoubtedly, the value of this report lies in the fact that a wealth of evidence has been collated on all possible facets of the interrelation of television and children's emotions. It seems however, that there are relatively few empirical findings on the direct connection between formal features and the emotional responses of children; this may be a result of the general methodological problem of reliably measuring such extremely context-related reactions as research into emotions is confronted with. It should not be forgotten that the predominant form of television as seen by many children in the world — presumably for most of the regular viewers — 11

is commercial television. That implies being bombarded for hours by commercial spots with formats which make maximum use of all conceivable elements of presentation in film and television production, conventional montage as well as electronic animation. We can assume that there are tons of reports on pre-testing and formative evaluation, or guidelines for effective production with a subtle psychological basis which deal in detail with the effects of numerous modes of presentation on emotions, subliminal cognitions or — for purposes of sales promotion — on socially relevant attitudes and behavioural patterns. Practically none of this material is generally accessible. Business implies business secrets, and the best "tricks of the trade" may be, quite logically, among the best-guarded secrets. The influence of commercial television has been a source of public concern especially in the United States, (cf. Adler et al., 1980); the "selling face of television" (Palmer and Dorr, 1980) is one of the main areas of investigation. Ellen Wartella and Linda Hunter have summed up for us the most important points and tendencies dealt with by American communication researchers in this field. They also present their own new approach to analysing the discourse structure of television commercials directed to children and to the informative or persuasive aspects of the particular formats and presentational styles of TV advertising. Sveriges Radio, the Swedish Broadcasting Corporation, can be proud of the fact that it possesses one of the best research units, i.e. the Audience and Programme Research Departments (SR/PUB), at least among the European stations. Within this department there is a highly active group which deals particularly with the relationship of children to the broadcast media and their programmes (cf. von Feilitzen et al., 1979). Ingegerd Rydin has for many years been a member of this specialised research team. She has focused her attention on the problems of how children process information and understand television messages, has herself carried out several programmerelated studies (e.g., Rydin, 1972; 1976a; 1976b) and has also gone into the effects of formal features (cf. Murray, 1980). In her contribution to this volume she reports on several Swedish investigations and places them into the context of current research on children's comprehension of television worldwide. It is a Swedish view of how television is understood at various developmental stages and on the consequences for children's learning processes resulting from the medium. The influence of moving pictures on children, i.e. of television and other audio-visual media, interested French researchers at the beginning of the seventies, particularly in connection with teaching processes and imparting cognitive abilities (cf. Allouche-Benayoun, 1975). In recent years in France 12

interesting research activities have developed which look into the significance of television in the lives of children and young people from a sociopsychological aspect, its influence on their individual development and their social behaviour. Pierre Corset, from the research team Recherche prospective of the Institut National de l'Audiovisuel, Paris, has summed up for us the latest research trends in France and described some of the methodologically interesting studies and their findings, using a developmental perspective. From what has been said before about the work of Gavriel Salomon in Israel, it seemed obvious that we should ask him for a contribution on the interaction of media-specific codes and cognitive skills. To our surprise, however, he declined, referring to the comprehensive reports already published on this series of investigations (Salomon, 1979), and offered us instead a report on his new line of research dealing with the Amount of Invested Mental Effort (AIME) which children expend while viewing television or turning to other media activities such as book reading. The relation of children to television, he found, is largely influenced by the perceptions and preconceptions both of their own abilities and of what the medium's modes of presentation and contents demand from them (Salomon, 1981). Thus, especially with children highly experienced in televiewing, even the comprehension of the more complex features and of montage becomes more or less automatic and demands little mental effort, which leads to the paradox that the more "literate" viewer might even learn less from television. Keith Mielke's contribution serves as a link, as it were, between the two aspects covered by this book. Now Associate Vice-President for Research at the Children's Television Workshop (CTW), Mielke has long been known as an ardent advocate of a closer cooperation between researchers in communication and producers of educational television programmes (Mielke, 1968, 1970) as well as of the application of instructional technology principles to communications research and theory (Mielke, 1972). His article is a slightly shortened version of his contribution to a research monograph entitled Television as a Teacher, commissioned and published by the Directorate for Children and Youth of the National Institute of Mental Health in the U.S A. (Coelho, 1981). Mielke understands the experimental development of formal features — "production factors", as he calls them — as well as formative research as the two key elements in a highly complex process aiming at the purposeful and constructive uses of television to produce intended effects on children and youth. He sums up the early history of the interrelation of production techniques and production-related research and discusses, in the light of 13

experience gathered at the CTW, the limits and potential as well as the approaches and procedures of formative research. The contribution from Edward Palmer, the Vice-President for Research of the CTW, is an updated and expanded version of a fundamental presentation of the starting points, the rationale and the methods of formative research as it was initially developed, practically at the moment the idea of Sesame Street was born. In its essential aspects this article has been published elsewhere (Palmer, 1973, 1974). The arguments it contains, however, "set much of the general conceptional framework and the language of formative research which have become standard in the field" (Palmer, in a letter to the editor), and we are of the opinion that the contribution in this part of the book, which is devoted to an introductory presentation of the complex of problems of formative research and, as stated above, is meant to reach a special group of readers, fulfils a purpose of its own. Not quite so well-known as the world-famous CTW although hardly less important in the U.S.A. and Canada is the Agency for Instructional Television (AIT) in Bloomington, Indiana, known as National Instructional Television (NIT) up to 1973. Saul Rockman, its Director of Research, describes AIT's research and evaluation activities as they relate to the production of instructional television programmes for schoolchildren. The significance of these activities is necessarily quite different from what it is in the production of children's programmes, one of the main reasons being that through the "consortium development process" the actual realisation phase of the programmes is preceded by a period in which the evaluation already plays an essential part during the conception and initiation phase. This safeguard for those participating in the consortium — both with respect to funding, cost-efficiency and the use of the programmes on as broad a front as possible in different school systems — is likely to be of value and interest to European producers particularly with regard to the growing efforts to cut costs by means of cofinancing and coproductions. In the last contribution, finally, Konrad J. Burdach discusses some methodological aspects of formative research, asking, for example, to what extent the "classical" evaluation criteria for empirical basic research, such as reliability, validity and generalisability, have any relevance for the formative research approach. Burdach clarifies his points by analysing and commenting on three studies carried out in connection with CTW productions in different stages of their development, thus offering an insight into early deficiencies and major improvements in this field over the last 15 years. 14

In connection with our publications we have always been guided by the question as to what kind of information the producer of television programmes regards as the most suitable for his work and in what form he can make the best use of it. We asked all the authors to take the special information needs of this group into account and to bring out the overview character, the introductory information, and only if necessary to mention the disputes and difficulties of interpretation existing among scholars. But we also asked them — with the programme-producing colleague as the focal point — wherever possible to draw conclusions and implications from the research findings which might be useful in planning and implementing programme projects. In conclusion, my particular thanks are due to the contributors of this book, who, in a constant fight against the time moving towards an urgent deadline, were extraordinarily helpful in complying with all our requests for providing the articles in the way we suggested, working out speedy revisions and supplementations, and cooperating in every possible way. I should also like to acknowledge the excellent work of our translators of the non-English parts (Mary-Louise Eisenberger, Norman Jones and Geoffrey P. Burwell in Munich; Charlyn Hulten in Sweden); the advice concerning stylistic improvements and proper terminology given by Konrad J. Burdach and, again, Geoffrey P. Burwell; and the editorial assistance of Rosemarie Hagemeister, reliable, competent and indefatigable as always. Munich, July 1983

Manfred Meyer

References Adler, R.P., Lesser, G.S., Meringoff, L.K., Robertson, T.S., Rossiter, J.R. and Ward, S. (Eds.) The effects of television advertising on children: Review and recommendation. Lexington, MA: Lexington Books, 1980. AIlouche-Benayoun, B.J. The influence of the moving pictures on children and young people: French research activities from 1970 to 1975. Fernsehen und Bildung, 1975, 9 (Special English Issue), 137—157. Anderson, C.M. In search of a visual rhetoric for instructional television. A V Communication Review, 1972, 20 (1), 43—63. Anderson, D.R., and Lorch, E.P. Looking at television: Action or reaction? In 15

J. Bryant and D.R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Anderson, D.R., and Smith, R. Young children's TV viewing: The problem of cognitive continuity. In F.J. Morrison, C. Lord and D.F. Keating (Eds.), Advances in Applied Developmental Psychology. New York: Academic Press, in press. Baggaley. J., and Duck, S. Dynamics of television. Westmead: Saxon House, 1976. Baggaley, J., Ferguson, M., and Brooks, P. Psychology of the TV image. Westmead: Gower, 1980. Baggaley. J., and Sharpe, J. (Eds.) Proceedings of the Second International Conference on Experimental Research in Televised Instruction. St. John's: Memorial Univ. of Newfoundland, 1979. Baggaley, J. (Ed.) Proceedings of the Third International Conference on Experimental Research in Televised Instruction. St. John's: Memorial Univ. of Newfoundland, 1980. Chu, G.C., and Schramm, W. Learning from television; What the research says. Washington, D.C.: National Assocation of Educational Broadcasters, 1968. Coelho,G.V. (Ed.) Television as a teacher: A research monograph. Rbckville, Md.: National Institute of Mental Health, 1981. von Feilitzen,C., Filipson, L., and Schyller, I. Open your eyes to children's viewing: On children, TV and radio now and in the future. Stockholm: Swedish Broadcasting Corporation, 1979. Gulliford, N.L. Current research on the relative effectiveness of selected media characteristics. Report prepared for Westinghouse Electric Corporation, Pittsburgh, Pa. 1973. Huston-Stein, A.C., and Wright, J.C. Children and television: Effects of the medium, its content, and its form. Journal of Research and Development in Education, 1979, 13, 20—31. McLuhan, H.M. Understanding media: The extensions of man. New York: McGrawHill, 1964. Meyer, M., Krusenstjern, F.v., Nissen, U., and Huth, S. Lernen durch Fernsehen in Schule und Ausbildung. Muenchen: Saur, 1980. Mielke, K.W. Asking the right ETV research questions. Educational Broadcasting Review, 1968, 2, 54—61. Mielke, K.W. Media-message interactions. Viewpoints, 1970, 46, 15—31. Mielke, K.W. Renewing the link between communications and educational technology. AV Communication Review, 1972, 20, 357—400. Murray, J.P. Television and youth: A review of research, commentary and controversy. In J. P. Murray, Television and youth. Boys Town: The Boys Town Center for the Study of Youth Development, 1980. Palmer, E. L. Formative research in the production of television for children. In G. Gerbner, L. P. Gross, W. H. Melody (Eds.), Communications technology and social policy. New York: Wiley, 1973. Palmer, E. L. Formative research in the production of television for children. In D. Olson (Ed.), Media and symbols: The forms of expression, communication and education. Chicago: Univ. of Chicago Press, 1974.

16

Palmer, E. L., and Dorr, A. (Eds.) Children and the faces of television: Teaching, violence, selling. New York: Academic Press, 1980. Pearl, D., Bouthilet, L., and Lazar, J. (Eds.) Television and behavior: Ten years of scientific progress and implications for the eighties. Rockville, Md.: National Institute of Mental Health, 1982. Rice, M. L., Huston, A. C., and Wright, J. C. The forms of television: Effects on children's attention, comprehension, and social behavior. In D. Pearl, L. Bouthilet, and J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties. Rockville, Md.: National Institute of Mental Health, 1982. Rydin, I. Information processes in pre-school children: 1. How relevant and irrelevant verbal supplements affect retention of a factual radio programme. Stockholm: Swedish Broadcasting Corporation, 1972. Rydin, I. Children's understanding of television: Pre-school children's perception of an informative programme. Stockholm: Swedish Broadcasting Corporation, 1976. Rydin, I. The tale of the seed: Facts and irrelevant details in a TV-programme for children. Stockholm: Swedish Broadcasting Corporation, 1976. Salomon, G. Can we affect cognitive skills through visual media: Explication of a hypothesis and initial findings. A VCommunication Review, 1972,20(4), 401—423. Salomon, G. The language of media and the cultivation of mental skills. The Hebrew University, 1977. (b) Salomon, G. Interaction of media, cognition, and learning. San Francisco: JosseyBass Publishers, 1979. (b) Salomon, G. Introducing AIME: The assessment of children's mental involvement with television. In H. Kelly, and H. Gardner (Eds.), Viewing children through television. San Francisco: Jossey-Bass, 1981. Salomon G., Eglstein, S., Finkelstein, R., Finkelstein, I., Mintzberg, E., Malve, D., and Veiner, L., Educational effects of Sesame Street on Israeli children. New York: Children's Television Workshop, 1972. Schramm, W. What the research says. In Schramm, W. (Ed.), Quality in instructional television. Honolulu: Univ. Press of Hawaii, 1972. Sturm, H. Die kurzzeitigen Angebotsmuster des Fernsehens. (The short-time programme patterns of television.) Fernsehen und Bildung, 1975, 9, 39—50. Sturm, H. Der Vielseher im Sozialisationsprozess. Rezipientenorientierter Ansatz und der Ansatz der formalen medienspezifischen Angebotsweisen. (The heavy viewer in the socialization process. Recipient-orientated approach and the approach of media-specific programme patterns.) Fernsehen und Bildung, 1981, 15, 137-149. Sturm, H., Vitouch, P., Bauer, H., and Grewe-Partsch, M. Emotion und Erregung — Kinder als Fernsehzuschauer. Eine psychophysiologische Untersuchung. (Emotion and arousal — Children as television viewers. A psychophysiological study.) Fernsehen und Bildung, 1982, 16, 11—114. Surgeon General's Scientific Advisory Committee on Television and Social Behavior. Television and growing up: The impact of televised violence. Washington, D. C.: United States Government Printing Office, 1972.

17

Part I

Children and the formal attributes of television: Findings of research, implications for production

Mabel L. Rice, Aletha C. Huston, John C. Wright

The forms of television: Effects on children's attention, comprehension, and social behavior*

Introduction Children like to watch télévision. It is a family entertainment medium equally accessible to children and to adults. Much of the appeal of television is attributable to its form, the manner in which it communicates its messages. It is a visual medium unlike print, still pictures or oral descriptions. On television, a stream of constantly changing images can be generated by techniques that are not replicated in real-world experience. Camera cuts, pans across scenes, zooms in and out, slow and fast motion, and special effects of all kinds are used in unique ways - musical accompaniments, sound effects, unusual cries and noises, canned laughter, and faceless narrators. Finally, of course, television is a verbal medium. The verbal and linguistic conventions of television are not unique. They are the language conventions of the real world, but the ways in which language is used to convey content (rather than the content it conveys) is an important formal property of the television medium. In the past few years, the attention of researchers studying television's influence on children has returned to the forms of the medium itself as distinct from the content presented with those forms. The approach is indebted to the seminal ideas of Marshall McLuhan (1964) who suggested that television contained representational codes fundamentally different from those of print. McLuhan's ideas remained a vague formulation, until Salomon (1979) and Huston-Stein and Wright (1977) began to elaborate the implications of that notion for developmental theory. Salomon focused particularly on the influence of visual media codes on children's mental processing and mental skills. Huston-Stein and Wright attempted to place television forms (both visual and auditory) in the context of a broader theory of developmental change in patterns of attention and information processing. In 1978, we established the Center for Research on the Influences of Television on Children (CRITC). Research in the Center has been concerned with several questions about the relationship of the forms of television to * Authors' note: Much of this research and the preparation of this review were supported by a grant to CRITC from the Spencer Foundation.

21

children's attention, cognition, and social behavior. An initial thrust was to describe the production conventions of the medium, and how those conventions are used to convey different kinds of content. Taxonomies of formal features and linguistic forms were developed. Current work includes descriptions of categories of program content and the formal features associated with fantasy/reality judgments of content. Concomitantly, we have completed a series of laboratory studies in which we investigated children's responses to television forms: their visual and auditory attention, their cognitive processing, and their social behavior during and after viewing. A study in progress investigates how children's viewing history influences their responses to television forms. Families of 320 preschool children recorded their home viewing at six month intervals for 2 1/2 years. Children then participated in a series of laboratory studies. The data analysis in progress is designed to determine the relationship between patterns of children's home viewing and their responses to and understanding of television form and content. The practical concerns of producing effective educational programming have been a second impetus for research on television form. Much of this work has been carried out in conjunction with the Children's Television Workshop productions, Sesame Street and The Electric Company. The goal is to identify the program attributes or production techniques that are maximally effective in gaining and holding children's attention and in communicating information to them in ways they will understand and remember. A third reason for studying form has been increasing suspicion that many effects attributed to television content may be partially due to the forms in which the content is presented. We have argued, for example, that some of the aggression-arousing effects of violent television may be a result of the high levels of what in the States in called "hype," a term that connotes stimulus properties (high salience, such as rapid cuts, visual special effects, and loud noise) and arousing effects on viewers (stimulating, exciting). High hype production features typically accompany violent content. Similarly, most of the research on the effects of prosocial television on very young children has used Mister Rogers' Neighborhood, a program notable for its slow pace, gentle style, and unusual language forms. It is possible that some of the positive effects of that program may be a function of its verbal and nonverbal formal properties. One may question whether it is possible to distinguish crisply between content and form. Although they can be defined independently, we acknowledge that in practice the forms of television and content messages co-occur in systematic ways. The relationship is probably analogous to that between grammar and 22

meaning in verbal language. Among linguists there is a growing belief that the grammar of verbal language cannot be isolated from semantic meanings (e.g., Fillmore's case grammar, 1968). To the degree that form and content are confounded in the real world, all studies of television content are subject to the criticism that their results may be partially a function of the forms in which that content was presented. Theory and research focusing on form independent of content may redress the imbalance so that their interactive effects can be better understood.

Representational codes of television Verbal and nonverbal forms are the representational codes of television. Because children view television at a very early age, it is tempting to assume that these representational codes are simple and of little interest. However, television is a medium that can be processed at differing levels of complexity. There is a difference between superficial consumption of interesting audiovisual events and mental extraction of information from coded messages, a distinction formulated by Salomon (1979). He used the term "literate viewing" to refer to " a process of information extraction by the active negotiation of the coding elements of the message" (p. 189). The notion of "literate viewing" is closely related to the more informal term "media literacy." With age and viewing experience, children's attention to, and comprehension of, television programs change (e.g., Collins, 1979; Krull and Husson, 1979; Calvert, Huston, Watkins, and Wright, 1982). It is presumed that these developmental changes reflect increasing facility with television's conventions and content, i.e., the beginning television viewer is not "media literate" but instead gradually acquires such competence as a function of experience with the medium and the attainment of certain minimal cognitive abilities. Just because one can become a literate viewer at an early age and without conscious effort does not demonstrate that the task is simple. The representational codes of television range in complexity from literal visual depiction to the most abstract and arbitrary symbols, including verbal language and audiovisual metaphor. The child's task is not an easy one. The change from infants' sensory-motor awareness of alternations in patterns of visual and verbal stimuli (Hollenbeck and Slaby, 1979) to the literate viewing skills of elementary school-age children involves a major qualitative advance, accompanied by developmental growth in related perceptual and cognitive skills.

23

Levels of

representation

The simplest level of representation is literal visual and/or auditory portrayal of real-world information, e.g., a shot of a car moving on the highway. A child's ability to process this level is presumably dependent primarily on perceptual and cognitive skills used in interpreting real-world stimuli. But even at this literal level, object recognition at unusual angles of viewing, lighting, and distance requires perceptual generalization and constancies not yet fully developed in the youngest viewers. On the second level of representation are media forms and conventions that do not have an exact real-world counterpart, some of these, such as cuts and zooms, are analogs of perceptual experience. For example, a zoom-in is a perceptual analog of moving close to an object. Other media conventions are more distinct from real-world experience. Dissolves, slow motion, musical accompaniments, sound effects, and electronically generated visual special effects are relatively specific to film and television. These features provide a structure for the presentation of content in a manner analogous to syntax in language. A literate viewer must be able to decode the structural meanings of formal features. For example, fades and dissolves often indicate major transitions in time, place, or content; cuts are more often used for minor shifts from one character or viewing angle to another {Huston, Wright, Wartella, Rice, Watkins, Campbell, and Potts, 1981). In American children's programs, distinctive visual "markers" are used to separate programs from commercials; the literate viewer must understand their function. This understanding ist not automatic. For example, 5- and 6-year-olds did not understand the meaning of separators between programs and commercials in one recent study (Palmer and McDowell, 1979). Media codes can also serve as models for mental representation or mental skills. That is, the child can adopt the media forms as modes of representations in her own thinking. Salomon (1979) has demonstrated, for example, that children can learn to analyze a complex stimulus into small parts by observing camera zooms in and out. Apparently, the camera provided a model of the mental process of focusing on specific parts of the stimulus. Media codes can be internalized as forms of mental representation, as suggested by McLuhan (1964), so that people can think in moving pictures with flashbacks, fast and slow motion, changes from color to black and white, and other media conventions (Salomon, 1979). The forms of television can also take on connotative meaning, either because of their repeated association with certain content themes or because of their metaphorical similarity to real-world objects and symbols. For example, rapid action, loud music, and sound effects are often associated with violence in 24

children's programs (Huston et al., 1981). Commercials for masculine sextyped toys are made with high action, rapid cuts, and loud noise, whereas feminine sex-typed toys are advertised with fades, dissolves, and soft music {Welch, Huston-Stein, Wright, and Plehals, 1979). The forms themselves may come to signal violence or sex typing to children, even when the content cues are minimal or nonexistent. Two studies demonstrated that children as young as five or six understand that productions using loud noise, fast pace, and high action have masculine connotations while soft music, frequent fades and dissolves connote femininity. This is evident for content-reduced synthetic ads {Greer, Huston, Welch, Wright, and Ross, 1981) and for professionally produced naturalistic ads {Leary and Huston, 1983). The third level of representation consists of symbolic forms not unique to the medium. Such forms may be nonlinguistic (e.g., a red stoplight) or linguistic. It is also possible for verbal language to encode forms at the other two levels. For example, dialog can encode the literal representation of reality (the first level), as when a speaker describes on-screen objects or events, or dialog can encode the conventional significance of a production feature (the second level), as when a fade is accompanied by the line "Once upon a time, long, long ago. . . ." In this sense of double encoding, it is possible for the first two levels to be nested in the linguistic codes. Such piggybacking of representational means could aid children in understanding the message and also, by association, facilitate their mastery of the codes themselves (cf. Rice and Wartella, 1981). It is apparent that the second and third levels of representational codes found in children's television programs not only have different surface characteristics but also are derived from different sources or experiences. The second level of representation, specific knowledge, is probably acquired largely as a function of experience with the medium. That is not the case with the third level, where symbols are shared by the wider culture. By definition, these codes have currency outside the medium of television and can be learned without viewing television. They also have a different utility in the world, leading to slightly different reasons for investigating them. The media-specific codes are important insofar as they reveal what is involved in a child's processing of televised information. The verbal language of television is of special interest insofar as it contributes to a child's processing of televised messages and other media codes and also, perhaps more importantly, as it serves to facilitate a child's mastery of the general linguistic code (cf. Rice, in press, a; in press, b). While the representational functions of the linguistic system have been described by linguists in a long research tradition, the production 25

conventions, or codes, of television have only recently come to the attention of behavioral scientists. The first step in understanding these codes and their functions is to develop descriptive taxonomies for formal features and to describe the ways in which they are used in television productions. Most descriptions of formal features have been developed for the purpose of studying television's influence on children. This is not to imply that formal features are without relevance for adults or that studies conducted with adult subjects are without implications for understanding children's television viewing experiences. A general discussion of how formal features may influence adult viewers is, however, beyond the scope of this review; only those studies immediately pertinent to child-directed issues and investigations are presented. Readers interested in the effects of television forms on adult audiences may wish to refer to television and film broadcasting and production publications, where issues of form are often discussed in regard to editing techniques. For example, Messaris et al. (1979) argue that editing techniques (the sequence and composition of visual shots) influenced how adult audiences perceived the nature of the interchanges between Carter and Ford during the televised 1976 presidential debates. Descriptive analyses of television

forms

Two groups have attempted descriptions of the occurrence and co-occurrence of formal features in existing television programs. One such analysis of adult programs was based on the information theory construct, "entropy" ( Watt and Krull, 1974). Entropy or form complexity was defined by the "variability" of sets, characters, and speakers and by the "unpredictability" with which each set, character, or speaker might appear next. Operationally, the entropy measure included the number of different sets, characters, and speakers, and the amount of time during which each of these appeared in the program. These investigators coded a sample of adult programs and demonstrated by factor analysis that the formal features of the programs could be clustered in two major groupings: "dynamism" (roughly the rate of change in scenes and characters) and "unfamiliarity" (roughly the variability or number of different scenes and characters). In a later study, these investigators found that form complexity was correlated with violent content in prime time programs (Watt and Krull, 1977). Wartella and Ettema (1974) used the same coding system on a set of commercials designed for children and adults but found that the two factors emerging were visual and auditory features. Formal features of children's television programs have been analyzed in our work (Huston et al., 1981) to determine what features co-occur, what features 26

characterize animated and live programs, and how formal features differ as a function of target audience or production goals. In two samples of children's programs selected from Saturday morning, prime time, and daytime educational programming (primarily PBS), action (physical activity of characters), variability of scenes (number of different scenes), and tempo (rate of scene and character change) were grouped with visual special effects, rapid cuts, loud music, and sound effects. This package of features was labeled "perceptually salient" because it was characterized by high intensity, rapid change, and rapid motion. Commercial programs for young children are packed with these perceptually salient forms. Although such formal features are more frequent in animated than in live shows, Saturday morning live programs have higher rates of perceptually salient features than prime time or educational programs. This pattern of heavy reliance on perceptual salience suggests an image of the child in the minds of producers as a being whose attention must be captured and held by constant action, change, noise, and visual onslaught. Although much of what children watch is family adult programming, these children's programs may be particularly important developmentally because they constitute the child's earliest experience with the medium. They may set the standard for what the child expects from television. In addition, they are less likely than adult programs to be mediated or buffered by parents' or adults' viewing with the child. We do not know what effects early experience with heavily saturated television "hype" and violence has on later development, later viewing patterns, or on tastes and preferences in the medium, but these questions are critically important for future research. Educational programs for young children use some perceptually salient visual features that characterize Saturday morning programs, though at more moderate levels. They combine these features, however, with other forms that have considerable potential for helping children to understand, rehearse, and remember a message. These include child dialog — probably the best form of speech to gain and hold children's attention — as well as songs, long zooms, and moderate levels of physical activity. All of these features provide opportunities for reflection, rehearsal, and review of content. Songs are frequently used to repeat themes and as a device for helping children to rehearse. Long zooms involve slow presentation and/or emphasis of important content. Because young children often understand content that is demonstrated in action, the moderate levels of action may be a particularly important means of conveying information in a form that is interesting and comprehensible to a young child. Educational programs package their content in a set of forms that is quite different from commercial programming for 27

children, and they appear to be designing programs that have good potential to hold attention and to communicate a message effectively. The findings concerning forms in children's programs can also be seen from the perspective of media literacy and its antecedents ( Wright and Huston, 1981). Recall that Saturday morning cartoons were characterized by high levels of action, variability, and tempo. These clusters consist of perceptually salient events, such as physical activity, music, sound effects, scene changes, and visual special effects. The conspicuous nature of these features may allow the forms themselves to become the message. That is, the child may pay more attention to how the information is conveyed then to what the message is, especially when the plot lines are thin to begin with. Unnoticed in the entertainment value of the features is the tutorial nature of the experience. The child is receiving explicit cues about how messages are communicated on television. In this case, the relationship between form and content is the opposite of the usual assumption. That is, the forms overpower the content (from the young viewer's perspective), whereas the problem is usually regarded as a matter of the content controlling the form (from the producer's perspective). Linguistic

codes

The coding systems inherent in verbal language constitute another component of the forms of television. In television programs, verbal language is a code within a code. Descriptive studies of the language of children's television can provide information for two purposes: (1) Knowing the nature of television's linguistic conventions or codes and how they interact with other forms of communication in children's programs is a critical part of any attempt to understand how children process televised information; and (2) analysis of television's linguistic codes may show how they are adjusted in different programs to different levels of linguistic competence in the viewer and therefore how they may, under certain conditions, play an important role in furthering language acquisition itself. In a pilot study of the linguistic structure of children's programming in relation to formal feature use, Rice (1979) analyzed 25 categories of linguistic coding in six programs. The programs represented animated stories with high, low, and no dialog; a live program representing situation comedy; and educational programs differing in age of intended audience and format (Mister Rogers' Neighborhood and The Electric Company). Three sets of linguistic descriptors were scored: (1) "Communication flow" consisted of measures of length, variability, rate, and repetition of utterances; (2) "language structure" contained measures of grammatical completeness, 28

descriptive qualifiers, and stressed single words; (3) "meaning/content" variables included focusing (i.e., giving selective prominence to a particular linguistic constituent), nonliteral meanings, explicit instructions, novel words, and immediacy of reference. Distinctive patterns of language usage were evident in the two educational programs. Mister Rogers' Neighborhood, the educational program for preschoolers, presented a moderate pattern of verbal communication: a moderate amount of dialog, without the use of nonliteral meanings or novel words, combined with moderate amounts of focusing and some use of stressed single words. The Electric Company, an educational program designed for early school-age children, used the most dialog of all the shows sampled and incorporated techniques for drawing attention and interest to dialog (e.g., focusing, stressed single words, novel words, nonliteral meanings) while at the same time adjusting for easier comprehension of grammatical forms (e.g., short comments, partial grammatical units, low variability in length) and content (e.g., reference to immediately present events). While it is widely recognized that the purpose of The Electric Company is to enhance children's reading skills, the fact that it does so by means of intensive verbal presentation is generally overlooked. Both Mister Rogers' Neighborhood and The Electric Company used techniques that are likely to facilitate children's comprehension of language (stressed single words, focusing), but the latter also used a more complex pattern of verbal presentation designed to challenge the more linguistically competent schoolage viewer. Unlike the educational programs, the commercial programs containing dialog showed little evidence that language codes were adjusted to the level of the child viewer. Some contained frequent nonliteral meanings and little focusing. The situation comedy was particularly high in descriptive qualifiers and nonreferential content, and it did not share any distinctive language features with the other shows. Comparison of the linguistic features with the formal production features of the six programs revealed that the shows with low amounts of dialog were high in action, pace, cuts, fades, zooms, visual special effects, vocalizations, sound effects, and music. All of these production features are perceptually salient ones that attract and hold visual attention in young viewers. The two verbally complex shows contained some distinctive uses of salient formal features: one had very high pace, frequent cuts, pans, and background music; the other had a high number of vocalizations. Such findings suggest a continuum of difficulty of representational coding in this range of children's programs. We would expect linguistic coding to be 29

more difficult for young viewers than the perceptually salient visual and auditory nonverbal codes. The packaging of American cartoons seems well suited to young children of limited media or linguistic competence. Similarly, the simple, comprehensible speech in Mister Rogers' Neighborhood is well suited to a preschool audience. More complex packaging in shows aimed at an older audience requires considerable linguistic sophistication and comprehension of distinctive uses of formal features. In some cases, the codes are judiciously mixed in packages of information presentation well suited to the communicative competencies of the intended audience. A moderate level of complexity may be important to maintain interest among older relatively sophisticated viewers. Just as the conventional meanings of production features can be suggested by exaggerated, perceptually salient presentations used to convey redundant content, so there is evidence of adjustments of the linguistic and production codes that are designed to draw attention to and clarify language forms themselves. For example, the frequent focusing operations and stressed single words on Mister Rogers' Neighborhood and Thé Electric Company serve to draw attention to the language codes. Furthermore, in these two programs, the meanings of the linguistic forms are often explicitly depicted. Frequently, the content is a visual representation of the verbal meaning, sometimes highlighted by attention-maintaining visual production techniques, such as cuts to a closer focus or different perspective. At least some children's programs appear to combine language adjustments with selective and supportive use of nonlinguistic salient features, at first to supplement and later to challenge the emerging cognitive competencies of the child viewer. The language of commercials aimed at children warrants explicit attention from research insofar as the intent goes beyond the communication of messages to the selling of products. Presumably, the effectiveness of commercials is dependent upon the nature of the linguistic codes presented (i.e., their basic understandability), their referential accuracy, and their use within the social context. Bloome and Ripich (1979) analyzed the social message units of commercials and how the messages related to plot or social context and/or the product. They found that many of the product-tied references were ambiguous in regard to certain features of products, such as the use of flavorings. Also, there was a subtle shift within commercials from using language in a social context to using language to promote products. Language served to establish the social occasion and then to lead the child to a product and its role in enhancing the social occasion.

30

The influence of television forms on children's mental processes When children watch television, they can just sit passively and stare at the set if they choose, but a growing body of empirical evidence suggests that this is not the usual level of response. Instead, children are more likely to become involved in the viewing experience, to work at extracting information from coded messages, to respond cognitively, affectively, and socially to program content. They are mentally and socially active viewers (Wright et al., 1978; Singer, 1980). At least some (if not most) of their mental responses are influenced by how the information is packaged, i.e., the media-specific and general representational codes employed (Rice and Wartella, 1981). The ones for which there is empirical evidence are discussed here: children's visual attention while viewing, and their understanding of television forms, program events, and relationships among characters. Formal features and

attention

Visual attention to television forms Studies using different types of programs found that certain production features or program attributes attract and hold children's visual attention while viewing television (Anderson et al., 1979; Anderson and Levin, 1976; Anderson et al., 1977; Bernstein, 1978; Susman, 1978; Wartella and Ettema, 1974; Wright et al., 1980; Rubinstein et al., 1974). Even though different systems of scoring production features have been used, there is consistency in the findings: 1. Auditory features, such as lively music, sound effects, children's voices (but not adult dialog), peculiar voices, nonspeech vocalizations, and frequent changes of speaker attract and hold children's attention. 2. Conventional visual features, such as cuts, zooms, and pans have less influence, but visual special effects do attract children's attention. 3. In most studies, high levels of physical activity or action elicit and maintain children's attention. 4. Changes in scene, characters, themes, or auditory events are especially effective in eliciting attention, though they are less important for maintaining it once the child is looking. 5. Features that lose children's attention include long complex speeches, long zooms, song and dance, men's voices, and live animals. Auditory

attention

The finding that auditory events, action, and change elicit and hold children's visual attention, while visual features have less influence, serves to remind us 31

that audition and vision interact in a complex manner during information processing. While there is considerable evidence describing visual attention, little information is available describing auditory attention (or the interaction of the two modalities) while viewing television. Any general conceptual model of how children attend to television (including the factors that are proposed as controlling attention) must take into account both visual and auditory attention. The measurement of auditory attention while maintaining a naturalistic viewing situation has been a challenging experimental problem. Looking behavior can be recorded directly in a reliable and unobstrusive manner; listening is a private mental event without easily observable indicators. We have devised a promising method for measuring auditory attention independently of visual attention. The procedure involves an intermittant degrade of the television picture or sound. As the child views, she can press a lever to restore the picture or sound track when it degrades. The child's responses are automatically recorded and time-referenced to the program. Initial findings with preschool children suggest a close relationship between looking and listening (Rolandelli, Wright, and Huston, 1982). In addition to direct measures, auditory attention can be inferred by testing comprehension of material presented in the auditory modality or material presented when the child is not looking at television. Repeated findings that children receive and understand fairly complex messages from exposure to Mister Rogers' Neighborhood, despite low rates of visual attention, have led to speculation that children were often listening even when they were not looking (Tower et al., 1979). Obviously, auditory attention can facilitate comprehension only for material that is presented in an auditory modality, usually speech. Studies in our laboratory, involving microanalysis of short time intervals within a program, indicate close connections among visual presentation of content, visual attending, and recall (Calvert et al., 1982). Similar precision in specifying the mode through which content is presented would be required to infer that auditory attention mediated comprehension. Auditory attention can also be inferred by observing visual attention to the screen (or lack thereof) and by observing what children talk about while viewing. If they are talking about things unrelated to the television content, they are probably not listening. Even if they are looking at the set, their attention may be only at the level of monitoring instead of active processing. On the other hand, auditory features, such as foreground music and children's speech, recruit visual attention for children who are looking away from the screen — evidence that some form of auditory processing is taking place.

32

Form and content interactions One of the original reasons for our interest in television form was the hypothesis that formal features in children's television were more important determinants of attention than violent content. The relative contributions of form and violent content are difficult to disentangle because conventions of production lead to correlations of certain forms with violence. Violence in children's programs is usually portrayed with high levels of action and salient auditory and visual features (Huston et al., 1981). Yet, formal features can be separated conceptually and operationally from violent content. In two studies of preschoolers, we selected programs that were high in both action and violence, or high in action and low in violence, low in action and high in violence, or low in both action and violence. Children's total attention differed as a function of action, not of violence. That is, they were as attentive to high action without violence as they were when it accompanied violence, and less attentive to low action (Huston-Stein, Fox, Greer, Watkins, and Whitaker, 1981; Huston, Wright, and Potts, 1982). A more molecular analysis was performed for these three programs and for four other cartoons by dividing each program into 15-second intervals and correlating attention with formal features and violent content. Multiple regressions were performed to determine which features were the best predictors of attention in each program. Violence did not enter any of the seven multiple regressions as a predictor that contributed significant variance independently of formal features, but considerably more data on different programs and different age groups are needed to establish the generality of this null conclusion (Huston-Stein, 1977). Form, content, and viewership ratings The relation of form and content to children's interest in television programs has also been studied by analyzing feature occurrence rates in nationally broadcast television programs in relation to national audience ratings for different ages, sexes, and regions of the country. For a sample of 34 Saturday morning programs, high action and violent content were predictors of viewership for preschool children. Each made an independent contribution. Among children from age 6 to 11, variability and tempo were the best predictors of viewership ( Wright et al., 1980). In a similar analysis of general adult audience ratings in relation to violent content of prime time adventure programs, violence accounted for a minuscule and nonsignificant portion of the variance in viewership (Diener and DeFour, 1978).

33

How formal features influence

attention

Salience and informativeness Basic research on young children's attention indicates that perceptual salience of the stimulus environment is one determinant of attention. The attributes of a stimulus that make it salient include intensity, movement, contrast, change, novelty, unexpectedness, and incongruity (Berlyne, 1960). Many of the production features that attract and hold young children's attention fit these criteria defining perceptual salience. We have proposed a developmental model hypothesizing that perceptual salience is a particularly important determinant of attention for very young viewers and/or for viewers with little media experience (Huston-Stein and Wright, 1977; 1979). The theory guiding our work was derived from the more general theoretical work of Wright and Vlietstra (1975) concerning developmental change from "exploration" to "search" in children's modes of information getting. Exploration as a mode of response is governed by the most salient features of the stimulus environment. It involves short duration, discontinuous, and impulsive responding to whatever features of the environment are perceptually dominant from moment to moment. Habituation to the salient features of a particular stimulus environment occurs as one becomes more familiar with it. Application of this model to television experience leads to the hypothesis that, among the youngest and least experienced viewers, the viewing experience consists of the consumption of perceptually salient events as entertainment in their own right. The child's attention is controlled primarily by feature salience. Until the powerful effects of salience have partially habituated, the child is essentially a passive consumer of audiovisual thrills and does not engage in deeper levels of processing ( Wright and Huston, 1981). Consummatory stimulus-controlled exploration gives way in familiar contexts to perceptual search, a kind of information getting in which the activity is instrumental, rather than consummatory, active rather than passive, and guided by the child's desire to abstract information, rather than by just entertainment, from perceived events. The child's progress from perceptual exploration to perceptual search is believed to be as much or more a function of familiarization through experience and habituation as it is a consequence of cognitive maturation, though, of course, the two are usually confounded. Thus, the older and more experienced viewers are more interested in the content of program and its meaning and less responsive to salient formal features. When older children do attend to formal features, they may use them as syntactic markers to develop a structural framework in which to 34

organize and integrate their comprehension of content meaning (Huston and Wright, 1983). Singer (1980) also proposed that high rates of salient audiovisual events on television absorb children's attention, not only because they are perceptually interesting, but because they are affectively involving. His theory does not, however, contain the proposition that developmental shifts will occur as consequences of cognitive development and familiarity with the medium. Instead, he seems to imply that extensive exposure to salient features in the medium will inhibit other forms of interest (e.g., books and verbal media) and will leave the child focused on the absorbing stimulus features of the moving picture on the screen. Studies comparing attention patterns of preschool children (age 4—6) with those of children in middle childhood (age 8—10) have provided minimal support for the hypothesis that younger children are more attentive to salient formal features than are older children (Calvert et al., 1982; Wartella and Ettema, 1974; Wright et al., in press). Both age groups attend to high levels of action and audiovisual "tricks" (visual special effects, sound effects, and unfamiliar scenes). Although these studies are consistent with the notion that younger children's attention is affected by the perceptual salience of television's formal features, they suggest that many of these features serve other functions as well in the child's processing of televised information. The complementary hypothesis is that older children's attention is guided more by the informativeness of features. Informativeness depends on the program context and the child's level of processing. For example, in a study comparing high and low pace programs, eight to ten-year-old children patterned their attention according to the length of scenes so that their average duration per look at the screen was longer in low pace programs (with long scenes) than during high paced programs (with short scenes). Younger children (5- to 7-year-olds) did not show this pattern ( Wright, Huston, Ross, Calvert, Rolandelli, Weeks, Raeissi, and Potts, in press). These findings suggested that the older children used the formal cues in the program to determine natural breaks between scenes with more skill than the younger children did. When children try to follow a plot or engage in a logical search for meaning, they probably attend to features that provide cues about time sequences, locations, characters, and events in the program. Studies by Krull and Husson (1979), in fact, suggest that older children may attend to form cues that signal content and form changes during the upcoming 1 or 2 minutes. Preschool children did not show these anticipatory patterns of attention to formal cues. Media literate children may learn temporal associations so they can anticipate 35

what will occur in a program. Older children also attend differentially to informative action and signals associated with scene changes, bit changes, and changes to and from commercials. Comprehensibility A somewhat different perspective on the relationship between attention and formal features is proposed by Anderson and his associates, who link attention with the comprehensibility of program content (e.g., Anderson and Lorch, 1983). They suggest that features such as animation or children's voices may serve as signals that the content is designed for children and is therefore likely to be comprehensible. Children may attend to such features, not because of the inherent qualities of the features, but because their media experience leads them to expect meaningful and understandable program content. The fundamental determinant of attention, according to this formulation, is the comprehensibility of the content. Two sets of data have been used to support this hypothesis. In one study (Lorch et al., 1979), children's attention to Sesame Street was manipulated experimentally by varying the availability of toys and distractions during viewing. Despite the fact that the nondistraction treatment produced very high levels of attention, it did not produce improved comprehension. Within the distracted group, however, the children who attended more comprehended more of the content. This finding was interpreted as demonstrating that comprehensibility guided attention rather than attention determining comprehension. In a subsequent study (Anderson, Lorch, Field, and Sanders, 1981), children attended less to a television program in which the speech was incomprehensible because it was backwards or in a foreign language than to a program with understandable speech. Although the influence of comprehensibility on attention has been tested thus far only by varying language features of programs, the hypothesis suggests that the comprehensibility of nonlinguistic formal features should affect attention through a similar mechanism. This line of research provides important evidence that very young children are actively processing content when they watch television rather than merely passively consuming audiovideo thrills. It does not, however, establish that feature salience and other noncontent aspects of television programs are unimportant influences on children's attention. In the studies varying comprehensibility, feature salience has been held constant (and fairly high). If salience were low, would comprehensibility alone hold children's attention? Again, the relatively low rates of attention usually found for Mister Rogers' Neighborhood suggest not, despite its outstanding comprehensibility. Second, the full range of comprehensibility has not been systematically explored. 36

In a study recently completed in our Center, these issues were addressed. Form and comprehensibility were varied independently. Short educational bits were constructed with identical content in animated, child-format versions and live, adult-format versions. Each format was used to produce bits varying in difficulty or comprehensibility but all were within the range of children's ability to understand. Children attended to the child-format versions more than to the adult-format versions, but they did not attend differently to the bits that were easy or difficult to understand (Campbell, 1982). It appears that there are many reasons why animated child formats may be attractive to children. Further, although complex, incomprehensible material loses children's attention in comparison to moderately easy, comprehensible material, but one cannot extrapolate that finding to conclude that very easy material would produce more attention than moderately difficult but still comprehensible content. In fact, the model to be proposed here suggests that both extremes of comprehensibility will be less likely to maintain attention than material in the middle range. Moreover, the model explicitly cautions against trying to define moderate comprehensibility as a stimulus feature without taking into account both the cognitive level and the viewing experience of the child. An integrative model of attention and development These seemingly divergent explanations of the determinants of attention can be integrated in the framework of one established model for attention and interest as a function of familiarity and complexity {Hunt, 1961). That model is illustrated in Figure 1. The abscissa is a compound of familiarity and complexity of both form and content. On the left end are highly familiar and oft-repeated bits, like the standard introductions and closings of familiar program series, whose informative content is minimal, and whose formal features have become habituated and no longer elicit attention among habitual viewers. The joint processes of habituation and familiarization (Wright, 1977) serve continually to depress attention on the left side of the inverted U-shaped function. By contrast, the forms and content at the high end of the abscissa are unfamiliar, complex, and incomprehensible to the child viewer. They, too, elicit little interest and attention because the child is incapable of understanding their meaning and their relation to other parts of the program. Their decoding requires comprehension of standards the child has not yet acquired and logical integration for which the child is not yet cognitively ready. They also often make reference to outside information and contextual knowledge that only adult viewers possess. Thus, attention on the right side of the curve is also low, owing to incomprehensibility. But cognitive development and the child's 37

Figure 1: Theoretical model High

.ο "

125

Viewing history influencing

emotions

The final way in which the youthful viewer may contribute to influencing his or her emotional responses to programming is the viewing history he or she has accumulated. There is some evidence that older children and adolescents with a history of more televiewing of violence and aggression are less emotionally responsive to new televised incidents of violence and aggression. This was demonstrated with children and adolescents when the viewing history considered relevant was either their own long-term viewing at home or short-term exposure provided in the laboratory by experimenters (Cline et al., 1973; Thomas, Horton, Lippincott, and Drabman, 1977). This same phenomenon of decreased emotional responsiveness following prior exposure is postulated as the explanation for a sobering finding from another study (Drabman and Thomas, 1974). In this study those older children who were exposed to televised violence prior to entering a playroom with a video monitor of children in the next room were much slower to intervene in an apparent fight in that room than were children not previously exposed to televised violence. To say that more frequent exposure to a particular type of content can diminish emotional responsiveness to a new instance of this content is not to say that all responsiveness is gone. Presumably some must remain if youth continue to view. Where else would the pleasure come from? Also, one study indicated that parents of younger children saw similar levels of emotional responsiveness in those younger children who watched a program regularly as in those who watched it infrequently {Linné, 1971). This suggests that frequency of viewing particular content is not the only, and probably not even the strongest, determinant of emotional responsiveness to that content. Moreover, it is well known that diminished emotional responses after intensive viewing of particular content can revive after a period of abstinence from that content. Thus, it is clear that viewers can become jaded to particular content but that we cannot yet explain how much this will happen, or when, or for which viewers. Viewing situation influencing

emotions

Because the situation in which youth view programming is not much under the program creators' control, this further influence on youths' emotional responses to programming will only be mentioned in passing. To the extent that program creators can anticipate or determine the circumstances in which their programming will be viewed, the information will be helpful. There are two major factors in the viewing situation that may moderate the emotional impact of programming: the physical characteristics of the setting and the 126

presence of co-viewers. The important physical characteristics are those that increase the "presence" of the program content and decrease other stimuli a darkened and quiet room, a large screen, and a color set (Himmelweit et al., 1958). The presence of co-viewers is important because it makes it possible for youthful viewers to use the co-viewers' emotional reactions as a guide to their own (cf., Himmelweit et al., 1958; Schachter, 1964). Most often we think of these as the reactions of older siblings and parents, both of whom are more likely to provide the interpretations, models of reactions, and physical comfort needed to decrease negative emotional responses to program. However, co-viewers may also enhance youths' responses to programming, often by laughing at the humorous dialogue and action (e.g., Sproull, 1973), and occasionally by themselves being fearful, angry, depressed, and the like. Implications Assuming the evoking emotional experiences in youth as they teleview is a desirable goal, the information we have presented in this section can be used by program creators to influence the amount and kind of emotions youth will experience. It is clear that all aspects of programming can be marshalled to help produce these experiences. These include the storyline, particular images and sounds, the setting, the time, the characters, roles, situations, colors, music, the point of view suggested for interpreting the content, the use of animation or live action, a humorous or more serious tone, the editing, camera angles and distances, and every other creative choice made by those who produce programs. We do not know what combinations of all these elements are most and least effective nor which choices are most crucial, but we do know that program creators should regard each and every programming decision as an opportunity to further the emotional effects they desire, including the effect of not feeling an emotion. An important point to take from the information here is that two different kinds of emotional responses can be influenced by television programming. One is probably best thought of simply as arousal level and the other as specific emotions such as fear, joy, and anger, or perhaps as positive and negative emotions. Specific emotions can be evoked by music, colors, and a humorous or serious tone, but they are probably most reliably evoked by a program's content, its story or messages. Thus, program creators should look long and hard at their storyline, content, messages, scenes, actions, characters, and dialogue to determine what emotions they suggest to viewers. They should rely primarily but not exclusively on these elements for creating particular emotions in viewers. Other production decisions, such as those about editing, camera angles and distances, pacing, music, sound effects, and 127

special effects, can then be made so as to increase or decrease viewers' arousal, thereby increasing or decreasing how much viewers feel the specific emotions associated with the program's story and messages. Among the several opportunities program creators have to modify how intensely specific emotions are felt are those that influence viewers' psychological involvement with a program's content. For non-humorous programming, emotional responses are usually greater when viewers are more psychologically involved. Such involvement can come from the use of characters, roles, situations, events, and storylines similar to those that are important in viewers' lives or to those that viewers anticipate or hope for in their lives; from the use of production techniques, settings, and time periods that enhance the realism of the content; and from the use of a point of view that emphasizes the experiences and feelings of the characters themselves rather than emphasizing a more distant, analytic, or defensive point of view. For humorous programming, the relationship between viewers' involvement in program content and their emotional responsiveness seems to depend somewhat on the type of humor in the program. To enjoy tendentious, aggressive, or ridiculing humor, viewers need some psychological distance from many if not all the characters. Otherwise, the insults, ridicule, and misfortunes would not be funny. More gentle humor, on the other hand, is probably enhanced when viewers are more psychologically involved in the program content. These opportunities to create more or less psychological involvement in viewers, and then to influence how much of any specific emotion they feel and perhaps which emotions are felt, surely vary according to the age and sophistication of the viewers. The kinds of emotions viewers will be aware of feeling and the messages, content, or storylines that will evoke them are also likely to vary by the viewers' age. If we had more research with youth of different ages, more like that reviewed in the section "Understanding emotions portrayed on television," we would have organized this "Implications" section, as that one was, by the age of the intended audience. However, we do not know enough to do that. There are, nonetheless, a few points to be made for program creators about the importance of the ways in which preschoolers, younger and older children, and adolescents are likely to differ in their emotional responses during televiewing. These points enlarge and elaborate implications we have already suggested here. It will be easy for program creators to evoke a limited number of emotions in preschoolers. These include fear, happiness, humor, and anger. They will be easiest to evoke when they are very clearly portrayed using many program elements and are presented singly rather than in combination. Fear can be 128

evoked by monsters, danger to animals, long sharp weapons, and overt aggression. Humor can be evoked by slapstick comedy, nonsense wordplay, visual incongruity, and adult foibles. Romantic or sexual content, ridicule and repartee, and implied but not portrayed actions, events, or feelings will be difficult to use to evoke emotional responses in preschoolers. Finally, program creators will find it challenging to affect preschoolers' level of psychological involvement with programming. At this age, they lack much of the knowledge and thinking ability needed to distance themselves from otherwise involving programming, and there are a limited number of characters, roles, situations, and events with which they can identify. Just as with preschoolers, it will be easy for program creators to evoke only a limited number of emotions in younger children, and these are most likely to be felt when they are portrayed clearly and one at a time. The elements that program creators can use to provoke fear and humor in younger children generally include all those described for preschoolers, plus a few of particular importance. Program creators can now use riddles, nonsense jokes, and insults to evoke laughter, and death to evoke several negative emotions. Romantic or sexual content is still not very useful to those creating programs for this age group, but implied rather than portrayed content can sometimes be used to good effect and psychological involvement with program content can sometimes be manipulated. The options program creators have for evoking emotional experiences increase considerably when the intended audience is older children. Program creators will find it quite possible to use implied rather than portrayed content and to manipulate older children's psychological involvement. They can plan to evoke several different emotions, more complex emotions, and more than one emotion at the same time or over a short period of time. The content which program creators will find most effective for evoking specific emotions is rather different from that they would use with younger children and preschoolers. In particular, monsters, nonsense words, most visual incongruities, much slapstick comedy, and most danger to animals will not be very effective. On the other hand, program creators will be able to use content such as romance, sex, impending good or bad fortune, repartee, tendentious humor, and play on literal meanings of expressions to evoke emotions in viewers of this age. Whatever program creators can do with older children's emotional responses can be more easily and often done with adolescents. Their knowledge and thinking ability are such that program creators can use a wide range of content, storylines, and production techniques to evoke emotions. Their abilities to adopt a point of view, to respond to attempts to increase or 129

decrease their psychological involvement with the program, and to infer events and feelings are such that program creators can use virtually any and all program elements to convey the intended emotions. Having described program creators' several opportunities to influence the emotions preschoolers, younger and older children, and adolescents experience while watching programs, it is fitting to end this section with the observation that emotions youth experience while viewing are not entirely within the program creators' control. Youth bring with them their own emotions and patterns of emotional responding, their own knowledge and interests, their particular abilities to think and reason, and their own reasons for choosing to watch programs. In addition, they teleview in situations that can increase or decrease the emotions evoked by programming and which at times can even dramatically alter the type of emotion experienced. In the final analysis, the emotions youth experience from programming are determined by a complex interplay among the programming itself, the youth themselves, and their viewing circumstances. By knowing everything possible about how to construct programming that takes account of the likely characteristics of youthful viewers and their viewing situations, program creators maximize their opportunities to determine what emotions, and how much of each emotion, youth will experience as they watch television programs.

Emotion-related consequences of televiewing The preceding three major sections of this chapter have focused on youths' reactions to television programming as they are viewing that programming. The gratifications they receive from the act of viewing itself, what they are likely to understand about emotions in the programming they are viewing, and the emotions they are likely to feel while viewing have all been discussed. Now we will turn to longer range reactions in five areas: Learning about emotions, increasing emotional responses in everyday life, decreasing emotional responses, altering other learning and activity choices, and choosing televiewing. Learning about

emotions

As indicated in the section "Recognition of emotions", youth have much to learn about emotions and their expression as they grow up. Television has the opportunity to contribute to this learning. Learning of four different kinds may occur: the names applied to different emotional displays, the emotions typically felt in particular situations, the display rules for when and how to express (or to hide) particular emotions in particular situations, and the 130

emotions intended by cues specific to television. All these kinds of learning are essentially normative, in that one assumes all members of the culture share a system of beliefs about these things. The interesting issue for our purposes is the extent to which television teaches about them. So far, we know little about what television does in these areas, other than that preschool viewers of Sesame Street over a year's time learned somewhat more about the emotions preschoolers normally feel in particular situations than did preschoolers who watched Sesame Street infrequently or not at all (Bogatz and Ball, 1971). Television programs have great potential to enhance youths' knowledge about emotions. Perhaps in the future we will know more about how much programming does that and how it can do it better, should one choose that as a programming goal. Increasing emotional

responses

A second long-term impact that programming may have is increasing youths' emotional reactions or displays in their everyday lives. This may occur in three ways. First, programming may connect a particular person, object, action, or situation with a particular emotional reaction that may eventually transfer to youths' everyday experiences. For instance, white preschoolers who live in segregated areas of the United States may develop positive emotional responses to black people if they see them portrayed positively and associated with objects, activities, and situations to which these preschoolers also respond positively. Alternatively, they may develop negative emotional responses to blacks if they see them portrayed as aggressive, dangerous, hostile, dirty, lazy, subservient, and the like. A personal example of this phenomenon involves the Sesame Street segment about mummies that we described in the section "Emotions experienced while viewing programs." There we noted that Dorr's son was frightened while viewing this segment. This fright transferred to his everyday life in a way that became apparent during a visit to a local museum. Once there the boy traipsed happily, though quickly, through displays of modern sculpture and of heavily embroidered costumes. Leaving the elevator at the ground floor, he immediately saw a room with mummies, large statues similar to mummies, and other such artifacts all bathed in light similar to that on the Sesame Street segment. He grabbed his mother's hand, complained loudly about how frightening the place was, demanded to leave immediately, and headed directly back to the elevator. Twice since he has been invited back to this area. The first time he flatly refused. The next he managed a hasty walk from the elevators across the room to stairs leading up to the first floor. His fear will surely not stay with him always, but right now it is a fine example of the longer term impact television programming can have on youths' emotions. 131

A second way in which programming may influence youths' emotions in everyday life is by providing a set of emotional responses to what they have seen, where the emotional responses remain for an extended period of time. The affection of viewers for specific characters is well known. We can have no doubt that these affections are experienced at least occasionally at times other than during televiewing. In fact, in a study with older adolescents and a later study with older children, emotional responses to characters and to programs per se were found to persist largely unchanged for up to three weeks, the last point at which measurements were taken (see Sturm, 1982, for a summary). We have no idea for what types of programs or characters, what emotions, or what viewers such persistence of emotional responses would be found. But such emotional reactions are clearly part of what television programming can introduce into the long-term, day-to-day experience of youth. A third way in which programming may influence youths' emotions in everyday life is by promoting a more generally positive or negative emotional approach to everyday experience. For instance, showing a loving, friendly, supportive, happy, and interested approach to the many people and events of daily life may result in youth adopting this emotional stance in more of their own lives. Alternatively, showing an aggressive, hostile, and violent world may result either in youth adopting this emotional stance in their own lives or in youth becoming more fearful because they expect such behavior from others. Naturally, one does not claim that the television programming youth view primarily accounts for their emotional reactions in everyday life. Moreover, one recognizes that television programming offers viewers an enormous variety of emotional stances. There is, nonetheless, some reason to look to television programs as a means of increasing youths' emotional responses in everyday life. Suggestive evidence comes from four sources. First, there are experimental studies that demonstrate that youth can be "trained" to respond emotionally to particular stimuli (.Bandura and Rosenthal, 1966; Berger, 1962; Venn and Short, 1973). Second, there are studies with preschoolers viewing programming produced for them that demonstrate that they can develop emotional responses to particular stimuli (Bogatz and Ball, 1971) and more general patterns of emotional responsivity such as explaining one's emotions more often (Stein and Friedrich, 1972) and showing more positive emotions to others (Coates, Pusser, and Goodman, 1976). Third, there is one study suggesting that 8-year-olds' fearfulness is partially explained by the amount of violent television programs they watched several years earlier (Singer and Singer, 1982). Fourth, several studies with older children and adolescents viewing primetime programming, which in the United States has frequent 132

examples of hostile humor, aggression, and violence, suggest that they develop a somewhat more fearful expectation about their environment if they teleview more (Hawkins and Pingree, 1981, 1982). Altogether these lines of evidence suggest that television programming may increase youths' emotional responses in everyday life either through establishing new responses or through enhancing those already weakly established. Decreasing emotional

responses

In the preceding section we examined television's success in increasing or intensifying emotional responses that will show up in everyday life. Here we examine the opposite: its contribution to the diminution or disappearance of emotional arousal when one is exposed to selected stimuli in everyday life. Here there is very clear evidence that specially constructed television programs, especially when they are used in conjunction with other experiences, can help decrease the fear, distress, anxiety, alarm, and the like that youth experience in such situations as the dentist's office and hospital and with such animals as snakes and dogs (Rushton, 1979, 1982; Zillmann, 1982). The most effective programs present the experience or animal in situations and contexts that will diminish its threatening qualities, present the experience or animal in many different situations and contexts, show several different people pleasantly involved with the situation or animal, and use people who are similar to the youth for whom the program is intended (Bandura, 1977). They also depict the positive consequences that can ensue from pleasant involvement with the situation or animal. These studies demonstrate that program creators have the opportunity to help youth feel less fearful and anxious in their everyday lives. Whether this is an appropriate goal for a program creator must surely depend on the particular emotion, situation, and/or animal being depicted, as well as the program creator's own values. For instance, our values lead us to believe it is desirable to produce programming to decrease youths' fears of the dentist, doctor, or hospital, but only insofar as those fears are unrealistic. If youth will encounter medical procedures that will be painful or from which substantial recovery will be required, then they are best served by realistically anticipating these travails. If such realistic expectations engender some fear or anxiety, that is probably appropriate. From this perspective a program creator would be ill advised to entirely do away with the youths' anxiety and apprehension. Similarly, our values lead us to believe that it is desirable for youth to feel as much distress and disgust as possible when they encounter violence and aggression.

133

From these perspectives the findings reviewed in the earlier section "Viewing history influencing emotions" are distressing. It will be recalled that viewers who have a history of viewing more aggressive programs are likely to be less emotionally responsive to it. This shows up in their diminished physiological responses while viewing and in their greater tolerance for a fight ostensibly occurring near them. A program creator, sharing our values, would obviously be ill advised for this reason to direct his or her programming toward diminishing youths' negative reactions to violence and aggression. Altering other learning and activity

choices

A fourth way in which television programs may impact youth in their everyday lives is the effect their emotional responses while viewing can have on other activities. Emotional responses are by definition involving. Thus, when youth react emotionally to television content, they are naturally more caught up in that content and less caught up in other television content or in other activities going on around them. Several studies have demonstrated this phenomenon. In some in which the emotionally involving content was irrelevant either to the central content of the program or the main task, younger children did less well on the main task or in learning the central content than they would have if the irrelevant, emotionally involving content had not been available or if they had not been in an emotional state in which that content was especially appealing (Canon, 1967; Rydin, 1973, 1976). In others which assessed the amount of information learned from programs that differed in how arousing they probably were for younger children, the findings were mixed (Bryant et al., 1983; Wakshlag, Reitz, and Zillmann, 1982; Zillmann, Williams, Bryant, Boynton, and Wolf, 1980). That is, younger children learned more from programming with humor added than from programming without it, but they learned less from programming with fast-paced background music than from programming with slow-paced background music or no music at all. Still other studies, of older adolescents and younger adults only, have shown that viewers' arousal when they stop televiewing may feed into and enhance their subsequent activities (Tannenbaum, 1972; Zillmann, 1980; Zillmann, in press; Zillmann et al., 1974). For instance, those who are more aroused after viewing are more likely to be aggressive, helpful, solicitous, humorous, or sexual should the immediately succeeding circumstances call for such behavior than are those who are less aroused. All these lines of evidence suggest that the emotions youth experience when televiewing can influence what they learn, how well they do on other tasks, and how much energy they put into subsequent social behavior. 134

Choosing

televiewing

The final consequence in youths' everyday lives that we want to mention is that of choosing televiewing over other activities. As we noted in the first section of this chapter, televiewing provides youth with many gratifications. The more gratifications programs provide, the more likely they are to be watched (Kovaric et al., 1983). The more educational programs employ humor at all, faster paced as opposed to slower paced humor, and appealing fast-paced music as opposed to slower paced or no music, the more they will be watched (Wakshlag et al., 1981; Wakshlag et al., 1982). More exciting and suspenseful programs will be chosen over others. Surely, there are enough television programs broadcast now in most industrialized countries that these gratifications are often, if not always, available to youth. Why wouldn't youth then choose televiewing as an easy source of, shall we say, "cheap thrills"? Such a consequence, of course, ultimately has implications for the variety and extent of emotions youth encounter and for the experiences that evoke them. Most of us would, we believe, value a wide range of emotion provoking experiences and a wide range of emotions experienced. Given such a value, we all seek to keep television from being too attractive and too attainable a source of emotional gratification, a source that would then keep youth from the other experiences and emotions that are also important for their development.

Conclusion There are three approaches to producing television programs. One emphasizes the creative endeavor, placing faith first and foremost in the common sense, intelligence, sensitivity, and artistry of those who will create the programs. Another emphasizes an analytic approach, placing faith in the knowledge that has been developed about how to communicate well and utilizing that knowledge to design the programming. The third approach emphasizes formative research, taking successive approximations of programs out to members of the intended audience, gauging their reactions to them, and then adjusting the programs accordingly. This chapter addresses the second approach, the analytic one. There are certain things we know about why youth teleview, what they understand about emotions, how they react emotionally to programs, and what longer term effects on emotions are possible from televiewing. In an analytic approach the assumption is that once program creators find out what is known, they can and will use this knowledge to design more effective and beneficial programming. We hope that assumption is correct, but we also hope and trust that that is not all program creators will do. 135

The need for creativity, sensitivity, and artistry in making programming for youth is enormous, and the programs that gifted persons can create are truly outstanding. Never should this source of excellent programming go unappreciated. Nor, however, should the benefits of the artistic approach completely overshadow the benefits to be derived from understanding one's audience as completely as possible. Such understanding comes through learning what is already known about the audience (the analytic approach), and through formal and informal research for the particular program one is creating (the empirical approach). All three approaches together significantly increase the chances of creating programming that youth will understand, enjoy, and benefit from — the goals of us all.

References Aronfreed, J. The socialization of altruistic and sympathetic behavior: Some theoretical and experimental analyses. In J. Macaulay and L. Berkowitz (Eds.), Altruism and helping behavior. New York: Academic Rress, 1970. Bailyn, L. Mass media and children: A study of exposure habits and cognitive effects. Psychological Monographs, 1959, 73 (1), 1-48. Bandura, A. Social-learning theory of identificatory processes. In D. A. Goslin (Ed.). Handbook of socialization theory and research. Chicago: Rand McNally, 1969. Bandura, A. Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 1977,84, 191-215. Bandura, A. H., and Rosenthal, L. Vicarious classical conditioning as a function of arousal level. Journal of Personality and Social Psychology, 1966, 3, 54-62. Bardon, R. C., Zelko, F. Α., Duncan, S. W., and Masters, J. C. Children's consensual knowledge about the experiential determinants of emotion. Journal of Personality and Social Psychology, 1980, 39, 968-976. Becker, S. L., and Wolfe, G. J. Can adults predict children's interest in a television program? In W. Schramm (Ed.), The impact of educational television. Urbana, 111.: University of Illinois Press, 1960. Berger, S. M. Conditioning through vicarious instigation. Psychological Review, 1962, 69, 450-456. Berlyne, D. D. Humor and its kin. In J. H. Goldstein and P. E. McGhee (Eds.), The psychology of humor: Theoretical perspectives and empirical issues. New York: Academic Press, 1972. Birdwhistell, R. L. Kinesics in context. Philadelphia: University of Pennsylvania Press, 1970. Bogatz, G. Α., and Ball, S. The second year of Sesame Street: A continuing evaluation. Princeton, N. J.: Educational Testing Service, 1971. 136

Brown, J. R., Cramond, J. K., and Wilde, R. J. Displacement effects of television and the child's functional orientation to media. In J. G. Blumler and E. Katz (Eds.), The uses of mass communications. Beverly Hills: Sage, 1974. Brown, Α., and Cunningham, J. G. Developmental interactions in the production and interpretation of affective facial expression. Paper presented at the biennial meeting of the Society for Research in Child Development, Boston, April, 1981. Bryant, J., Zillmann, D., and Brown, D. Entertainment features in children's educational television: Effects on attention and information acquisition. In J. Bryant and D. R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Camras, L. Facial expressions used by children in a conflict situation. Child Development, 1977, 48, 1431-1435. Camras, L. Children's interpretations of facial expressions used in conflict. Child Development, 1980, 51, 879-885. Canon, L. K. Motivational state, stimulus selection, and distractibility. Child Development, 1967, 38, 589-596. Chaffee, S. H., and Tims, A. R. Interpersonal factors in adolescent television use. Journal of Social Issues, 1976, 32, 98-115. Charlesworth, W. R., and Kreutzer, M. A. Facial expressions of infants and children. In P. Ekman (Ed.), Darwin and facial expressions: A century of research in review. New York: Academic Press, 1973. Cline, V. Β., Croft, R. G., and Courrier, S. Desensitization of children to television violence. Journal of Personality and Social Psychology, 1973, 27, 360-365. Coates, Β., Pusser, Η. Ε., and Goodman, I. The influence of „Sesame Street" and „Mister Rogers' Neighborhood" on children's social behavior in the preschool. Child Development, 1976,47. 138-144. DePaulo, Β. M., Jordan, Α., Irvine, Α., and Laser, P. S. Age changes in detection of deception. Paper presented at the annual meeting of the American Psychological Association, Montreal, September 1980. Deutsch, F. Female preschoolers' perceptions of affective responses and interpersonal behavior in videotaped episodes. Developmental Psychology, 1974, 10, 733-740. Dorr, A. No shortcuts to judging reality. In J. Bryant and D. R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Drabman, R. S., and Thomas, M. H. Does media violence increase children's toleration of real-life aggression? Developmental Psychology, 1974, 10, 418-421. Dysinger, W. S., and Ruckmick, C. A. The emotional responses of children to the motion picture situation. New York: Macmillan, 1933. Ekman, P. (Ed.) Darwin and facial expression: A century of research in review. New York: Academic Press, 1973. Ekman, P., and Friesen, W. V. Unmasking the face. Englewood Cliffs, N. J.: Prentice-Hall, 1975. Ekman, P., Friesen, W. V., and Ellsworth, P. C. Emotion in the human face: Guidelines for research and an integration of findings. New York: Pergamon Press, 1972. 137

Ekman, P., Liebert, R. M., Friesen, W. V., Harrison, R., Zlatchin, C., Malmstrom, E. J., and Baron, R. A . Facial expressions of emotion while watching televised violence as predictors of subsequent aggression. In G. A . Comstock, E. A . Rubinstein, and J. P. Murray, (Eds.), Television and social behavior (Vol. 5) Television's effects: Further exploration. Washington, D. C.: U. S. Government Printing Office, 1972. Eyre-Brook, C. (Ed.) Young people and television: An international study of juries, producers and their young audiences based on the prize-winning programmes of Prix Jeunesse 1970: Man in Metropolis and Baff. Munich: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen, 1972. von Feilitzen, C. Children and identification in the mass communication process. Stockholm: Swedish Broadcasting Corporation, Audience and Programme Research Department, 1974. von Feilitzen, C. The functions served by the media: Report on a Swedish study. In R. Brown (Ed.), Children and television. Beverly Hills: Sage, 1976. von Feilitzen, C., and Linné, O. Living habits and broadcast media behavior of 3-6 year-olds. Stockholm: Sveriges Radio, Audience and Programme Research Department, 1969. Fernie, D. E. Ordinary and extraordinary people: Children's understanding of television and real life models. In H. Kelly and H. Gardner (Eds.), Viewing children through television. San Francisco: Jossey-Bass, 1981. Feshbach, S. Reality and fantasy in filmed violence. In J. P. Murray, E. A . Rubinstein, and G. A . Comstock (Eds.), Television and social behavior (Vol. 2) Television and social learning. Washington, D. C.: U. S. Government Printing Office, 1972. Feshbach, N. D., and Roe, K. Empathy in six and seven year olds. Child Development, 1968, 39, 133-145. Field, T. M., and Waiden, T. A . Production and discrimination of facial expressions by preschooler children. Child Development, 1982,55, 1299-1312. Filipson, L. The role of radio and TV in the lives of pre-school children. Summary. Stockholm: Swedish Broadcasting Corporation, Audience and Programme Research Department, 1976. Filipson, E., Schyller, I., and Hoijer, B. Pretesting of the programme ,, Why Must We Die?" Stockholm: Swedish Broadcasting Corporation, Audience and Programme Research Department, 1974. Flapan, D. Children's understanding of social interaction. New York: Teachers College Press, 1968. Flavell, J. H. Cognitive development. Englewood Cliffs, N . J.: Prentice-Hall, 1977. Frijda, Ν. H. Recognition of emotion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 4) New York: Academic Press, 1969. Garry, R. (Ed.) Findings and cognition on the television perception of children and young people based on the prize-winning programme of Prix Jeunesse 1968: The Scarecrow. Munich: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen, 1970.

138

Greenberg, Β. S. Gratifications of television viewing and their correlates in British children. In J. G. Blumler and E. Katz (Eds.), The uses of mass communications. Beverly Hills: Sage, 1974. Harris, P. L., Olthof, T., and Terwogt, M. M. Children's knowledge of emotion. Journal of Child Psychology and Psychiatry, 1981, 22, 247-261. Hawkins, R. P. The dimensional structure of children's perceptions of television reality. Communication Research, 1977, 4, 299-320. ^ Hawkins, R. P., and Pingree, S. Using television to construct social reality. Journal of Broadcasting, 1981, 25, 347-364. Hawkins, R. P., and Pingree, S. Television's influence on social reality. In D. Pearl, L. Bouthilet, and J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (Vol. 2). Rockville, Md.: National Institute of Mental Health, 1982. Himmelweit, H. T., Oppenheim, Α. Ν., and Vince, P. Television and the child. London: Oxford University Press, 1958. Honkavaara, S. The psychology of expression. British Journal of Psychology Monograph Supplement, 1961, No. 32, 25-50, 58-67, 72-86. Huston, A. C., Wright, J. C., Wartella, E., Rice, M. L., Watkins, Β. Α., Campbell, T., and Potts, R. Communicating more than content: Formal features of children's television programs. Journal of Communication, 1981, 31 (3), 32-48. Inside/Out: A guide for teachers. Bloomington, Ind.: National Instructional Television Center (NITVC), 1973, (ERIC Document No. ED 084 864). Izard, C.E. Human emotions. New York: Plenum, 1977. Johnstone, J. W. C. Social integration and mass media use among adolescents: A case study. In J. G. Blumler and E. Katz (Eds.), The uses of mass communications. Beverly Hills, Sage, 1974. Keilhacker, M. (Ed.) Findings and cognition on the television perception of children and young people based on the prize-winning programmes of Prix Jeunesse 1966: Patrik and Putrik and Clown Ferdl. Munich: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen, 1969. Kohlberg, L. Stage and sequence: The cognitive-developmental approach to socialization. In D. A. Goslin (Ed.), Handbook of socialization theory and research. Chicago: Rand McNally, 1969. Kovaric, P., Dorr, Α., and Nicol, J. What's gratifying about different types of television programs. Unpublished manuscript, University of Southern California, January 1983. Krauss, R. M., and Morency, N. L. The nonverbal encoding and decoding of affect in first and fifth graders. Paper presented at the annual meeting of the American Psychological Association, Montreal, September 4, 1980. Laosa, L. M. Viewing bilingual multicultural educational television: An empirical analysis of children's behaviors during television viewing. Journal of Educational Psychology, 1976, 68, 133-142. Lazarus, R. S., Averill, J. R., and Opton, E. M., Jr. Toward a cognitive theory of emotion. In M. Arnold (Ed.), Feelings and emotions. New York: Academic Press, 1970. 139

Lazarus, R. S., Kanner, A. D., and Folkman, S. Emotions: A cognitive-phenomenological analysis. In R. Plutchik and H. Kellerman (Eds.), Emotion: Theory, research and experience (Vol. 1). New York: Academic Press, 1980. Lazarus, R. S., Speisman, J. C., Mordkoff, A. M., and Davidson, L. A. A laboratory study of psychological stress produced by a motion picture film. Psychological Monographs, 1982, 76 (34). Lesser, G. S. Children and television: Lessons from Sesame Street. New York: Random House, 1974. Lesser, G. S. Personal communication, 1977. Leventhal, H. Toward a comprehensive theory of emotion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13). New York: Academic Press, 1980. Linné, O. Reactions of children to violence on TV. Stockholm: Sveriges Radio, Audience and Programme Research Department, 1971. Lyle, J., and Hoffman, H. R. Children's use of television and other media. In E. A. Rubinstein, G. A. Comstock, and J. P. Murray (Eds.), Television and social behavior (Vol. 4). Television in day-to-day life: Patterns of use. Washington, D. C.: U. S. Government Printing Office, 1972. Maccoby, E. E., and Wilson, W. C. Identification and observational learning from films. Journal of Abnormal and Social Psychology, 1957, 55, 76-87. Maccoby, E. E., Wilson, W. C., and Burton, R. V. Differential movie-viewing behavior of male and female viewers. Journal of Personality, 1958, 26, 259-267. McGhee, P. E. Toward the integration of entertainment and educational functions of television: The role of humor. In P. H. Tannenbaum (Ed.), The entertainment functions of television. Hillsdale, N. J.: Erlbaum, 1980. Mandler, G. Mind and emotion. New York: Wiley, 1980. Messaris, P. The film audience's awareness of the production process. Journal of the University Film Association, 1981, 33, 53-56. Messaris, P., and Gross, L. Interpretations of a photographic narrative by viewers in four age groups. Studies in the Anthropology of Visual Communication, 1977, 4, 99-111. Mischel, W. Sex-typing and socialization. In P. H. Müssen (Ed.), Carmichael's manual of child psychology (3rd ed.). New York: Wiley, 1970. Moad, D. W., Johnson, J. E., and Shantz, C. U. Social comprehension and affectmatching in young children. Merrill-Palmer Quarterly, 1978, 24, 63-66. Murray, J. P. Television in inner-city homes: Viewing behavior of young boys. In E. A. Rubinstein, G. A. Comstock, and J. P. Murray (Eds.), Television and social behavior (Vol. 4). Television in day-to-day life: Patterns of use. Washington, D. C.: U. S. Government Printing Office, 1972. Murray, J. P., and Kippax, S. From the early window to the late night show: International trends in the study of television's impact on children and adults. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12). New York: Academic Press, 1979. Noble, G. Children in front of the small screen. London: Constable, 1975. (In the United States published in Beverly Hills, Calif., by Sage).

140

Odom, R., and Lemond, C. Developmental differences in the perception and production of facial expressions. Child Development, 1972, 43, 359-369. Pazer, S., Slackman, B., and Hoffman, M. Age and sex differences in effects of extenuating circumstances on anger. Paper presented at the biennial meeting of the Society for Research in Child Development, Boston, April 1981. Phelps, Ε. M. Knowledge of the television industry and relevant first-hand experience. Paper presented at the annual meeting of the American Psychological Association, Washington, September 1976. Roberts, D. Testimony before the Federal Trade Commission's rulemaking on children and TV advertising. San Francisco, January 1979. Rubin, A. M. Television use by children and adolescents. Human Communication Research, 1979, 5, 109-120. Rushton, J. P. Effects of prosocial television and film material on the behavior of viewers. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12). New York: Academic Press, 1979. Rushton, J. P. Television and prosocial behavior. In D. Pearl, L. Bouthilet, and J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (Vol. 2). Rockville, Md. : National Institute of Mental Health, 1982. Rydin, I. Information processes in pre-school children, I. Stockholm: Swedish Broadcasting Corporation, Audience and Programme Research Department, 1973. Rydin, I. Information processes in pre-school children, II. The tale of the seed. Stockholm: Swedish Broadcasting Corporation, Audience and Programme Research Department, 1976. Saarni, C. Children's understanding of display rules for expressive behavior. Developmental Psychology, 1979, 15, 424-429. Saarni, C. Observing children's use of display rules: Age and sex differences. Paper presented at the annual meeting of the American Psychological Association, Montreal, September 1980. Saarni, C. Social and affective functions of nonverbal behavior: Developmental concerns. In R. Feldman (Ed.), Development of nonverbal behavior. New York: Springer Verlag, 1982. Sapolsky, B. S., and Zillmann, D. Experience and empathy: Affective reactions to witnessing childbirth. Journal of Social Psychology, 1978, 105, 131-144. Schachter, S. The interaction of cognitive and physiological determinants of emotional state. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 1). New York: Academic Press, 1964. Schramm, W., Lyle, J., and Parker, E. B. Television in the lives of our children. Stanford: Stanford University Press, 1961. Shantz, C. The development of social cognition. Chicago: University of Chicago Press, 1975. Shapiro, B. N. Comprehension of television programming designed to encourage socially valued behavior in children: Formative research on "Sesame Street" programming with social and affective goals. Unpublished manuscript, University of Massachusetts, Amherst, 1975. 141

Singer, J. L., and Singer, D. G. Progress report to the Spencer Foundation: Relationships between television-viewing and the cognitive and emotional development of children. Yale University, July 1, 1982. Speisman, J. C., Lazarus, R. S., Mordkoff, Α., and Davison, L. Experimental reduction of stress based on ego-defense theory. Journal of Abnormal and Social Psychology, 1964, 68, 367-380. Sproull, N. Visual attention, modeling behaviors, and other verbal and nonverbal meta-communication of prekindergarten children viewing Sesame Street. American Educational Research Journal, 1973, 10, 101-114. Stein, A. H., and Friedrich, L. K. Television content and young children's behavior. In J. P. Murray, E. A. Rubinstein, and G. A. Comstock (Eds.), Television and social behavior (Vol. 2). Television and social learning. Washington, D. C.: U. S. Government Printing Office, 1972. Sturm, H. Emotional effects and formal features on TV: A recipient-oriented approach. Paper presented at the annual meeting of the International Communication Association, Boston, May 1982. Sutton-Smith A developmental structural account of riddles. Paper presented at the biennial meeting of the Society for Research in Child Development, Philadelphia, March 1973. Tannenbaum, P. H. Studies in film- and television-mediated arousal and aggression: A progress report. In G. A. Comstock, E. A. Rubinstein, and J. P. Murray (Eds.), Television and social behavior (Vol. 5). Television's effects: Further explorations. Washington, D. C.: U. S. Government Printing Office, 1972. Tannenbaum, P. H., and Gaer, E. P. Mood change as a function of stress of protagonist and degree of identification in a film-viewing situation. Journal of Personality and Social Psychology, 1965, 2, 612-616. Thomas, Μ. H., Horton, R. W., Lippincott, E. C., and Drabman, R. S. Desensitization to portrayals of real-life aggression as a function of exposure to television violence. Journal of Personality and Social Psychology, 1977, 35, 450-458. Thompson, R. Α., and Paris, S. G. Children's inferences about the emotions of others. Paper presented at the biennial meeting of the Society for Research in Child Development, Boston, April 1981. Venn, J. R., and Short, J. G. Vicarious classical conditioning of emotional responses in nursery school children. Journal of Personality and Social Psychology, 1973, 28, 249-255. Wakshlag, J. J., Day, K. D., and Zillmann, D. Selective exposure to educational television programs as a function of differently paced humorous inserts. Journal of Educational Psychology, 1981, 73, 27-32. Wakshlag, J. J., Reitz, R. J., and Zillmann, D. Selective exposure to and acquisition of information from educational television programs as a function of appeal and tempo of background music. Journal of Educational Psychology, 1982, 74, 666-677. Waiden, Τ. Α., and Field, T. M. Discrimination of facial expressions by preschool children. Child Development, 1982, 53, 1312-1320. 142

Wartella, E. Individual differences in children's responses to television advertising. In E. L. Palmer and A. Dorr (Eds.), Children and the faces of television: Teaching, violence, selling. New York: Academic Press, 1980. Weiner, B., and Graham, S. An attributional approach to emotional development. In C. Izard, J. Kagan, and R. Zajonc (Eds.), Emotions, cognition, and behavior. Cambridge, Mass.: Harvard University Press, in press. Weiner, B., Graham, S., Stern, P., and Lawson, M. E. Using affective cues to infer causal thoughts. Developmental Psychology, 1982, 18, 278-286. Werner, H. Comparative psychology of mental development. New York: Science Editions, 1948. Winner, E., Wapner, W., Cicone, M., and Gardner, H. Measures of metaphor. In E. Winner and H. Gardner (Eds.), New directions for child development: Fact, fiction, and fantasy in childhood (No. 6). San Francisco: Jossey-Bass, 1979. Zillmann, D. Anatomy of suspense. In P. H. Tannenbaum (Ed.), The entertainment functions of television. Hillsdale, N. J.: Erlbaum, 1980. Zillmann, D. Television viewing and arousal. In D. Pearl, L. Bouthilet, and J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (Vol. 2). Rockville, Md. : National Institute of Mental Health, 1982. Zillmann, D. Transfer of excitation in emotional behavior. In J. T. Cacioppo and R. E. Petty (Eds.), Socialpsychophysiology. New York: Guilford Press, in press. Zillmann, D., and Bryant, J. Viewer's moral sanction of retribution in the appreciation of dramatic presentations. Journal of Experimental Social Psychology, 1975, 11, 572-582. Zillmann, D., Bryant, J., and Sapolsky, B. S. The enjoyment of watching sport contests. In J. H. Goldstein (Ed.), Sports, games and play: Social and psychological viewpoints. Hillsdale, N. J.: Erlbaum, 1979. Zillmann, D., and Cantor, J. R. Affective responses to the emotions of a protagonist. Journal of Experimental Social Psychology, 1977, 13, 155-165. Zillmann, D., Hay, Τ. Α., and Bryant, J. The effect of suspense and its resolution on the appreciation of dramatic presentations. Journal of Research in Personality, 1975, 9, 307-323. Zillmann, D., Hezel, R. T., and Medoff, N. J. The effect of affective states on selective exposure on televised entertainment fare. Journal of Applied Social Psychology, 1980, 10, 323-339. Zillmann, D., Mody, B., and Cantor, J. R. Empathie perception of emotional displays in films as a function of hedonic and excitatory state prior to exposure. Journal of Research in Personality, 1974, 8, 335-349. Zillmann, D., Williams, B. R., Bryant, J., Boynton, K. R., and Wolf, M. A. Acquisition of information from educational television programs as a function of differently paced humorous inserts. Journal of Educational Psychology, 1980, 72, 170-180.

143

Ellen Wartella, Linda S. Hunter

Children and the formats of television advertising

Preliminary remark For the past decade, American research on children and television advertising has focused primarily on children's abilities to make sense of advertising messages. Public policy hearings on the effects of television commercials directed to children, held by the U.S. Federal Trade Commission in 1971 and 1979, and inquiries into children's television practices by the U.S. Federal Communications Commission between 1971 and 1974, prompted an early interest in children's attention to and comprehension of advertising messages. Opponents of advertising directed to young children claimed that such advertising was "deceptive" or "unfair", in that the target audience had neither the experience nor the cognitive abilities necessary to evaluate such messages (see Rossiter, 1980). Research on children and advertising in the 1970's thus tended to address two primary questions: (1) At what age and how do children discriminate television commercials from television programs? and (2) At what age do children understand the selling intent of commercial messages? In addition to research attention to the public policy issues regarding the vulnerability of the child audiences for television advertising, other American studies have examined children's comprehension of television advertising messages. Also, the persuasiveness of advertising in changing children's attitudes towards products and in effecting their request for advertised products has been studied, but less frequently than the nature of how children make sense of advertising claims. Thus, two major issues have been addressed in the research on television advertising directed to children: (1) Can children understand the information in commercial messages? and (2) Are television commercials persuasive to children and how do children understand this persuasive aspect of commercials? In this paper, we will review the status of the current research on how children make sense of television advertising with an aim toward elaborating the nature of the informative and persuasive aspects of television commercials directed to children. 144

Secondly, it should be noted that much of the research on children's processing of advertising messages has not paid much attention to commercials as a unique type of discourse with particular types of formats and styles of presenting information. Other than defining an advertisement as a television message which tries to sell a product, most of the American research on children and television advertising ignores how these commercials are constructed. In the second part of this paper we will attempt to offer a scheme for examining the discourse structure of television advertisements directed at children which might identify their informative and persuasive aspects. In addition, we will apply our analytic scheme to a content analysis of a sample of American children's television commercials.

Children's processing of television advertising Attention

to

commercials

How much attention do children devote to television advertising? Adler et al. (1977) report that after age 2, most American children spend about 3-4 hours per day watching television which yields an estimate of about 3 hours of exposure to television advertisements per week. However, observational evidence of the amount of attention given to advertisements by children varies across studies, an important variant being whether the study was conducted within homes, in school viewing groups or in special laboratories. Bechtel, Achelpol, and Akers (1972) reported an in-home observational study which found that the 1- to 10-year-old children they observed watched the commercials only 40% of the time they were on the air while the 11- to 19-year-olds watched them 55% of the time they were on. In another in-home observation, Winick and Winick (1979) reported that commercials were regarded as relatively unimportant by the over 300 children they observed and that children as young as 2 regularly left the room when a commercial was on. Self-report studies of children's attention patterns while viewing at home (Atkin, 1975d) indicated that younger children reported heavier exposure to Saturday morning television, and thus perhaps to commercials, than did older children. Mothers reported that their preschool through third-grade children paid close attention over half the time (59% for mothers of preschool and kindergarten children and 54% for mothers of first- and third-grade children). In contrast, only 29% of the mothers of fourth- and fifth-grade children reported that their children paid close attention (Atkin, 1978). Ward, Levinson, and Wackman (1972) reported data from 65 mothers who observed their 5- to 12-year-old children watching television in the home. 145

Overall, these children showed slightly higher attention to commercials than found by Bechtel et al. (1972). Children 5 to 8 years old were at full attention 67% of commercial viewing time, and 9- to 12-year-olds averaged full attention 75% of the viewing time. Zuckerman, Ziegler, and Stevenson (1978) videotaped second-, third-, and fourth-grade children watching a program with commercials in a special laboratory viewing room that had toys and other distractors. They reported uniformly low attention to both the 15-minute program (either a cartoon or a program about children on a camping trip) and the eight breakfast commercials embedded in it. On the average, only 17% of the children attended to the commercials. Zuckerman et al. found attention to the commercials decreased with increasing age of the child as in the in-home study. In constrast to the above research, two in-school studies found high attention to commercials among school children. Wartella and Ettema (1974) and Atkin (1975a) report high attention to commercials embedded in programs. Moreover, both studies found higher attention among older children (8 to 10 years in A tkin 's study; second-graders in the Wartella and Ettema study) than younger preschool or kindergarten children. In general, these findings suggest that where observations or reports of children's viewing have been conducted in more naturalistic environments such as homes or special viewing rooms with distractors present, attention to commercials decreases among older children. Preschool and early-elementary school-age children show more attention to commercials than do older children. However, in-school studies have found that older children show slightly higher attention to commercials than do younger children. Demand characteristics of the in-school viewing setting may help account for this divergent outcome. Effects of formal features on children's attention to

commercials

What factors of the commercial production affect attention? Recently, several studies have examined the impact of production factors, such as the specific audio-visual non-content characteristics of commercials on attention. Wartella and Ettema (1974) reported that nursery school children showed greater differences in attention to high as compared to low perceptual complexity commercials than did kindergarten and second-grade children. Their data were derived from in-school observations of 120 upper-middle class children viewing commercials embedded in a half-hour comedy program. 146

Two other recent studies have examined the effects of audio-visual formal characteristics on children's attention. Although the specific stimuli used were segments from Sesame Street, the research of Welch and Watt (1982) has implications for television advertising aimed at young children. Welch and Watt investigated the manner in which specific formal attributes of television influence visual attention among young children. For a 30-minute sequence of Sesame Street, variations in static and dynamic complexity were determined at one-second intervals. Static complexity refers to the diversity of the visual field and its organization, whereas dynamic complexity refers to the activity or movements of objects across a screen. For children between the ages 4 to 6, dynamic complexity was found to have a strong positive impact on attention, whereas static complexity was found to have a negative impact; i.e., the greater and more unpredictable the pattern of objects on the screen, the lower the visual attention among these young children. Greer, Potts, Wright, and Huston (1982) investigated the effects of commercials that varied both in formal features (high vs. low perceptual salience) and placement in a television program (dispersed vs. clustered) on children's attention and social behavior. For preschool children, high salience commercials tended to maintain attention better than low salience commercials, particularly when the commercials were dispersed. Greer et al. also found a sex difference in attention. Males were more attentive than females and their attention was maintained across commercials, whereas females' attention tended to decrease after the first 5 or 10 seconds. Although the results of their study indicate that the dispersed commercial format was more effective in maintaining attention, it should be noted that behavorial effects were generally greater in the clustered format. These findings differ from other studies of the effects of clustering of commercials on attention by At kin (1975a) and Duffy and Rossiter (1975). Both studies were conducted in classroom situations, and both report that young children show greater attention to clustered commercial formats. At kin (1975a) found higher attention to clustered commercials in his study of preschoolers through fifth-graders; Duffy and Rossiter (1975) found firstgraders more attentive to clustered commercials, but fourth-graders were slightly more attentive to the dispersed commercials. Thus both the audio-visual formal features of commercials and the format for presenting commercials (clustering or dispersing them) would appear to affect children's attention. Younger, preschool and kindergarten children would appear to be most heavily influenced by these perceptual characteristics of commercials. 147

Effects of separators between program and

commercials

As noted earlier, substantial public policy debates occurred in the United States during the 1970s focusing on children's abilities to distinguish television programs from television advertisements. In 1974, the U.S. Federal Communications Commission (FCC) required all television licensees to make a clear separation between programs and commercials. Several studies have examined the effectiveness of various program/commercial separators for aiding young children's discrimination between the advertisement and the program. In an early study, Palmer and McDowell (1979) examined the success of the three U.S. networks' attempts to make a clear separation between programming and commercials as required by the FCC in 1974. Sixty kindergarten and first-grade children were assigned to one of four viewing groups: a control group which viewed a television program and commercials with no separators and three experimental groups each utilizing a particular network's program/commercial separator format. The videotape was stopped at predetermined points during the commercials and the program, and children were asked whether what they had just seen was part of the show or part of the commercial. Children in the control group were able to distinguish programs from commercials as well as children in any of the experimental groups, correctly identifying commercials in about two-thirds of their response opportunities. Stutts, Vance, and Hudleson (1981) suggest that both the type of program/commercial separator and the age of the child should affect the speed with which commercial information within a program is recognized. One hundred and eight 3-, 5-, and 7-year-old children viewed a Bugs Bunny cartoon containing a Nestle's Quik Rabbit commercial. The children were assigned to one of three viewing conditions: a half second of black between program and commercial, an audio separator with unrelated video or titles, or an audio-video separator with related titles. With the exception of the 7-year-olds, speed of recognition did not appear to be a function of separator type; however, for the 7-year-olds only, the audio-visual separator appeared to be most effective in increasing speed of recognition. There was no difference in the ability of the 5- and 7-year-olds to distinguish between program and commercial, although the 3-year-olds appeared to have considerable difficulty with the distinction. Butter, Popovich, Stackhouse, and Garner (1981) also attempted to assess preschoolers' discrimination of program and commercials and to evaluate the effectiveness of several separation techniques. Eighty preschool children were divided into two equal groups (4.13 and 5.15 years), and assigned to one of four treatment conditions. All children viewed a 30-second commercial 148

inserted within a segment of Captain Kangaroo, with either an audio-visual break, an audio break, a visual break, or a blank separator between program and commercial. The children were instructed to let the experimenter know if a commercial was expected, in addition to informing the experimenter "when something different from the Captain Kangaroo show comes on." Older children recognized more commercials than younger children, although the younger children performed better than chance. Also, for older children, audio separators appeared to be most effective in priming a child to the onset of a commercial, although only 30% of the older children responded during this period. Levin, Petros, and Petrella (1982) attempted to assess the effect of the visual and auditory attributes of a message on preschoolers' ability to correctly identify programs and commercials. Seventy-two preschoolers (3, 4, and 5 years of age) viewed three videotapes (audio-visual modality, audio modality, and visual modality), each containing a random order of 14 TV program segments, 7 children's commercial segments, and 7 adult commercial segments (for a total of 84 segments). Each segment was 10 seconds long and the children were asked to identify each segment as a "commercial" or a "program." Five-year-olds (79%) identified segments better than 3- (66%) or 4-year-olds (74%), although correct identifications were significantly above chance at each age level. The 3- and 5-year-olds showed no modality preference, while the 4-year-olds performed best on the audio-visual mode. Children's commercials were easier to identify than adult commercials for the 3- and 4-year-olds, and programs (62% correct) were more difficult to identify than commercials (77%), possibly due to the unusually short segments utilized. A slightly different technique for measuring children's abilities to distinguish the program from commercials were used by Gianinno and Zuckerman (1977). They found that about 50% of the 4-year-old children they interviewed could, in eight of ten paired comparisons, correctly pick out a picture of a television commercial character paired with a television program character. On the other hand, nearly all of the 7-year-old children they interviewed could recognize the commercial characters in all ten paired comparisons. When asked to choose the picture of a character who showed products on television, nearly all of both the 4- and 7-year-old subjects demonstrated at least 80% accuracy. As these studies suggest, younger children, particularly preschoolers and kindergartners have difficulty in distinguishing television programs from commercials compared to older children. However, even young children do demonstrate some ability to at least perceptually distinguish programs from 149

commercials; this ability for perceptual discrimination would appear to occur somewhere between 3 and 5 years of age. However, there is little evidence that the use of program/commercial separators aids these young children's discrimination of the commercials. Comprehension

of the purpose

of

advertising

Research on children's understanding of the purpose of commercials has relied on children's abilities to articulate the persuasive aspect of advertising. Results of the various survey studies seem to indicate that below age 6 the vast majority of children cannot articulate the selling purpose of advertising (Adler et al., 1977; Bever, Smith, Bengen, and Johnson, 1975; Donohue, Meyer, and Henke, 1978; Meyer, Donohue, and Henke, 1976; Roberts, 1979; Robertson and Rossiter, 1974; Sheikh, Prasad, and Rao, 1974; Ward, Wackman and Wartella, 1977). Between kindergarten and third grade, that is between the ages of about 5 and 9 years, the majority of children have usually been shown to be able to articulate the selling intent of advertising. Variations in the different studies in the percentage of children between kindergarten and third grade judged to understand the purpose of commercials appear to be the result of variations in the measurement context as described by Wackman, Wartella, and Ward (1979) and Wartella (1980). For instance, of the 108 children in the Stutts, Vance, and Hudleson (1981) study, 78% said that they knew what a commercial was, based on a yes/no answer. There were no significant differences between 3-, 5-, ad 7-year-olds in response to this question. When asked to describe what a commercial is and why commercials are shown on television, however, the 3-year-olds were generally unsuccessful, whereas 11% of the 5-year-olds and 64% of the 7-year-olds defined the selling intent of television commercials. A more difficult behavioral response measure yielded lower estimates of commercial intent and perceptual discrimination of programs from commercials. Donohue, Henke, and Donohue (1980) attempted to assess children's comprehension of television through nonverbal measures. Ninety-seven children (2- to 6-year-olds) viewed a cereal commercial and were then asked to point to one of two pictures in order to indicate what the character in the commercial wanted them to do. In a second phase of the study each child was given a board with 12 cut-out figures arranged on it, from which the child selected figures representing the members of his or her family (including him or herself). Then, the children viewed six program and commercial segments, three of which were designed to appeal to adults and three to children. At the conclusion of each segment, the experimenter asked the child the following questions: "Who in your family would like to watch that?" and "Who in 150

your family would not like to watch that?" All of the storyboard figures indicated by the child were noted by the experimenter. The results indicated that, although the grasp of the concepts improves with age, children as young as three demonstrated some understanding of selling intent and were able to identify programs intended for adults compared to children. As children develop beyond early elementary school, a more complete and fuller understanding of persuasive intent and commercial role certainly develops. Several researchers put the demarcation between rudimentary understanding and grasp of persuasive intent at age 8 or older (Atkin, 1979; Roberts, 1979; Robertson and Rossiter, 1974), although the work of Wackman et al. (1979) might put the dividing line at as young as kindergarten age. In defining comprehension of persuasive intent, a slightly different criterion than that used by Wackman et al. (1979) is offered by Roberts (1979), and perhaps this accounts for the difference in the chosen demarcation points. Roberts argues that just understanding that commercials want someone to buy or to try a product is not sufficient evidence that a child understands the purpose and persuasive aspect of advertising. Calling on general cognitive developmental concepts and research on the formation of children's role-taking skills, Roberts contends that not until age 9 or 10 can children take the advertiser's motivations into account when considering a commercial. In his view, children who lack role-taking skills are unable to recognize that because advertisers are trying to sell them products the presentation of the product information may be biased. Therefore, children below at least 8 years of age cannot be wary consumers of advertising messages. His argument is based on general cognitive development theory and research, but it has not been put to direct empirical test. The Donohue et al. (1980) study noted above is difficult to interpret as a test of 3- to 6-year-olds' role-taking skills although the authors wish to argue that it is. In summary, during the elementary school years children's understanding of advertising proceeds from showing evidence of perceptual discrimination ability and rudimentary conceptual distinctions to an increasingly better articulated grasp of the concept of advertising. Conceptual understanding is articulated verbally by some kindergartners and appears to be well articulated by nearly all children by the time they reach third grade. It is difficult to provide evidence of when children begin to take advertisers' motivations into account when assessing any particular advertisement's claims, although there is reason to argue, as Roberts (1979) does, that this does not occur until after more advanced role-taking skills are acquired.

151

Memory for advertising

information

Research on children's memory for advertising information has indicated that children's recognition and recall of advertising messages increase as a function of age. Major increases in memory seem to occur between kindergarten and third grade. Ward (1972) and Ward et al. (1977) studied 5- to 12-year-old children's recall of their "favorite" television commercial in an interview conducted outside a television viewing situation. The children's responses were content analyzed. While the kindergarten children tended to recall a single element of the commercial (e.g., there was a girl playing with a doll), the older children tended to recall more product and commercial plot line information. They related the information in a story-line sequence and generally gave a more unified multidimensional description of the commercial and product. According to Ward et al. (1977), the major shift in recall from memory for one dimension to multidimensional memory seemed to occur between kindergarten and third grade. Studies of children's memory for commercials they have just been shown tend to support these findings. For instance, Atkin (1975a, 1975b, 1975c, 1975d) reported that immediately after viewing a cereal commercial, 90% of the 8- to 10-year-olds interviewed could recall two or more details from the commercial while about two-thirds of the 4- to 7-yearolds could do so. Similarly, Wackman et al. (1979) report that when kindergarten and third-grade children were shown groups of television commercials for candy or toy products embedded in a half-hour cartoon show, kindergartners generally performed at levels slightly above chance on recognition measures of the product or commercial story-line elements. When information was presented» about a relatively simple game product in which the kindergartners appeared interested, their performance improved and they answered three-fourths of the recognition measures accurately. Nevertheless, kindergartners' recognition memory overall trailed third graders' recall and recognition memory; across studies, third graders remembered between twothirds and three-fourths of the product and commercial elements while kindergartners retained 40%. Despite generally low recognition and recall scores, young children may find certain kinds of information more salient than other kinds. Several studies suggest that children as young as six may be more likely to remember premium information than brand name {Rubin, 1972; Shimp, Dyer, and Divita, 1976). However, there is evidence that memory for commercial slogans is well established by the time children reach grade school {Atkin, 1975d; Burr and Burr, 1977; Hendon, McGann, and Hendon, 1978; Katz and Rose, 1969; Lyle and Hoffman, 1972). 152

In addition to changes in children's memory for commercials, understanding commercial claims also appears to increase as a function of age. For instance, several studies of grade-school children's comprehension of toy product disclaimers indicated that disclaimer information needed to be presented in simple terms if young children were to understand it (Atkin, 1975b; Barry, 1978; Liebert, Sprafkin, Liebert, and Rubinstein, 1977). In the Liebert et al. (1977) study fewer than half the 5-year-olds understood the wording "partial assembly required" whereas nearly all children could comprehend a disclaimer such as "you have to put it together." Two studies have examined children's understanding , of the concept "balanced breakfast" which is included in all American cereal commercials. Atkin and Gibson (1978) found low understanding of this concept among 4to 7-year-old children. Palmer and McDowell (1981) also examined young children's understanding of the concept of a "balanced breakfast." Sixtyfour kindergarten and first-grade children were assigned to one of four terminology conditions — balanced, big, good, and right kind — and were asked what they would have if they were having a breakfast (with one of the four terms inserted in the space). The children's answers were recorded by response prevalence across each of the four basic food groups. The results indicated that the responses were not very evenly distributed across the four groups. The term big elicited the highest percentage of responses across the four categories, and balanced the fewest, with right kind and good intermediate. In summary, the research on children's memory for commercial messages indicates that as children grow older and have more advanced cognitive skills, their comprehension of a variety of advertising and nutrition information appears to increase. However, it is difficult from this research to identify exactly what is most informative in commercials for young children. Part of the problem is that this memory research has looked at a variety of characteristics of commercials in a piecemeal fashion. No systematic examination of children's memory for a variety of product- and plot-related information in commercials has been made. It would appear, however, when reviewing the research that has been conducted that young children may be more likely to remember product-related information (such as premium or brand name) than plot or story-line information. However, we have no systematic analysis of the variety of commercials used in these studies - or how they varied in informativeness regarding the product and narrative or storyline. Our understanding of children's memory for commercial messages would be improved by more systematic examination of the effects of different advertising formats on the viewer's memory. 153

Persuasive effects of television advertising to children In this section we will consider the persuasive impact of television advertising on children's desire for advertised products and purchase requests. In general, research studies have been directed at examining the conditions which enhance or limit advertising's persuasiveness. For instance, a number of recent studies have investigated the effects of television advertising on children's food preferences. Stoneman and Brody (1981) attempted to determine how peer-modeled food preferences combine with television food commercials to influence the food preferences of schoolaged children. The subjects were 40 black and 40 white fourth-graders who were randomly assigned to one of four treatment conditions, plus forty children who served as peer models. All the experimental subjects viewed four commercials for salty snacks; however, one group viewed the commercials alone, one group viewed the commercials and observed a peer model who expressed a preference for a salty snack and one group viewed the commercials and observed a peer model who expressed a preference for a nonsalty snack. A fourth group served as a control. After viewing the commercials and observing the peers, the children had to express a preference for salty or non-salty snacks. Stoneman and Brody found that when the peer model reinforced the commercial message by expressing a preference for salty snacks, children were most likely to express a preference for salty snacks, However, children in the television ad/peer model dissimilar condition were less likely to prefer salty foods than did children in the television ad only condition. Children in the control condition preferred salty food less frequently than any other experimental group. Thus when peer preference reinforces the commercial messages, the effect of commercials' messages on food preferences is strongest. Another study by Galst (1980) founds adults' comments can be effective in influencing food preferences stimulated by advertisements. Galst investigated the role of observational learning from television commercials on children's actual behavioral selection of snacks. Sixty-five children (3.5- to 6.75-years of age) were exposed to 4.5 minutes of commercials within a cartoon program 'each day for four weeks. The commercials were either for food products with added sugar content or for food products with no added sugar content and pro-nutritional public service announcements. After commercials for sugared food products, one group of children heard highly negative evaluative comments from an adult co-observer, while a second group heard positive evaluative comments after viewing commercials for unsugared products and the public service announcements. A third and fourth group viewed one or the other type of commercial without adult comment and a fifth group served as 154

a control. The dependent measure was the weekly proportion of snacks chosen which contained added sugar. Although at no point did any group select unsugared products the majority of the time, the condition that was most effective in reducing the consumption of sugar snacks was the presentation of commercials for products without added sugar and the pronutritional public service announcements, accompanied by positive evaluative comments by an adult. Following completion of the study, the children were also administered a questionnaire in which they were asked to identify products as either being healthy or containing too much sugar. Children in the sugared products/negative comments condition received significantly higher scores than children in the other conditions, indicating retention of nutritional information from the post-viewing comments although they had no effect on behavioral selections of snack foods. Further research on pro-nutritional public service announcements was conducted by Cantor (1981). He attempted to assess the relative persuasiveness of a humorous vs. nonhumorous pro-nutritional public service announcement (PSA). Three- to nine-year-old children viewed a humorous or nonhumorous PSA advocating the eating of oranges and the avoidance of sweets, followed by a commercial for either Hostess Cupcakes or a toy. The commercials were inserted about halfway within a 24-minute segment of children's programming. The children's choices of dessert (a fruit vs. a sweet dessert) were then recorded for the week following exposure. A significant behavior change was found only for those children who saw a nonhumorous PSA followed by a toy commercial. Cantor suggests that the humorous PSA may not have been seen as a credible source of information or that the Hostess Cupcake commercial may have interferred with storage of the PSA's specific content. Gorn and Goldberg (1982) note that PSAs are considered to be lacking in emotional appeal for young children, and must be aired frequently if they are to be effective. Scammon and Christopher (1981), however, conclude that length and frequency of PSAs does not appear to be as critical as quality in affecting food choices and nutritional knowledge and that, although exposure to pro-nutritional ads does not always lead to increased consumption of healthy foods, it does appear to dampen increased consumption of sugared foods. Unfortunately, each of these studies examining the effects of food commercials on product choice provide little information about the commercials employed so that we might assess what is persuasive about the commercial or the "quality" of the commercial that Scammon and Christopher refer to. One study which does attempt to address a potentially "persuasive" element in commercials is that of Desmond and Loughlin (1981). They note that 155

television commercials directed to young children frequently include peer interaction among the children appearing in the commercial; and they investigated the effect of such commercials containing social interaction on the affective responses of young children. Two 30-second commercials were professionally produced, both of which featured a wooden toy puppet not currently available. One commercial depicted four children playing with the puppet, whereas the other depicted a single child. The voice-over for both commercials was exactly the same. The sample included 48 first graders and 51 third graders; one-half of each age and sex group viewed each commercial. For both commercials, younger children exhibited more desire for the product than did older children, although all children liked the social interaction commercial significantly more than the noninteraction commercial. No significant relationship was found between desire for the product and commercial type; however, more children who viewed the social interaction commercial indicated that they would give the toy to their best friend than those who viewed the noninteraction commercial. Ruble, Balaban, and Cooper (1981) suggested that only children "high" in gender constancy would be differentially affected by sex-role information in a toy commercial, whereas children "low" in gender constancy would not be affected. Fifty boys and fifty girls (44 to 77 months) viewed a commercial for a neutral toy, the Fisher-Price Movie Viewer. The commercials depicted either two boys or two girls playing with the toy. Forty children viewed the same-sex models, 40 viewed opposite-sex models, and 20 served as a control group (no commercial). Following presentation of a short cartoon and the commercial, each child was free to play with a variety of toys, including the Fisher-Price Movie Viewer. Only high-gender-stage children spent significantly less time playing with the viewer when they saw opposite-sex models in the commercial, compared to high-gender-stage children who saw the same-sex models or no commercial. They were also more likely to state that the toy was more appropriate for an opposite-sex sibling. Overall, boys spent more time with the viewer than did girls, however the pattern of differences was very similar for both boys and girls. What these two studies suggest is that affective content of television advertisements such as the presence or absence of children and social interaction on the screen does affect children's evaluation and desire for the products advertised. Similarly, affective responses to commercials can override children's cognitive defenses to advertising. For instance, Ross et al. (1981) recently examined how celebrity endorsement in toy commercials influences children's perceptual and cognitive responses to the advertisements in two separate experiments with 8- to 14-year-old boys. Celebrity 156

endorsement appeared to affect the boys' product choice and the effect was at least as strong for older boys as younger boys. Most importantly, the endorsement techniques influenced preference independently of cognitive response to the commercials. Indeed even for the older boys, understanding that an endorsement was used, that the events on the endorsed product commercials were staged, and being less deceived about the physical characteristics of the "endorsed" racers did not lead to less preference for the "endorsed" model racer over other brands. That is, the older boys wanted the endorsed brand in spite of their presumed cognitive defenses and understanding of the purpose of the advertisements. Another study suggests that children's motivational state when watching commercials affects their desire for the advertised products. Galst and White (1976) found that the harder a preschool child worked to maintain TV commercials on a TV monitor, as compared to the program narrative, in an experimental situation and the more commercial television the child was exposed to at home, the greater the number of purchase requests directed to mothers in a supermarket observation situation after viewing. This study suggests that differential motivation for attending to television advertisements and for processing advertising information, seems to create individual differences in the conditions under which children are influenced by the advertisements. These studies indicate that some advertisements are more likely to produce an affective response in children than others. For instance, the presence of celebrity endorsers in toy commercials is appealing to grade school children. On the other hand, current American public service announcements for nutritional foods appear to lack affective appeal. What is needed to further identify what is persuasive in advertising directed to children is systematic analysis of what commercial elements elicit an affective response in children. A first step in this direction would be to conduct an analysis of the structure of commercials or how information is presented in television commercials directed to children. In the next section we offer such an analysis of the formats of American children's television advertising focusing primarily on the informative and persuasive aspects of these formats.

Discourse structure in children's television advertisements A variety of content-analytic studies have examined the content of television advertisements directed to children (Barcus, 1976, 1980; At kin and Heald, 1977). These studies typically examine the commercials for the presence or 157

absence of such content characteristics as the mention of brand name, the types of products advertised, the presence or absence of premium information and presence or absence of a variety of qualifiers and disclaimers. However, none of these content analyses offer a structural description of the formats of advertisements directed to children. Such a structural description of advertisements could help to differentiate the informative and persuasive aspects of advertisements. The literature on discourse structure would appear to offer such a scheme for structural analysis of advertising formats. Brewer (1980) has recently proposed that written texts can be classified on the basis of their discourse structure and discourse force. For example, an expository text is defined as discourse that attempts to represent, in linguistic form, such underlying logical processes as induction, classification, and comparison in order to explain or expound upon a particular topic or phenomenon. A narrative, on the other hand, is defined as discourse which embodies a sequence of temporal events, in which the events are causally or thematically related. Narrative discourse can be further subdivided into narratives that simply describe an event sequence, and stories. Stories are narratives that relate a causal sequence of events relevant to a protagonist pursuing a goal or resolving some problem (.Black and Bower, 1980). The discourse force of various structures of discourse can be described as either entertaining, informative or persuasive. According to Brewer certain structural patterns in discourse can manipulate the force of the discourse. In the present analysis of television commercials, this suggests that one should first examine the types of discourse structures present in television commercials directed to children in order to determine the potential discourse force, particularly the persuasive force of advertisements. In order to determine the various discourse structures present in television commercials directed to children, 58 American advertisements were categorized according to Brewer's (1980) analysis of discourse types. The advertisements were videotaped from the three major television networks between 8 a.m. and 12 p.m. on a Saturday morning in December, 1979. Since the sample was collected just prior to Christmas, advertisements for toys are overrepresented; at other times of the year, advertisements for food products predominate in Saturday morning children's TV hours. The products advertised fell into seven categories: fast food (4), cereal (7), candies (2), games (5), toys (32), public service announcements (4), and other (5). All four public service announcements addressed the issue of "the smart way to watch television"; products classified as other included advertisements for toothpaste (2), the Boy Scouts, Sherwin Williams paint, and pickles.

158

The audio portion of the commercials was categorized according to a 2 χ 2 classification scheme: the presence or absence of narrative structure and the presence or absence of expository structure according to Brewer's (1980) definitions of these discourse types. Of the 58 commercials in the present sample, the number which fell into each of the six possible categories were as follows: narrative-story/expository (7), narrative/expository (24), narrativestory/nonexpository (3), narrative/nonexpository (2), nonnarrative/expository (22), and nonnarrative/nonexpository (0). Three examples of the narrative, expository and story discourse structures examined in this analysis are presented below; the audio tract of each type of commercial format is described. Narrative Structure (Play Dough commercial): "The real Play Dough Fuzzy-Pupper Pet Shop! Little Spotty needs a clip Off it comes, snippety-snip Poodles hair was such a fright A permanent made it look just right Kitty's such a pretty girl And monkey's hair grows and curls The fun just seems to grow and grow You can make it with Play Dough!" Expository Structure (Wheelies commercials): " Kids are wheeling! Wheelies' Dragstrip Showdown, with loops you shape into different exciting layouts. Wheelies! Through the loops. Wheelies! Speedburner racers that wheelie all the way. Wheelies! Wheelies! Wheelies are a wheeling all the way! Wheelies' Dragstrip Showdown comes complete as shown, assembly required. By Mieko." Story Structure (MacDonald's foods): "It was Christmas Eve in MacDonaldland. Everyone was asleep except Ronald, who had other plans. I'm giving MacDonald gift certificates to all my friends, so Hamburgler can get a hamburger, Goodness can get a shake. So up to the chimney Ronald did reach. You can get 'em at MacDonald's, 50c each. 159

When suddenly, 'Merry Christmas!' MacDonald's gift certificates for me? Oh, gee. And a Merry Christmas to all of you from all of us at MacDonald's. Merry Christmas!" It is apparent that it is possible to differentiate television commercials on the basis of the type of discourse structure in which the verbal content of the advertisement is embedded. A further question is what types of information appears in these discourse types. First, mention of product name appears to remain relatively constant across discourse structures. However, information about product attributes does appear to vary with the type of discourse. In the present sample, stories and narratives contain little product information, with the exception of product or brand name. For example, the commercials for MacDonald's relate the various adventures of Ronald MacDonald, but make no explicit claims about the virtues of their fast food products. The content of expository commercials, on the other hand, focuses primarily on product information: description of the product, explanation of how the product works and/or what can be done with it, and disclaimers. Narrative/expository commercials contain both product and productirrelevant information, although the manner in which this information is presented varies considerably across commercials. Bloome and Ripich (1979) present an analysis of television commercials in which some message units were found to relate only to the plot (or social situation), others relate to both the plot and the product, and still others to the product only. Bloome and Ripich suggest that "it is almost as if the commercial was easing its way over from the plot (or social context) only tied message... to messages that are tied to the product only" (p. 224). Although a number of the narrative/expository commercials in the present sample fit such a description, others interweave expository and narrative content throughout the commercial message. One can also speculate about the relative informativeness and persuasiveness of the discourse force of these structures of television commercials. For instance, Brewer (1980) suggests that discourse with a story structure produces an affective response. If this indeed is the case, then commercials with a story structure would be expected to have more persuasive appeal than commercials with an expository structure, independent of the amount of product information present in the commercial since, as was reviewed earlier, most of the research suggests that commercials with an affective dimension are higher in persuasiveness than commercials without such a dimension. 160

Furthermore, although our analysis indicates that commercials with an expository structure provide more product-relevant information than commercials with a narrative structure, it is generally assumed that expository discourse is more difficult for children to process than is narrative discourse. However, Spiro and Taylor (1980) point out that there is little available data from research on children's processing of written text to support the latter belief. Also, the story grammar literature indicates that there are age-related differences in children's abilities to process these different discourse structures. In short, there is reason to believe that the different discourse structures present in advertising directed to children vary both in how much information about the product they present the child, that is, in informativeness, as well as in their affective appeal, or their persuasiveness. In order to better relate discourse structures of commercials to the informative and persuasive force of these commercials, we would need to empirically examine the effects of various commercial formats or structures on children's memory for the product and plot-related information in the commercials and the children's desire for the product advertised. We hope to do this in future research. We believe that such attention to the structural aspects of commercials would better help us understand both how children make sense of advertising's messages and the ways in which advertisements influence children's desire for products. Certainly, as the earlier review of the literature on children's processing of television advertisements indicates, the relative inattention to the unique aspects of the structure or formats of television advertising in past research makes it difficult to precisely assess what is informative or persuasive in children's television advertising. Future research which relates the structure of commercials to viewers' processing outcomes will redress the current imbalance in the literature.

References Adler, R., Friedlander, Β., Lesser, G., Meringoff, L., Robertson, T. S, Rossiter, J. R., and Ward, S. Research on the effects of television advertising on children. Washington, D. C.: United States Government Printing Office, 1977. Atkin, C. K. Effects of television advertising on children. First year experimental evidence. Technical report, Michigan State Univ., 1975. (a) Atkin, C. K. Effects of television advertising on children. Second year experimental evidence. Technical report, Michigan State Univ., 1975. (b)

161

Atkin, C. Κ. Effects of television advertising on children. Survey of children's and mothers' responses to television commercials. Technical report, Michigan State Univ., 1975. (c) Atkin, C. K. Effects of television advertising on children. Survey of preadolescents' responses to television commercials. Technical report, Michigan State Univ., 1975. (d) Atkin, C. K. Observation of parent-child interaction in supermarket decision making. Journal of Marketing, 1978, 42, 41-45. Atkin, C. K. Testimony before the Federal Trade Commission's rulemaking on children and TV advertising. San Francisco, January 1979. Atkin, C. K., and Gibson, W. Children's nutrition learning from television advertising. Unpublished manuscript, Michigan State Univ., 1978. Atkin, C. and Heald, G. The content of children's toy and food commercials. Journal of Communciation, 1977,27, 107-114. Barcus, F. E. Pre-Christmas advertising to children: A comparison of the advertising content of children's advertising programs broadcast in April and November of 1975. Newtonville, Mass.: Action for Children's Television, 1976. Barcus, F. E. The nature of advertising to children. In E. L. Palmer and A. Dorr (Eds.), Children and the faces of television. New York: Academic Press, 1980. Barry, T. The effect of a modified disclaimer on inner city vs. suburban children. Paper presented at the meeting of the American Marketing Association, Boston, 1978. Bechtel, R. B., Achelpol, C., and Akers, R. Correlates between observed behavior and questionnaire responses on television viewing. In E. A. Rubinstein, G. A. Comstock, and J. P. Murray (Eds.), Television and social behavior (Vol. 4). Television in day-to-day life: Patterns of use. Washington, D. C.: United States Government Printing Office, 1972. Bever, T., Smith, M., Bengen, B., and Johnson, T. Young viewer's troubling response to TV ads. Harvard Business Review, 1975, 53 (6), 109-120. Black, J. B., and Bower, G. H. Story understanding as problem-solving. Poetics, 1980, 8, 223-250. Bloome, D., and Ripich, D. Language in children's television commercials: A sociolinguistic perspective. Theory into Practice, 1979, 18, 220-225. Brewer, W. F. Literary theory, rhetoric, and stylistics: Implications for psychology. In R. J. Spiro, Β. C. Bruce, and W. F. Brewer (Eds.), Theoretical issues in reading comprehension. Hillsdale, N. J.: Erlbaum, 1980. Burr, P., and Burr, R. M. Product recognition and premium appeal. Journal of Communciation, 1977, 27(1), 115-117. Butter, E. J., Popovich, P. M., Stackhouse, R. H., and Garner, R. K. Discrimiation of television programs and commercials by preschool children. Journal of Advertising Research, 1981,21, 53-56. Cantor, J. Modifying children's eating habits through television ads: Effects of humorous appeals in a field setting. Journal of Broadcasting, 1981, 25, 37-47. Desmond, M., and Loughlin, R. J. Social interaction in advertising directed to children. Journal of Broadcasting, 1981, 25, 303-307. 162

Donohue, T. R., Henke, L. L., and Donohue, W. A . Do kids know what T V commercials intend? Journal of Advertising Research, 1980, 20, 51-57. Donohue, T. R., Meyer, T. P., and Henke, L. L. Black and white children's perception of the intent and values in specific adult and child oriented television commercials. Unpublished manuscript, Univ. of Hartford, 1978. Duffy, J., and Rossiter, J. R. The Hartford experiment: Children's reactions to TV commercials in blocks at the beginning and end of the program. Paper presented to the Conference on Culture and Communications, Temple Univ., Philadelphia, March, 1975. Galst, J. P. Television food commercials and pro-nutritional public service announcements as determinants for young children's snack choices. Child Development, 1980, 51, 935-938. Galst, J. P., and White, M. A . The unhealthy persuader: The reinforcing value of television and children's purchase-influencing attempts at the supermarket. Child Development, 1976, 17, 1089-1096. Giannino, L. J., and Zuckerman, P. A . Measuring children's responses to television advertising. Paper presented at the meeting of the American Psychological Association, San Francisco, 1977. Gorn, G. J., and Goldberg, M. E. Behavioral evidence of the effects of televised food messages on children. Journal of Consumer Research, 1982, 9, 200-205. Greer, D., Potts, R., Wright, J. C., and Huston, A . C. The effect of television commercial form and commercial placement on children's social behavior and attention. Child Development, 1982, 53, 611-619. Hendon, D., McGann, Α., and Hendon, B. Children's age, intelligence and sex variables mediating reactions to T V commercials: Repetition and content complexity implications for advertisers. Journal of Advertising, 1978, 7 (3), 4-12. Katz, M., and Rose, J. Is your slogan identifiable? Journal of Advertising Research, 1969, 9 (1), 21-25. Levin, S. R., Petros, T. V., and Petrella, F. W. Preschoolers' awareness of television advertising. Child Development, 1982, 53, 933-937. Liebert, D., Sprafkin, J., Liebert, R., and Rubinstein, E. Effects of television commercial disclaimers on the product expectations of children. Journal of Communication, 1977, 27(1), 118-124. Lyle, J., and Hoffman, H. Explorations in patterns of television viewing by preschool age children. In E. A . Rubinstein, G. A . Comstock, and J. P. Murray (Eds.), Television and social behavior (Vol. 4). Television in day-to-day life: Patterns of use. Washington, D. C.: United States Government Printing Office, 1972. Meyer, T., Donohue, T., and Henke, L. Black children's perceptions of TV advertising: A cognitive developmental study. Paper presented at the meeting of the Speech Communication Association, San Francisco, December, 1976. Palmer, E. L., and McDowell, C. N. The program commercial separators in children's television programming. Journal of Communication, 1979, 29 (3), 197-201.

163

Palmer, E. L., and McDowell, C. Ν. Children's understanding of nutritional information presented in breakfast cereal commercials. Journal of Broadcasting, 1981, 25, 295-301. Roberts, D. F. Testimony before the Federal Trade Commission's rulemaking on children and TV advertising. San Francisco, January, 1979. Robertson, T. S., and Rossiter, J. R. Children and commercial persuasion: An attribution theory analysis. Journal of Consumer Research, 1974, 1, 13-20. Ross, R. The effects of celebrity endorsement in TV ads: Children persuaded despite cognitive defenses and general skepticism toward advertising. Report to the Federal Trade Commission, University of Kansas, Dept. of Human Development, 1981. Rossiter, J. R. Children and television advertising: Policy issues, perspectives, and the status of research. In E. L. Palmer, and A. Dorr (Eds.), Children and the faces of television. New York: Academic Press, 1980. Rubin, R. An exploratory investigation of children's responses to commercial content of television advertising in relation to their stages of cognitive development. Unpublished doctoral dissertation, Univ. of Massachusetts, 1972. Ruble, D. N., Balaban, T., and Cooper, J. Gender constancy and the effects of sextyped televised toy commercials. Child Development, 1981, 52, 667-673. Scammon, D. L., and Christopher, C. L. Nutrition education with children: A review. Journal of Advertising, 1981, 10, 26-48. Sheikh, Α. Α., Prasad, V. K., and Rao, T. R. Children's TV commercials: A review of research. Journal of Communication, 1974, 24 (4), 126-136. Shimp, T., Dyer, R., and Divita, S. Advertising of children's premiums on television: An experimental evaluation of the FTC's proposed guide. Unpublished manuscript, George Washington Univ., 1976. Spiro, R. J., and Taylor, Β. M. On investigating children's transition from narrative to expository discourse: The multidimensional nature of psychological text classification. (Tech. Report No. 195). Urbana, 111.: Center for the Study of Reading, University of Illinois at Urbana-Champaign, 1980. Stoneman, Z., and Brody, G. H. Peers as mediators of television food advertisements aimed at children. Developmental Psychology, 1981, 17, 853-858. Stutts, Μ. Α., Vance, D., and Hudleson, S. P. Program-commercial separators in children's television: Do they help a child tell the difference between "Bugs Bunny" and the "Quik Rabbit?" Journal of Advertising, 1981, 10, 16-25. Wackman, D., Wartella, E., and Ward, S. Children's information processing of television advertising. Unpublished manuscript, Univ. of Minnesota, 1979. Ward, S. Children's reactions to commercials. Journal of Advertising Research, 1972, 12 (2), 37-45. Ward, S., Levinson, D., and Wackman, D. Children's attention to television advertising. In E. A. Rubinstein, G. A. Comstock, and J. P. Murray (Eds.), Television and social behavior (Vol. 4). Television in day-to-day life: Patterns of use. Washington, D. C.: United States Government Printing Office, 1972. Ward, S., and Wackman, D. Children's information processing of television advertising. In P. Clarke (Ed.), New models for mass communication research. Beverly Hills, Cal.: Sage, 1973. 164

Ward, S., Wackman, D., and Wartella, E. How children team to buy. Beverly Hills, Cal.: Sage, 1977. Wartella, E. Individual differences in children's responses to television advertising. In E. L. Palmer and A. Dorr (Eds.), Children and the faces of television. New York: Academic Press, 1980. Wartella, E., and Ettema, J. A cognitive developmental study of children's attention to television commercials. Communication Research, 1974, 1, 69-88. Welch, A. J., and Watt, J. H. Visual complexity and young children's learning from television. Human Communication Research, 1982, 8, 133-145. Winick, M. P., and Winick, C. The television experience: What children see. Beverly Hills, Cal.: Sage, 1979. Zuckerman, P., Ziegler, M. E., and Stevenson, H. W. Children's viewing of television and recognition memory of commercials. Child Development, 1978, 49, 96-104.

165

Ingegerd Rydin

How children understand television and learn from it: A Swedish perspective The significance of comprehension Television's potential for imparting information has been a perennial topic of debate among producers and researchers. Some stress the negative effects of the medium: television makes people passive, inhibits their curiosity and dampens their thirst for knowledge, or else it gives rise to aggressive and violent behaviour. Finally, television, it is asserted, offers a distorted, unrealistic conception of reality. Others say less about negative effects, but they have most faith in the medium's ability to convey vicarious experiences. A seasoned Swedish producer commented: "Personally, I've come to the conclusion that TV may not be the best medium for information, but it is great when it comes to arousing feelings and interest... What really comes across on the TV screen is a producer's interest and involvement in the programme." While this statement is optimistic in a sense, it expresses doubt about the medium as a conveyor of knowledge. Mass communication research offers another view of the medium. In many countries, Sweden among them, a good share of programming — both factual and fictional — is addressed specifically to children. Many years' research on children's programmes in Sweden tells us that children can indeed learn quite a bit from television — from individual programmes, but especially from extended series of programmes. The repetition of essential information and messages in segment after segment can produce lasting learning effects. Programmes can also activate their viewers, that is, children can be inspired to do things by what they see on television. A Swedish series a few years ago taught children the sign-language used by the deaf. Children using signlanguage was a common sight throughout the country while the series was on the air. It is important, however, that children watch television in the company of an adult. Adults can help them apply what they have learned from television in a constructive way (cf. Singer and Singer, 1976). Many researchers claim, however, that television teaches rather little, mainly because the medium itself is passive, i.e. it does not involve mutual 166

communication. Harvey Lesser (1977), for example, explains Sesame Street's success by the fact that the programme succeeded in creating a conducive atmosphere in the home, an atmosphere that stimulates further activity and knowledge-seeking on the part of its audience. Lesser maintains that programmes must actively create rapport with the viewers. As an example he mentions a programme in which a young woman develops a "friendship" with her audience and speaks directly to the children, trying to relate to their experiences. A similar technique is used in the popular American series Mr Rogers' Neighborhood, in which Mr. Rogers speaks "directly" to his viewers and furthermore tries to establish a sense of mutuality. This technique is open to question. Such efforts may be perceived as artificial, inasmuch as there is no real communication between the viewers and the programme host, while on the other hand very young children may actually believe they can be seen and heard. Although the medium itself may be passive, we should not underestimate children's ability to learn using "their minds alone". A small child may be active without direct physical manipulation, just as he or she may be mentally passive while sitting and tinkering with various objects. That is to say, children can absorb a lot simply by looking and listening, provided they are interested. As a means of establishing a dialogue between the viewer and the screen, Lesser (1977) proposes various feedback systems whereby children might experience response to their communications on their TV screens. Computer technology, plugged into, say, a cable system makes it possible to communicate directly with the audience, but this will not be everyday reality for some time to come. Nor is such a development necessarily desirable. In its extension, such a system could amount to a form of televised programmed instruction that will demand advanced technology that few households can afford. Instead, resources should be devoted to making programmes that are "conducive to learning" among children. Parents or other adults should also be encouraged to keep children company in front of the set. Programmes can also be made that get through to viewers more effectively. In the first place, one may wonder whether the producers have done everything in their power to reach their viewers. According to an English study (O'Hare, 1980), psychologists and educationalists find few children's programmes to be really well adapted to their target audiences. They are often too sophisticated for the age groups they are intended for. A greater effort must be made to relate the form and content of programmes to the viewers' experience and maturity. The study also asserts that producers of educational programming are not sufficiently aware of their viewers' cognitive abilities. 167

According to Gwen Dunn, a former producer, British programme-makers tend to overestimate children's knowledge, experience and vocabulary. On the basis of a number of studies Dunn concludes that English preschoolers' TVviewing is largely a question of habit, and that their understanding of what they see is poor {Dunn, 1977). A Finnish study (.Liikanen, 1974) concludes that information programmes intended for preschool audiences satisfy less than half the criteria that should be applied to children's programmes, e.g. careful use of language, agreement between sound and picture, not allowing extraneous information to interfere with the essential message. Thus, failure to reach one's audience is partly due to the producers' not knowing enough about their audiences. But it also has to do with the aims and motives the programmes are to serve. In some cases the medium is used in ways that clearly militate against comprehension and learning. In commercial television systems the prime aim is to keep viewers watching one's own channel, which is best achieved by making entertaining programmes with striking sound and visual effects and "action". That the viewers may not comprehend the programmes is a secondary concern. In Scandinavia we still have systems that allow room for programmes that address themselves to children on the basis of their capacity and interests. Neither need producers compete with other channels for juvenile audiences. Norway and Denmark .each have only one channel; Sweden has two, but children's programmes are "protected" from competition on the opposite channel. Sesame Street, and various adaptations of that series, apply commercial production techniques to get pedagogic content across to young viewers. Unfortunately, a number of studies indicate that children nevertheless understand surprisingly, little, despite a high level of attention and interest. It seems that the tempo, variety and comic features sometimes gain the upper hand (cf. Pearl, Bouthilet, and Lazar, 1982). Interest and comprehension are obviously related. Attention and interest are necessary prerequisites to understanding, but recent findings (Anderson et al., 1981) suggest that children in fact not only appreciate engaging sound and visual effects, but also the very fact that they understand the content: when they notice they understand something, they become more interested. Thus, understanding can be a prerequisite to attention and interest. The optimal goal for producers is naturally to make children's programmes that are both attractive and comprehensible, but this is a difficult combination to achieve, especially comprehensibility. One must try to put oneself in the 168

child's situation and study the relationship between children's understanding of television and their overall level of maturity on the one hand and programme form and content on the other. Comprehension is intimately related to level of cognitive development, i.e. how children of different ages think, apprehend and process information. In the following we shall review some of the current research on this vital aspect.

How children understand television — what the research tells us Much of the research on comprehensibility has focussed on how children of different ages understand television programmes in a more general sense, that is, researchers have sought to trace developmental trends in understanding that can be related to and explained by what we know about cognitive development. There is a risk, however, that in searching for the general we may overlook, or underestimate, the importance of the specific. In looking for age trends, it is easy to forget that comprehension is highly dependent on the form and content of the stimuli, the programmes tested. Comparability between studies may be complicated by other factors, as well. In the first place, understanding is defined variously. Secondly, the definition used will determine one's choice of measures, which is of decisive importance. It makes a big difference whether understanding is measured by asking the children to reconstruct the story in their own words, or whether they are asked to respond to questions. Finally, comprehension can be measured with or without aids to recall, such as pictures from the programme, which the child is asked to place in the right order. Such methodological differences between studies make it difficult to make totally congruent generalisations about children's understanding of film and television. Preschool

children

Most research to date has indicated that preschool children, and particularly 4- and 5-year-olds, have difficulty understanding the programmes they watch on television. Noble (1975), who has summarized ti.e research prior to the 1970s, states that 4- and 5-year-olds tend to recall isolated episodes or scenes, but are unable to relate a story in its entirety. This has to do with the preoperational stage of the children's thinking, which is to say that they live totally in the present and have difficulty connecting the beginning, middle and end of a programme. Nummenmaa (1969) presented a well-organised version and a randomly assembled version of the same film content to children aged four to six. The 169

children seemed to have tröuble following the plot in either treatment. When asked to relate the story, some children started with the ending or the middle. Those who had seen the well-organised version did not benefit from its logical structure in trying to recount the story, nor did those who had seen the "random" version seem to notice that it lacked a logical structure. Not even programmes made especially for a given age group seem particularly comprehensible. Leifer et al. (1971) had children of different ages see a film version of a well-known folk-tale but noted widespread failure to comprehend among the 4-year-olds in the group. All research does not offer quite so dark a picture with respect to preschoolers' comprehension, however. In the following I shall try to point out various nuances in the findings to date, taking account — to the extent possible — of the form, structure and content of the stimulus as well as the measures applied. In a Swedish study (Rydin, 1976a), 5- and 7-year-olds were shown a factual programme that treated a biological process: how a seed sprouts and grows into a pine, which is ultimately made into a telephone pole. Understanding with respect to the whole of the process (as measured "with a picture test) was surprisingly good, even among the youngest children. They had difficulty, however, in recounting the factual information offered in the programme. Another Swedish study (Filipson and Schyller, 1969) confirms that 5-year-olds can follow and grasp a biological process (breathing) if it is presented well enough. Even 4-year-olds showed a good understanding of the programme. The children were asked to reconstruct the breathing process with the help of an illustration. The programme format was apparently an important factor. Another programme in the same series, describing the digestive process, was not at all as effective among 4- and 5-year-olds: hardly any of them was able to reconstruct the process. This was attributed to the manner in which the information was presented. The children had far greater difficulty recounting the factual information in the programme. A similar pattern is noted in an American study of an excerpt from the perennial CBS series Captain Kangaroo (Friedlander, Wetstone, and Scott, 1974). Here, children of about four years of age had great difficulty apprehending the factual information in the programme, although they were able to follow the story line quite well. Even if one succeeds in imparting a good grasp of the whole, this is still far from bringing about an understanding of the deeper meaning that often underlies both fairy tales and pedagogically inspired programmes. In one Swedish case producers tried to communicate an understanding of the 170

situation of immigrant children from Finland. They did so by describing the day-to-day lives of a Swedish and a Finnish girl. The underlying message they intended to get across was roughly: "We are quite alike, after all". The film did not succeed, even though the children were able to follow the story and relate various details (Preisler, 1975). But, in order to grasp the message, the viewers had to draw conclusions from the material presented, and this they were unable to do since they were still in the preoperational phase of development. Comprehension may also vary, depending on the medium by which content is presented. The sensations imparted via television are quite different from those one gets when someone reads a story. Rather little attention has been accorded the relationship between medium and understanding to date. Meringoff (1980), however, found that children's recounting of a story communicated to them by a picture-book and by an animated television programme, respectively, differed quite markedly. Children who had listened to the picture-book reading tended to be more bound to the text, while they also seemed more apt to interject their own spontaneous comments and draw their own conclusions than those who had seen the TV version. The latter group, on the other hand, were more bound to the picture and emphasized the visual events more. Typical of film and television compared with other media is that information is presented sequentially. Perhaps this factor makes it more difficult for younger children to perceive continuity in the stream of images, thereby impairing their ability to gain a grasp of the whole. Even the simplest of stories demands the mental association of a couple of more scenes in order to gain a sense of the whole. Brown, who has especially studied logical thinking in small children, presented a picture-story to preschoolers in two ways: (1) sequentially, one picture at a time, and (2) simultaneously, all the pictures at once. When the children were then asked to recount the story, the simultaneous treatment produced the better results {Brown, 1976a). Thus, it seems easier for small children to create a meaningful whole from a series of pictures when they have the whole story in front of them. Comic-books afford such an advantage. Another disadvantage associated with television is not inherent to the medium per se, but a function of how we use it. Many common techniques and conventions of televised narration are not suited to younger children's level of development. Producers have a penchant for fragmenting the messages in order to create "action" or to condense or "streamline" events and processes. Furthermore, discontinuities in time or space are frequent, although they may 171

obliterate the continuity of the story for the untrained viewer. The soundtrack seldom carries information that might help link the images together. On the other hand, moving pictures would seem to afford a definite advantage over, for example, still pictures in their ability to depict reality and show all the steps in a process or event. In Sweden and the rest of Scandinavia and Finland television programmes of still pictures are not uncommon. Therefore, in a Swedish study (Rydin, 1979), we tested the hypothesis that an animated film can better portray a biological process than a film of stills for audiences of 5- and 7-year-olds. It turned out, however, that only the 7-year-olds benefited from the "extra" information the animation contributed. Among the younger group retention/comprehension was consistently low for both versions, nor did animation help. It seems that even more reinforcement of the message, e.g. repetition and explanations in the verbal complement, is needed if 5-year-olds are to understand better. Perhaps television may have an unused potential for communicating with preschool audiences if only the medium were used in accordance with the children's cognitive abilities. Several well-known psychologists have stressed the importance of form, content and presentation for the effectiveness of messages. They contend that psychologists like Piaget, for example, have underestimated preschoolers' capacity for logical thinking because they have used too difficult or uninteresting stimuli (cf. Donaldson, 1978; Brown, 1976b; Mandler and Johnson, 1977). Donaldson, for instance, claims that children are not at all as egocentric in their thinking as is generally assumed on the basis of Piaget's theories and observations. In studies of young children's recounting of stories, Brown concludes that children seem to appreciate logically structured sequences and benefit by logic, which would indicate, he submits, that they at least are capable of semilogical thought. So far, we have focussed on preschoolers. In recent years, however, interest has increasingly focussed on school-aged children. Schoolchildren in Sweden and other countries are among the most avid consumers of TV fare, and they practically exclusively watch programmes intended for adult audiences. This gives us special reason to examine their comprehension and interpretation of television programmes, particularly entertainment such as detective/ adventure stories and "thrillers". School-aged

children

By "school-aged children" we generally refer to children aged about seven to twelve years. This is an interesting period from the perspective of developmental psychology, since children's thinking in this period is still very different from adult thought processes. Meanwhile, children of these ages 172

show some signs of mental maturity; their command of the language, for example, is nearly complete. What particularly distinguishes school-aged children is their ability to draw conclusions from the material presented to them. That is to say, their thinking is more flexible and not so closely bound up with the situation in which they find themselves at the moment. They freely relate their experiences to others. When recounting a story, they frequently introduce their own interpretations, which were not part of the original version. Making one's own interpretations and drawing conclusions are part of the mental processing by which we create structures and contexts in what we have perceived, which facilitates recall (cf. Paris and Mahoney, 1974; Paris and Lindauer, 1976; Paris and Upton, 1976). Despite the impression of maturity, studies have found that children's comprehension of television up to about eight years of age is poor. Collins et al. (1978) let children of different ages (8 to 12 years) see an adventure film of the type that is popular among school-aged children. The film was edited in three versions: (1) a simple version, containing the information essential to the story only; (2) a more complex story involving a subplot as well as the essential story; and (3) a randomly assembled series of scenes. Collins and his colleagues found that 8-year-olds, particularly the boys, understood very little of any of the versions and that comprehension differed little between the logically structured and the random versions. In other words, the 8-year-olds seemed unconscious of structure. As regards the recounting of episodes and extracting the essentials of the story, the logically structured versions produced better results, which suggests that context aided the retention of the central scenes. Furthermore, comprehension was somewhat better among younger girls who had seen the simple version, compared to the more complex version. On the whole, however, comprehension was equal for the complex and simple versions, which indicates that school-aged children were able to extract the essential information out of the material and were not distracted by the subplot in the complex version (Collins et al., 1978). It seems as though the period between eight and eleven is critical with respect to comprehension: both boys' and girls' cognitive abilities increase markedly between these two ages. Differences between 11- and 13-year-olds are not nearly as pronounced. These differences seem to apply generally, that is, in relation to different types of films. A Swedish study of an informative programme intended for schoolchildren found a similar pattern (Hedbom, 1970). The pattern probably reflects levels of cognitive development and particularly, as noted above, the ability to draw conclusions from the material presented, which appears to improve progressively between eight and eleven 173

years. The differences are clearly apparent if you stop a film at a critical point in the plot and ask the viewers to predict the subsequent development of the plot on the basis of what they have seen. Eight-year-olds' predictions tend to be stereotyped and simplistic, with little bearing on the foregoing material, while older children make use of what the material has told them (cf. Collins et al., 1978). Dramatized stories generally require viewers to interpret what they see and to draw their own conclusions inasmuch as the plot is carried forward in the dialogue and actions of the characters. The viewers must fabricate the context in which the dialogue and actions take place. In a Swedish study (von Feilitzen, Filipson, and Schyller, 1979), schoolchildren (7 to 12 years) watched an adventure film in three different treatments: one version in which the dialogue was dubbed in Swedish; one furnished with subtitles (the customary form of translation in Sweden); and one with the addition of a voice-over speaker text, which summarized the essential development of the plot. The hypothesis was that the dubbed version would result in better comprehension than the other two treatments. Surprisingly enough, the voice-over speaker text turned out to work best in all age groups. On closer examination, the speaker text was found to explain and "link together" the elements in the story so as to facilitate comprehension. The viewers did not have to make their own interpretations and could thus follow the plot more easily. Involvement and interest in the programme content also play a role in this regard. Children can make better interpretations or conclusions when the content engages their interest. Even 7- and 8-year-olds can make rather mature judgements when emotionally charged events are involved (Minkkinen, 1969; Kvart, 1978). Recalling facts and details One reason children may have difficulty following the story line or main theme in a programme may be that they lack strategies for extracting the essentials out of a body of information. Preschoolers may totally lack strategies for selecting out information and become preoccupied with whatever catches their attention. This could suggest that it would make no difference to young children whether or not a television programme has a structure — but, as we have seen, several of the findings reviewed above contradict such an assumption. A number of studies in the field of cognitive psychology have been done to determine how children of different ages select out information. A distinction is made between "task-relevant" and "irrelevant" information. 174

The first such studies were done in laboratory settings using information that was judged to be "neutral". The findings revealed rather consistent patterns, the main features of which were: Children of different ages have different information processing strategies. Preschoolers do not seem to select information in any systematic manner. They frequently become preoccupied with details and irrelevant information. From age six, they start building up a reserve of knowledge, storing both relevant and irrelevant information. At 12 to 13, however, children seem to grow more discriminating; they can avoid irrelevant and unrelated information and concentrate on essentials (Maccoby, 1969; Hägen, 1972). Subsequently, such research began to make use of meaningful material such as tales, films and television programmes. The findings remained by and large consistent with the above patterns {Hale, Miller, and Stevenson, 1968; Collins, 1970). But they are not totally unequivocal, which is probably due to the fact that the stimuli are no longer "neutral" and are of widely varying form and content. Hawkins (1973), for example, presented a "juvenile western" and an "adult western" film to children aged eight to fifteen. The films differed in their degree of difficulty. In the case of the juvenile film recall of essential, important scenes increased consistently with age, but recall of the "irrelevant" content also increased and did not subside at 12 to 13 years as had been anticipated. In the case of the adult film, the developmental "rule" that recall of irrelevant information declines at 12 to 13 years of age was confirmed. A conceivable explanation may be that the information load in the more complex film reached a ceiling, i.e. that the adult film contained altogether too much information for the children to process, so that they let less important details pass by in order to keep abreast of the plot. In the case of the juvenile film the older children could afford to store away all kinds of information, including details and irrelevancies. No sifting or selecting was necessary in order to maintain a grasp of the essential story. Thus, when we leave the laboratory and start broadening our perspective, we reach somewhat different conclusions concerning children's information processing. How one selects information, then, depends on the medium in question and the nature of the material, e.g. its difficulty, how strong the story line is, how prominent the irrelevant information and whether the content is factual or fiction. Unfortunately, we know all too little about children's information processing in relation to film and television. We do have some further findings, however, concerning children's apprehension of tales, factual and prose narratives. 175

In a Swedish study (Rydin, 1972), 5- and 7-year-olds listened to a factual story describing a biological process (a seed sprouting, growing into a pine tree, which is subsequently made into a telephone pole). Both age groups recalled the irrelevant information much better than the factual content. The irrelevant details were easier to remember, most probably because they were familiar to the children, and some of them were also amusing. The factual content, on the other hand, was entirely new to most of the children and thus unfamiliar. "Psychological distance" may also have been a factor. In an American study by Christie and Schumacher (1975) who explored children's consciousness of narrative structure using fictional material the findings were different. Very briefly, the story was about an outing made by a group of children in a rowing boat. The story contained no particularly dramatic elements, but there was a certain amount of irrelevant information. It turned out that all the age groups (5, 7 and 10 years) more readily recalled and could recount the relevant content, and that their narrations tended to stay on the essential theme. This suggests once again that even 5-year-olds have a certain awareness of narrative structure. The structure seems to serve as an aid to recall of the essential information. Subsequent studies, using other kinds of prose narratives, have confirmed these findings (Brown and Smiley, 1977). Finally, in the Soviet Union Korman (1971) has determined that Russian preschoolers seem to be conscious of what features are more or less essential to the development of the plot or story. They are not always aware, however, that their listeners need coherence in the recounting of the story in order to understand. The children's ability to tell the story well seems to depend on whether or not they found it interesting in the first place. It is conceivable that Russian preschoolers may be more adept story-tellers than their counterparts in other countries, since they receive considerable training in recounting stories they have listened to. Story-telling is considered of value in developing retention and a capacity for logical thinking (cf. Köhler, 1980). Thus, we find that research using realistic material indicates that even younger children have a sense of structure and that they make use of this awareness in recounting stories they have been told. The content of the stories also seems to play a role, however. It makes a difference whether a factual narrative relates unfamiliar content or familiar, "everyday" occurrences. Even preschoolers can keep track of the story and select out the information that is essential to the development of the plot when the content is simple and familiar to them. If, on the other hand, the story is complex and unfamiliar, they regress to a less systematic manner of taking in the information and are more readily distracted by irrelevancies. Distractions impair their ability to 176

select out the essential content, despite its being reinforced through repetition and presented in an accessible, logically structured story. Children are not the only ones who store away extraneous information. We have many examples of how adults, too, become preoccupied with details and irrelevancies in, for example, newscasts. Viewers and listeners recall the location and persons or things involved better than the causes and consequences of news events (Findahl and Höijer, 1976). They are sometimes even distracted by the tie or hair-style of the newscaster or interviewee. Nonetheless, adults are far better equipped than children to judge what is significant, and they use this ability whenever they feel pressured by the amount of information at hand and are sufficiently motivated to make an effort to understand. Do details

distract?

It seems reasonable to assume that irrelevant information might distract attention from the central theme or plot and thereby impair comprehension of it. Younger children are most probably more easily distracted than older ones. A couple of Swedish studies have found that irrelevant informtion can be distracting, but the effect is not particularly strong. Only dominant and emotionally charged irrelevancies have such an effect. The inclusion of background details to enrich a narrative does not appear to impair comprehension. Such details can make the story more attractive and interesting. Indeed, when irrelevant information is smoothly integrated into the essential content, comprehension tends to increase, but when such details stand out and are poorly integrated into the story, they tend to interfere. It also makes a difference whether the irrelevant material is introduced into the sound or the picture. The picture can contain more irrelevant information than the soundtrack without affecting comprehension. The information in the soundtrack must be more precise and kept relevant to the essential theme. The sound component guides our attention and gives us a hint about where to look in the picture (Rydin, 1976b). Understanding

of motives,

causes and

consequences

To understand an event or sequence of events one must associate or link together the causes and consequences of the event in a given order. If all the links in a causal chain are presented in chronological order and, furthermore, occur in a relatively short span of time, viewers and listeners will probably have no great difficulty grasping the event. Many television programmes — 177

thrillers, detective stories and the like — have considerably more complex plots, however. Children under seven years of age tend to recall action and the more eventful scenes in films and stories. In the case of thrillers and "whodunits", preschoolers generally recall acts of aggression — the violence — best, but not the motives or consequences relating to them. Neither do young children seem to have any insight regarding emotional reactions. The next step in their development is that they start taking note of the consequences of an act or acts. Finally, at around age 12, they become able to link together motives, acts and their consequences. While this progression holds for the most part, certain deviations may occur that have to do with the structure of the film material. If, for example, motives are more vividly portrayed than consequences, even very young children may be able to grasp and recall the motives rather than the consequences (Collins, Berndt, and Hess, 1974). Preschool children find it easier to perceive consequences because they are often the concrete results of, or responses to, actions. This bias in favour of the concrete is even reflected in preschoolers' moral judgements. Piaget as well as Kohlberg, for example, observed that children often make somewhat bizarre judgements of others' acts because they base their judgements on tangible, observable behaviour rather than on underlying motives (cf. Piaget, 1965; Kohlberg, 1969). Thus, young children are likely to want to punish a person who accidentally drops three teacups on the floor more harshly than a person who deliberately smashes one. Preschoolers can, however, perceive motives if they are made clear and are presented adjacent to the actions they precipitate. Nevertheless, misperceptions of motives would seem to be common far up into the school years — probably because in many films (thrillers in particular) the motives are deliberately withheld or enshrouded in order to heighten the suspense. Maintaining continuity between motives, acts and consequences is naturally difficult if a film is repeatedly interrupted for, say, commercials. Collins (1973) showed two versions of a violent film to children of eight and older. In the one version, the children saw the entire film without interruption, so that the continuity between motive, act and consequence ramained intact. In the other treatment the film was interrupted by commercials, which were placed immediately following scenes of violence. This latter treatment produced a greater degree of aggression in the children, especially the 8-yearolds. No such difference between the two treatments was apparent among the older children. Thus, it appears that younger children tend to react 178

aggressively when they fail to understand why violence was committed and what it resulted in, i.e. when the causal chain is broken so that acts are separated from their consequences. Older children, able to link the elements together despite the commercials, seemed less influenced by the violent content. How experience and identification

influence

comprehension

Age and programme content or form are not the only factors to affect comprehension. Social background and differences in knowledge and personal experience also play in. There is evidence, for example, that social background influences how people identify with events, characters and settings in programmes. For another study by Collins, two versions of the same story were produced. In the one version the story took place in a white middle-class setting, and the characters behaved according to middle-class norms. The other version was set in a black, working-class milieu. The programmes were shown to audiences of black and white children of different ages (8, 11 and 13 years). White children showed a better understanding of the story that took place in the setting familiar to them, and the same was true of the black children. It should be noted, however, that despite these differences that seem to be totally determined by different social experience the age differences found in other studies were confirmed (Collins, 1979). A Swedish study of understanding an informative programme (Hedbom, 1970) also found differences that could be attributed to the children's social background. Children of highly educated parents absorbed the content more easily than others, as did children who had attended nursery school and kindergarten. Sex differences, too, have been noted. Leifer et al., (1971) tested the hypothesis that understanding of a film may be related to the extent to which boys and girls identify with various characters. The assumption was that girls would show a better understanding of episodes in which girls and women played leading roles, while boys would better understand episodes in which men and boys had dominant roles. The hypothesis was not confirmed, however, perhaps due to the children's age (10 years and younger). Most probably, the children had not yet become discriminating with regard to gender, nor were their own sex-roles yet firmly established. Findings of other studies indicate that adolescents and adults are more receptive to the motives, acts and feelings of characters of their own sex, which in turn influences their perception and comprehension.

179

Schemata for

understanding

Several researchers posit that a kind of schema, or set of expectations about programme structure is activated when we watch television. Such schemata are formed in the preschool years, when children become acquainted with tales, stories and television programmes {Dorr, Graves, and Phelps, 1980; Collins, 1981a). As early as the 1920s researchers studied how people apprehend narratives. Listeners to stories that were illogically structured or contradictory tended to "correct" the story in the re-telling so that it was better organized and more logical. The longer the interval between hearing and telling, the more "ideally" restructured the story tended to be (Bartlett, 1932). During the 1970s the idea of schemata has again surfaced in psychological research, especially in cognitive psychology. Schank and Abelson (1977) speak of scripts, by which they mean preconceptions of how stories are structured. These preconceptions are activated in the understanding process. People match what they see and hear to pre-stored groupings of actions that they have already experienced. For example, when we see policemen in a film, we begin to activate a number of expectations about their behaviour and characteristics based on previous observations: e.g. that they wear uniforms, drive cars with sirens, bear weapons, etc. As early as at the age of eight or nine children are believed to have developed ideas as to how stories "should" be constructed. They expect to find an initiating event, a reaction to the event, and the consequences of the reaction. If these elements are missing in a story, the children will fill them in. Children most readily understand and can recount their impressions of material that is presented in the "ideal" sequence. When the structure deviates from the ideal, the story is "corrected" to match it (Dorr, 1980). Similarly, viewers have an idea or model of the "ideal" television programme. Before one has developed such a model, one is not disturbed by deviations from it, e.g. an incoherent or randomly assembled plot, distortions in time or parallel subplots. Once the viewer has developed a schema of how programmes "usually are", deviations interfere. Such interference has been noted among audiences of 9- to 10-year-olds. The schema is initially quite simplistic and stereotyped; the plot follows a straight line, and the main characters are perceived in a stereotyped fashion. Collins (1981b) found that very young children have difficulty grasping complex character delineation; e.g. "masters of duplicity" (that is, persons who at first seem good, but later disclose evil motives) puzzle children. Young children cannot perceive such double-faced characters since it does not 180

correspond with their stereotyped and simplistic idea of how characters behave. As they grow older, viewers' schemata or expectations become more differentiated, extending to include more types of narratives and more nuances in television role patterns. Even if younger and older children have precisely the same experience of narratives and television programmes and identical knowledge reserves, however, they will nevertheless apprehend and understand television programmes differently. Older children are able to use their knowledge in a different way. They can discern deviations from the standard narrative techniques and grasp their significance (Collins, 1981b) Some have even posited some kind of universal schema of the most general narrative form of all, i.e. the consciousness that stories have a beginning or initiation, that various actions and reactions ensue, that attempts are made to resolve the situation, and — finally — that some form of outcome or solution obtains (cf. Mandler and Johnson, 1977). Beyond this there are features specific to "television reality", standard structural characteristics of television programmes. Such models or patterns are probably not universal and general, but more specific and based on the particular programmes and dramaturgical traditions the viewer has been exposed to. One may, for example, assume that American and Swedish children will have somewhat different models. American television carries a tremendous amount of mass-produced action-adventure and detective stories which follow standard "formulas". They follow a prescribed sequence with standard plot development and elements of suspense or excitement, standard character delineation, etc. (Lindung, 1981). Frequent exposure to the "formula" probably gives rise to a specific set of expectations as to what certain genres should contain. But, in countries like Sweden, where programming is less standardized, there is a rather wide range of programme types in respect of form and content. Viewers have presumably less definite expectations. Such models or sets of expectations may possibly be developed over an extended span of time. American programmes may follow one "formula", Swedish programmes another. So far, however, we know all too little about how viewers develop "media literacy" and how their awareness and expectations continue to develop in the course of their careers as TV viewers.

181

Summary and discussion Despite the fact that studies on media comprehension sometimes lack comparability because of differences in programme content and format, we can still trace some very general development trends in TV comprehension. Previous reviews of the literature have claimed that preschool children's understanding, especially that of 4- to 5-year-olds, is often fragmentary and erroneous. They have a tendency to experience stories as consisting of unrelated episodes rather than as coherent wholes. Furthermore, preschool children have a limited attention span as well a limited memory capacity, which also contributes to limit comprehension. But one also has to keep in mind that film and television for children rarely fulfil the demands of structure and logical continuity. Production techniques such as cuts and time- and space-lapses are often used and limit the viewer's capacity to create a meaningful whole of the images on the screen. Rapid speed and "action" also act in this direction. Too little time is allowed for reflections and synthesizing thoughts. One can suspect that programme elements of this kind are the main reasons for poor comprehension. But, we may in fact have underestimated small children's ability to comprehend televised information. If the medium were used more in line with children's developmental level, even the youngest viewers might follow the story-line and gain a more complete understanding. Their comprehension might not be so deep and detailed but schematic or general. With time, however, the "shell" will be filled with more facts and details. Preschool children have a tendency to pick up irrelevant information, but mainly when it is dominant or especially interesting, or if the story-line is difficult to follow, or if the essential content is too complex or unfamiliar. Comprehension increases considerably somewhere between the age of 5 and 7, which depends mainly on the shift from the preoperational to the concrete operational stage of development. Thought processes become more flexible and the memory capacity is improved. Programmes, even for young children, often have an underlying meaning which is not so easily observable. Such messages are difficult to grasp because conclusions beyond the information given are required. Interpretations of motives, actions and consequences are different from those of adults. Children mainly notice the observable actions, secondly the consequences. But they cannot discern motives, especially when they are subtle. 182

A voice-over speaker text can be used as an extra source of information in order to facilitate comprehension. Picture information and dialogues alone may be too fragmentary and cause misconceptions. Comprehension improves considerably during the elementary-school period and reaches a peak at 11 to 12 years of age. Thereafter, improvement is insignificant. Eight-year-olds have difficulties understanding certain programme types, such as adventure films or thrillers, when the story-line is complex and characterized by time-lapses and subplots. On the other hand, children of this age are now able to go beyond the information given, i.e. make conclusions and connect sequences of events. Furthermore, they have an improved attention and memory span. If programmes are packed with information, children of about age 12 or older tend to absorb information selectively in order to concentrate on the more central content. Television's long-term cognitive

effects

Mass media researchers and philosophers have raised the question that television might have its own "language". That is, television uses codes, techniques and styles which should be interpreted according to special rules. This television "language" is to be acquired as one grows up with television. If such a "language" exists, the next question is whether it is universal or not. So far, we do not know enough to answer this question. However, in some countries, such as the United States, one can certainly distinguish some kind of a TV-specific "language". People must learn that certain techniques, formats and symbols should be "read" in a specific way. For example, children who grow up with Sesame Street seem to apply a selective attention pattern, which is shaped by experience with the special format of this programme. We tend to look at the screen only when we expect something interesting to happen (cf. Levin and Anderson, 1976). Knowledge of the conventions of television may of course facilitate understanding, but may also make the viewer less alert, so that deviations from the normal pattern pass by unobserved. Since children become television viewers at a very early age, researchers have recently raised the question: Can television affect cognitive development? If so, comprehension will not only be dependent on the child's developmental status or on programme features, but the opposite might also be true: cognitive development may be affected by television. 183

In Israel, Salomon (1979) has presented evidence in this direction. Exposing children to Sesame Street for a long period of time, he noted that as the camera makes certain "thought operations", such as zooming in and out, the viewers acquire these operations earlier than children who have no experience of the programme. As television stories seldom follow general logical rules, people are perhaps forced to be more flexible in their way of organizing thoughts. This can actually be noticed in lectures, speeches, etc, according to some media researches. Teachers, for example, give lectures which are characterized by deviations from the story-line, by interposed remarks and retrospective comments much more today than was the case before the television age. The result can be more vivid, even if the listener may have difficulties following the main theme (Novak, 1981). Most researchers are not so optimistic, however, and predict negative effects of growing up with television: television will create people whose thinking is incoherent and fragmentary and whose language is impoverished. Lately, researchers have presented evidence suggesting that children's reading ability has deteriorated during the last decades, probably because of heavy television consumption (Morgan, 1980; Zuckerman, Singer, and Singer, 1980). Other consequences are poor training of abstract thinking abilities and more "image-directedness". That is, youngsters tend to be more inclined to think in concrete and iconic terms. The conversion from the concrete stage to the abstract stage will thus be delayed. These more profound effects of television may, however, differ from one culture to another. In some countries, where television output is more stereotyped and certain genres and formats are repeated year after year, one may more easily trace effects of television on human behaviour, but in countries where the output is more diverse and restricted in quantity and less standardized, the effects might not be so pronounced. So far, these issues have hardly been explored at all. They are extremely important, however, and will open up an entirely new perspective for future research on the effects of television.

184

References Anderson, D. R., Lorch, E. P., Field, D. E., and Sanders, J. The effects of TV program comprehensibility on preschool children's visual attention to television. Child Development, 1981,52, 151-157. Bartlett, F. C. Remembering. A study in experimental and social psychology. Cambridge: Cambridge University Press, 1932. Brown, A. L. The construction of temporal succession by preoperational children. In A. D. Pick (Ed.), The Minnesota Symposia on Child Psychology. Vol. 10. Minneapolis: University of Minnesota Press, 1976. (a) Brown, A. L. Semantic integration in children's reconstruction of narrative sequences. Cognitive Psychology, 1976, 8, 247-262. (b) Brown, A. L., and Smiley, S. S. Rating the importance of structural units of prose passages: A problem of metacognitive development. Child Development, 1977, 48, 1-8. Christie, D. J., and Schumacher, G. M. Developmental trends in the abstraction and recall of relevant versus irrelevant thematic information from connected verbal material. Child Development, 1975 , 46, 598-602. Collins, W. A. Children's comprehension of television content. In E. Wartella (Ed.), Children Communicating: Media and development of thought, speech, understanding. Beverly Hills, Cal.: Sage, 1979. Collins, W. A. The effect of temporal separation between motivation, aggression, and consequences: A developmental study. Developmental Psychology, 1973, 8, 215-221. Collins, W. A. Learning of media content: A developmental study. Child Development, 1970, 41, 1132-1142. Collins, W. A. Recent advances on cognitive processing television viewing. Journal of Broadcasting, 1981, 25, 327-334. (b) Collins, W. A. Schemata for understanding television. In H. Kelly and H. Gardner (Eds.), Viewing children through television. San Francisco: Jossey-Bass, 1981. (a) Collins, W. Α., Berndt, T. J., and Hess, V. L. Observational learning of motives and consequences for television aggression: A developmental study. Child Development, 1974, 45, 799-802. Collins, W. Α., Wellman, A. H., Keniston, Α., and Westby, S. D. Age-related aspects of comprehension and inference from a televised dramatic narrative. Child Development, 1978, 49, 389-399. Donaldson, M. Children's minds. Lund: Liber, 1978. Dorr, A. When I was a child, I thought as a child. In B. Withey and P. Abeles (Eds.), Television and social behavior. Hillsdale, N. J.: Erlbaum, 1980. Dorr, Α., Graves, S. B., and Phelps, E. Television literacy for young children. Journal of Communication, 1980, 30, 71-83. Dunn, G. The box in the corner. London: McMillan, 1977. v. Feilitzen, C., Filipson, L., and Schyller, I. Open your eyes to children's viewing. On children, television and radio, now and in the future. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1979. 185

Filipson, L., and Schyller, I. Pretest of two programmes in the series, "People". Stockholm: Sveriges Radio, Audience and Program Research Dep., 1969. Findahl, O., and Höijer, Β. Fragments of reality. An experiment with news and TV visuals. Stockholm: Sveriges Radio, Audience and Program Research Dept., 1976. Friedlander, Β. Ζ., Wetstone, H. S., and Scott, C. S. Suburban preschool children's comprehension of an age-appropriate informational television program. Child Development, 1974, 45, 561-565. Hagen, I. W. Strategies for remembering. In S. Farnham-Diggory (Ed.), Information processing in children. New York: Academic Press, 1972. Hale, G. Α., Miller, L. K., and Stevenson, H. W. Incidental learning of film content: A developmental study. Child Development, 1968, 39, 69-77. Hawkins, R. P. Learning of peripheral content in films: A developmental study. Child Development, 1973 , 44, 214-217. Hedbom, A. Six to twelve year old's comprehension of a television film. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1970. Köhler, E. M. (Ed.) Grand-children of Lenin. Heisingborg: Wahlström & Widstrand, 1980. Kohlberg, L. Stage and sequence: The cognitive-developmental approach to socialization. In D. A. Goslin (Ed.), Handbook of Socialization, Theory and Research. Chicago: Rand McNally, 1969. Korman, T. A. Differences in verbal-semantic memory of younger and older preschoolers. In Α. V. Zaporozhets and D. B. Elkonin (Eds.), The Psychology of Preschool Children. Cambridge, Mass: M. I. T. Press, 1971. Kvart, A. M. Seven- to eight-year-olds' comprehension of a television program on justice-injustice. Stockholm: Department of Psychology, University of Stockholm, 1978. Leifer, A. D., Collins, W. Α., Gross, Β. M., Taylor, P. H., Andrews, L., and Blackmer, E. R. Developmental aspects of variables relevant to observational learning. Child Development, 1971,42, 1509-1516. Lesser, H. Television and the preschool child. New York: Academic Press, 1977. Levin, S. R., and Anderson, D. R. The development of attention. Journal of Communication, 1976, 26 (2), 126-135. Liikanen, P. The use of TV-programmes and other materials in the Finnish kindergarten. Educational Media International, 1974, (4), 6-13. Lindung, Y. TV-drama knows no bounds: On the suspense and adventure genres. Stockholm: Sveriges Radio Audience and Programme Research Dep., 1981. Maccoby, Ε. E. The development of stimulus selection. In J. P. Hill (Ed.), The Minnesota Symposia on Child Psychology. Minneapolis: University of Minnesota Press, 1969. Mandler, I. M., and Johnson, N. S. Remembrance of things parsed: Story structure and recall. Cognitive Psychology, 1977, 9, 111-151. Meringoff, L. K. Influence of the medium on children's story apprehension. Journal of Educational Psychology, 1980, 72 (2), 240-249.

186

Minkkinen, S. Experiments on comprehension and impact of movies among children and young people. Helsinki: Oy. Yleisradio AB, 1969. Morgan, M. Television viewing and reading: Does more equal better? Journal of communication, 1980, 30, 159-165. Noble, G. Children in front of the small screen. London, Constable, 1975. Novak, M. Television shapes the soul. In R. P. Adler (Ed.), Understanding television, New York: Praeger, 1981. Nummenmaa, T. Children's descriptions of simple events. Helsinki: Suomalainen Tiedeakatemia, 1969. O'Hare, M. Educational broadcasting and cognitive development: Programmes for children in the middle years of schooling. Lancaster: Charlotte Mason College of Education, 1980. Paris, S. G., and Lindauer, Β. Κ. The role of inference in children's comprehension and memory for sentences. Cognitive Psychology, 1976, 8, 217-227. Paris, S. G., and Mahoney, G. I. Cognitive integration in children's memory for sentences and pictures. Child Development, 1974, 45, 633-642. Paris, S. G., and Upton, L. R. Children's memory for inferential relationships in prose. Child Development, 1976, 47, 660-668. Pearl, D., Bouthilet, L., and Lazar, J. (Eds.) Television and behavior: Ten years of scientific progress and implications for the eighties. (Vol. 1). Summary report. Rockville, Md.: National Institute of Mental Health, 1982. Piaget, J. The moral judgement of the child. New York: The Free Press, 1965. Preisler, G. Finnish pre-school children's comprehension of a Swedish television program. Stockholm, Sveriges Radio, Audience and Program Research Dep., 1975/76. Rydin, I. Children's understanding of television. Preschool children's perception. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1976/77. (a) Rydin, I. Children's understanding of television II: From seed to telephone pole. Moving vs. still pictures. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1979. Rydin, I. Information processes in preschool children: How relevant and irrelevant verbal supplements affect retention of a factual radio program. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1972. Rydin, I. The tale of the seed: Facts and irrelevant details in a TV-program for children. Stockholm: Sveriges Radio, Audience and Program Research Dep., 1976. (b) Salomon, G. Interaction of media, cognition, and learning. San Francisco: JosseyBass, 1979. Schänk, R., and Abelson, R. Scripts, plans, goals and understanding. Hillsdale, N.J.: Erlbaum, 1977. Singer, D. G., and Singer, J. L. Family television viewing habits and the spontaneous play of preschool children. American Journal of Orthopsychiatry, 1976, 46 (3), 496-502. Zuckerman, D. M., Singer, D. G., and Singer, J. L. Television viewing, children's reading and related classroom behavior. Journal of Communication, 1980, 30, 166-174. , e7

Pierre Corset

Television in the lives of French children: A review of recent research

Introduction Research into social communication is relatively recent in France. When television began to develop, around 1955, intellectual circles were content to argue about the value of mass culture or to speculate about the effects of cultural standardisation and uniformity that television was likely to produce in the general public. Later public opinion became critical, inspired by studies in the English-speaking world describing the pernicious influence brought about by the consumption of TV programmes on the behaviour of young people. This critical attitude, at variance with the common experience of televiewers, hid a reserved, if not defensive attitude held by adults, who thus transferred to the younger generations the difficulties they were experiencing themselves in integrating this new medium into their behaviour. It is true that the actual viewing habits of young televiewers, being ignored by schools, was considered to compete with, and thus be harmful to school education (cf. Sultan and Satre, 1982). This delay on the part of research is explained partly by the French cultural tradition. For a long time television was considered a non-legitimate experience compared with the cultural field represented by the book, the theatre, music, painting and even the cinema. Moreover, for a long time intellectuals kept away from this social experience, which they considered too popular, and this lack of recognition did nothing to make television a worthwhile subject for researchers. The only development was in the semiotics of the image, but mainly with reference to the cinema, which had already been accepted as legitimate. About 1970, however, when the media were already politically and economically important, research was developed in France introducing new approaches in the field of communications. The media thus became a recognised subject for research. At the same time the educational sciences also became interested in the phenomena observed in connection with "social" learning. If all learning can be ascribed to any activity accomplished by the 188

"learner", as conceived by the operational theory of intelligence developed notably by Piaget, then it would seem that the "learner" needs no more than simply to observe behavioural models in order to learn. This would explain the enormous influence of the media on young people. Consequently, the behaviour of the young televiewer became the subject of studies. These attached special value to the "learner's" activity, and, taking into account the phenomena of messages being actively interpreted by audiences, succeeded in relativising the media's all-powerful ability to influence, and managed to destroy the image of their mechanical efficacy. This approach is more and more accepted in research aiming at a better understanding either of the modifications of attitudes resulting from the consumption of television programmes or of the psychological processes (perception, image formation, retention) which occur with communication through the media in the structuring of young people's personalities. These studies have set themselves the goal of exploring the relation between young people and television. This relation is integrated quite naturally in their daily lives, without in any way overwhelming them. The pattern of their activities and their social relations, the decisions they take on the use of their time exist beyond and independent of it. Television adds itself to this pattern by adapting itself to it and by enriching it with its own special properties. Thus the relationship of young people to television must be restored to a much broader context which will permit a better understanding of young people's behaviour, attitudes or concepts related to television. In the multiplicity of information supplied by these studies, we will retain here only certain aspects, without being able to reproduce quotations of the remarks made by young people that are the basis of these analyses and constitute their great value. In fact this information was collected from the young people themselves using different methods which, depending on their age groups, appeal to their creativity and develop in them an astonishing attitude in which they themselves think about their own behaviour, bringing in a critical distance. Moreover, the presentation of the results of these studies has an educational aim. Rather than being limited only to theoretical aspects, they invite readers to examine the remarks made by young people, thus introducing adults to a better understanding of the behaviour of young televiewers, at the same time explaining and amplifying the relations they could have with them.

189

Reception situations Before school age and parallel to it, the young child is subjected to a set of influences from outside its family which reach it in its home. These influences make themselves felt directly on it and indirectly by the changes brought to its family life. It is the latter that we will first take an interest in, when considering the place held by television in the life of children. First, however, it is appropriate to take the facts as a starting point. Audience studies permit a first approach to be made in measuring the time spent by young people watching television, and in establishing what the different factors are that occur in this common practice (Centre d'Etudes de l'Opinion, 1982). Audience

studies

On average the TV viewing time of young people between the ages of eight and fourteen is two hours a day: less in spring (1 hour 40 minutes), it increases in autumn (2 hours) and reaches its maximum in the Christmas holidays (3 hours 24 minutes). During school time, most viewing naturally takes place on Wednesday (holiday) and above all on Saturday and Sunday. During the Christmas holidays viewing is long on every day of the week, with a record of an average amounting to 4 hours 4 minutes for one of these days. These average viewing figures, however, mark the differences that exist in young people's viewing habits: 3.5% of children spend at least five hours a day in front of the set during school time and this proportion rises to 23% during Christmas week. As a general rule, the older a child becomes, the more TV it watches: Small children between 8 and 9 spend less time before the box, but youngsters between the ages of 13 und 14 devote a large part of their spare time to it, without there being any notable difference between boys and girls. There is little difference between young televiewers in different regions and different types of housing, as well as between those with different social backgrounds (taking into account the occupation of the head of the family), except that children whose fathers are in the managerial class view slightly less. On the other hand, the socio-cultural level of the parents appears to have an influence on their children's viewing habits. The children of skilled workers or labourers, above all those whose mother only had an elementary education, as well as children in large families (five children or more) view more than children enjoying a more favourable socio-cultural level in their family. The greatest differences in viewing time can be observed during the Christmas holidays. But contrary to accepted ideas, children whose mothers go out to work do not view more television than others. 190

This measurement of audiences only suggests a very abstract approach to the behaviour of young televiewers which in fact covers a set of attitudes and behaviour that is extremely extensive and inextricably bound up with their lives. Reception

conditions

We must begin here by recalling how the young child familiarises itself with its surroundings. Liliane Lurcat describes this process as follows: "Very soon young children know how to distinguish the functions of the objects that are handled around them. They know their face, they know how to identify what is related to food, cleaning, decoration, handicraft work, amusement. At first this classification is quite intuitive, resulting from the repetition of daily gestures. The corresponding concepts gradually graft themselves onto their daily habits. Daily life is associated with all these objects which, by their presence, create and meet needs. The child becomes a consumer of TV from the earliest age, just as it is a consumer of food. Whether it is greedy or choosy, by itself and under the influence of its parents and others, it ends up finding its bearings in the programmes and showing its preferences" (Lurcat, 1982). It is in terms of this permeation of its environment that attention must be paid to the family framework in which the child views TV. When children aged 8 to 12 are asked to "make a drawing of TV", their spontaneous expression provides revealing material on this point (cf. Souchon et al., 1973)*. It is generally the living room containing the TV set that represents the privileged scene of the child's relationship with TV. The furniture placed in the room has a special relationship with the TV set, as the clock in this drawing of the child makes clear the following: "On the left is my room, on the right the clock; I have to go to bed at 9 o'clock"; the clock it has drawn showing ten past nine. The table laid for the meal is also an important element, whether the whole family eats dinner watching TV, or whether the time of the meal is in competition with the TV programme. As perceived by children, the surroundings of the TV set are the basis of a certain emotional climate linked with the participation of the members of the * The research technique involved appealing to the children's free expression by means of games, drawings, the spoken word. In 10 groups (10 children), the children were asked to make „television drawings", which they commented on; the children in 10 other groups were invited to play at television using a camera and a video camera; lastly, 10 other groups were asked to speak on the subject of television. The groups were divided into sub-groups of 5 or 6 children for the video games and the discussion. Finally, the children kept a viewing diary covering one week.

191

family in the TV programme. Thus the festive character that colours the experience gained from the small screen seems to be reflected in certain elements, sometimes imagined, such as the bunch of flowers placed by the mother next to the set. In most of these drawings, the child is represented alone, or with its brother or sister, sometimes the mother is doing household tasks, or again the whole family is sitting in a circle before the set, each one in his favourite position, either sitting on the sofa right opposite the set or even sitting on the floor on the carpet. What the young televiewers experience thus also gives importance to the familiar framework of the TV programme. This big occasion of family life is also the occasion of relations between the members of the family. We will refer here only to those that bring relations between parents and children into play. The choice of

programmes

The choice of channel sometimes gives rise to conflicts of influence revealing what types of relations exist in families (Corset, 1982). Conflicts arise that are settled more or less harmoniously, from case to case. The means of achieving agreement vary greatly from family to family. Often those present with the most authority decide. If the youngest accept as natural the authority of their parents, it is less accepted by adolescents, who sometimes succeed in influencing their parents' choice. In some families the procedures used to establish this choice are not based on the principles of authority, whether the choice is made on a majority vote, or whether negotiations bring about a solution. When conflicts occur, mothers often act as mediator. Whatever habits exist, the family unit is often divided by sports transmissions into two unyielding opposing clans, one feminine and the other masculine, which nearly always wins. Limitations

and

prohibitions

For the youngest children (9-11 years of age) any discussion of the time spent watching TV amounts in most cases to discussing the limitations imposed on this activity. The most important concern of parents, or more exactly of the mothers, is at what time their children go to bed (see Pierre et al., 1982). The child has to go to bed at a reasonable hour. This rule has exceptions during the holidays and on the days before holidays. The requirements imposed by bedtime result in a kind of 'calender of viewing'. The exceptions to the bedtime rule are more or less numerous and more or less controlled, from case to case, and liberality can be very great. Children accept this limitation relatively well and on their own account apply the principles on which their 192

upbringing is based, in this point as in others. But this does not mean that from time to time some small conflicts do not arise, or that some ruses are not resorted to. Adolescents accept less well any limitations still placed on them. It is true that at this age constraints are beginning to slacken, until complete freedom is reached for the young person to view as he wishes. Having to do homework is another limitation on time spent in front of the TV set. Here, too, habits vary from family to family. Most often the mother insists that homework is done before turning on the set. There is more constraint for the younger children than for older children who take their work more seriously and who are less tempted by the programmes before the evening. There are also other things that have to be done, such as household chores, that prevent televiewing. In general mothers are responsible for limiting times of viewing, but it is rather the fathers who appear to have the task of forbidding viewing of certain programmes judged harmful. In certain families the children are allowed to see whatever they like, this being discussed with them occasionally. In other cases the parents do not allow themselves to view productions considered contrary to their ethical beliefs. In general this kind of ban exists for 14- to 15-year-olds and disappears completely later. Accepted by the youngest and by those adolescents that still have respect for paternal authority, this prohibition is based on reasons that remain rather woolly or mysterious, above all for 10- to 11-year-olds. Apparently fathers are not always coherent or in command of the situation. Sometimes not allowing televiewing is used as a kind of punishment, for a bad school report in particular. Absent from some families, it is applied more or less strictly in others. Television and family

life

An experiment with role playing carried out with a group of adolescents, investigating certain modes of behaviour in the family in front of the TV set, has allowed the experimenters to produce an informative analysis (Besenval, 1982)*. Television modifies the task of the family cell, giving a major place to pleasure and to the word in the confrontation of adolescents with the exterior world. * This study deals with the perception of television within the family background. In order to circumvent the familiar difficulty of speaking about this experience with children, work was carried out on the occasion of a theatre-forum session; Augusto Boal's method was used and the study was taped on video. Boal, a Brasilian man of the theatre, undertook to practise alternative stage forms which would appeal directly to the public and would call on them to act out their own oppression in an attempt to find an initial solution by making use of that analytical tool, the theatre. The device is approximately the following: A smal group of actors have prepared a short scene

193

Television makes the family the place where the child and the adolescent confront the exterior world and where questions will have to be discussed that before tended not to be approached. In this sense television reinforces the centring of the life of the adolescent and, even more so, the life of the child on the parent-child-relationship, giving the family an increasingly larger place in the perception of the world. Paradoxically, if television has the reputation of stopping talk, it also forces the members of the family, in particular the parents, to exchange a certain number of remarks that they do not always have the capacity and the freedom to make. The child who asks to be explained why a certain programme is on the banned list demands from its parents a difficult explanation, which is avoided every so often. In this case, television, far from stopping talk, stimulates it, and it would be the family that evades its task. But more fundamentally, it is the distribution itself of roles within the family that is questioned. By suggesting that pleasure is a dimension within reach of everyone, ages and sexes being unimportant, television obliges parents to justify their authority on these questions. If there is a regulation of the discussion of pleasure rather than a ban arising from a certain conception of the family, television confuses the rules according to which it functions. It is perhaps at this level that there is a reason to be found for this "fear of the image", which seems to sustain discussion of the effects of television. The language may be sufficiently coded for one to be able to allow this or that message to emerge selectively, but the image is less easily controllable. It ventures to displace the frontiers of the 'learnable' and the 'non-learnable' within the family, by threatening to provide room for questions that are taboo. Thus parents feel unprotected when faced by the TV screen and the inescapable stream of images passing by. The only alternative is to view or to switch off. Viewing means being in the position of the voyeur and feeling unprotected. One never knows what is going to happen, and one is left without room for maneuver before the unexpected. The impossibility of representing an oppression situation with which the audience is very familiar, in order to act it out. One of the oppressed tries somehow or other to escape but comes up against very strong opposition on the part of the other actors. After this scene has been rendered once, a compere asks the audience if it is possible to react in a better way. He then proposes that the scene be reenacted and that the spectators come and replace the actor. Although the expression „oppression" seems excessive in connection with television and its effects, these role-playing techniques which emphasise primarily the institutional dimension of conflicts as compared with the psychological drama, which places the individual at the centre of the scene, make it easier to perceive what is going on in the bosom of the family with regard to its relations with television. The experiment took place over one weekend, the instructors from the Boal group worked with a group of secondary school pupils who had already had theatrical experience.

194

having any hold on the time of the programme, of postponing it, the constraint of having to express one's opinion on unforeseen questions and the risk of disturbing the regulated functioning of discussion within the family, the claim of young people to freedom or to an explanation, the fear of losing face before certain facts by switching off the TV set and in this way allowing it to be considered incapable of confronting reality or to give an account of one's attitude in the face of this: these are the problems with which television confronts the members of the family.

Relation to the televised programme The effects of watching television on young people depends on the different stages of learning and attitude formation.

The preschool

child

Few studies take an interest in the relationship between children of an early age and television. Nevertheless, the child of pre-school age spends a good part of its day in a framework of domestic life with its mother or at its nurse, in a home most often containing a television set which is on during the day. Thus the child discovers television at a very early age, and it would be interesting to know the forms of this early learning, but conditions of observation based mainly on psychological analysis are long and difficult to put into operation. Observation of the behaviour of very young children in front of the television shows the importance of the adult's role as mediator (Blin-Basset, 1982)*. Left to itself the young child pays little attention to the source of light and sound provided by the television; but one can see it reacting to certain sequences of sound, following the rhythm and dancing to certain tunes. The only movements it seems to follow are those of the adult, the mother or her substitute looking after it. It is through the way the adult looks, guiding and inducing, that the child looks at the screen. We can therefore speak of a triangular relation, as the adult, above all the mother, is the indispensable intermediary in this. Towards the age of two the child's attention is retained by certain very short sequences, or rather by images that it punctuates by quoting what it sees in a way corresponding to the image presented but often also in a way modified or distorted according to desire. These images can even result from hallucinations; this last type of situation repeats itself very often * A doctoral dissertation is now in course of preparation on this subject by Dominique BlinBasset, psychologist. 195

with images (certain animals or even colours) that, for example, make the child afraid. Later, towards the age of four, the perception of the image and of the sound conform more to reality. The child listens and follows the story being told. It also becomes more independent and seems to be able to look at the screen alone for a moment without wanting the presence of adults. But the images shown by television relate a story and can bring into play in the viewing child all the psychological processes and defense mechanisms, as in real life. Television programmes therefore represent for the child a field of identification but also of self-projection, which it directs towards the hero of the story being told. This hero can be the subject of the phenomenon of introjection, a corrollary to projection that can be compared to incorporation, so much does the child "devour" the images with its eyes. This expression, evoking visual sensuality, is appropriate to the situation in which the child approaches the screen so much that it seems as if it wants to get inside it to see more, to see the inaccessible, which arouses the desire to see and to know more about it. The expression "in the telly" often used by the child indicates perhaps its desire to integrate a place and to discover all its secrets. The programmes preferred by the very young viewers (from three years onwards) are for the most part series in episodes or repetition. The pleasure of this kind taken by the child in repetition, which allows it to orientate itself better, recalls the compulsion it feels to have the same story always repeated, with the same words and the same changes. The regularity of the programme allows identification in the sense of recognition, of refinding so as to identify itself then better with the character and to incorporate it. The child that drinks in the words of its hero will one day succeed in reproducing the remarks it has chosen from among those it has heard. With these words, transmitted by the adults on the screen who address themselves to the child in adapted and reassuring language, the child carries around in itself something of its personal and intimate world, which can comfort it in moments of yearning. These words can also contribute to its social development by serving as a transition between the home and the nursery school, where they are recognised by other children; they can thus serve as stabilisers in the anxiety of separation. But the way children watch television is far from being uniform; the more one observes them in front of the screen, the more one discovers different ways of viewing, and perhaps there are as many of these ways as there are children. These viewing habits can vary according to parameters as manifold as age, sex, background or its place in relation to its siblings, and the biography of the child even if it has been a very 196

short one, but above all according to the attitude of the adults in its surroundings — in particular of its mother — towards television. Children between 4 and 6 On the basis of personal interviews with 113 children (between 4 and 6) at two nursery schools, one in the residential suburbs of Paris (66 children), the other in a mainly working-class district of Paris (47 children), Liliane Lurcat, a psychologist, attempted to explore the picture that the child can create for itself of television: "The influence of television on the young child can make itself felt at different levels. At the level of the real, it identifies itself with certain characters who are the heroes or models, and very often games in the playground are embroidered round the actions of these heroes. On the level of immediate judgment, the child formulates its preferences and distinguishes between the nature of the programmes.,It knows that certain programmes are meant for children. It also knows that adults have preferences: tastes and preferences are made explicit in the home, and the child also becomes imbued with them. At another level, it accumulates a set of impressions, influences, and knowledge closely associated with the viewing situation. In this way it familiarises itself with the rather private television world of the adults" (Lurcat, 1982). Encounter with

Goldorak

Another study carried out with 110 children aged between 5 and 6 makes it possible to analyse in depth the relationship between the young televiewer and one figure: Goldorak, the science fiction hero (Lurcat, 1982)*. Children have

* This study involved 329 children who were interviewed, often on their own, in school (infants' school). Three themes were treated: The first theme involves Goldorak, a science-fiction hero, who has hit the headlines, creating enthusiasm in the playground and the uneasy surprise of families. The second involves more current aspects of the relations with television: How can these children give expression to the place television occupies in their daily lives? How can a five-yearold picture to himself the source of the happenings viewed in the home? The third theme involves the intellectual world of children, a world largely fashioned by televised events and scenes. The statements regarding the first theme were collected in the course of individual interviews in the school. Each child had only just resolved conflict scenes of school life, presented in the form of a make-believe game carried out using Fisher Price toy figures. The following situation was then presented, whereby the style of the make-believe game was kept: „Michel says to Bruno: Ί haven't got television, myself, I have never seen Goldorak.' — What does Bruno say? What do the girls say? What do the boys say?" The conversation sometimes developed in a purely transposed form, in the statements attributed to the various protagonists of the make-believe game. But mostly the children steered the game

197

developed a favourable relationship with this figure, as is shown by the commercial success of records, toys and spacesuits. The convergence between the success of the programme and commercial interests means that few children have escaped its influence. The remarks made by the children have been analysed around three themes: the ambivalent figure of Goldorak, his unlimited powers, and the special relationship he creates with the young televiewer. In the first place Goldorak is perceived as an ambivalent figure. His strangeness is undoubtedly one of the factors in the fascination he exercises on the children. Neither a being nor an object, he is "something" unclassifiable meeting an irrational need and perhaps also maintaining a certain moral and emotional isolation. In their description most children interviewed made no mention of Actarus (another figure in the programme) and concentrated only on Goldorak. They trace his portrait at a level of objective realism, describing certain real features of the figure and certain of his actions, but also at a level of subjective realism when they add a personal esthetic or moral judgment to the description. Thus for sixty of these children, Goldorak is beautiful; his beauty is contained above all in his colours, his strength and in his way of moving. But his strange physique is a source of embarrassment made concrete by highlighting one special detail or feature. Sometimes the esthetic judgment is tempered by a moral judgment which lessens the effect of the fascination. This is notably the case with those that find him ugly. It seems that "beautiful" is identified with "nice" and that nobody can be beautiful and wicked at the same time. This description shows the ability of young children to establish a distance relative to the programme, which allows them not only to view but also to judge. The question: "Is Goldorak real?" evokes ambiguous responses from the children; the figure is real in its material aspects, in as much as it is a cartoon programme, but also because it is a disguised person. The children feel the contrivances of the film but they still cannot manage to distinguish between the filmed actor and the drawn cartoon figure. The ambiguity expressed in the contradictory answers is the result of either the fact that it is a repeated transmission or of Goldorak's nature, his humanised appearance leading to confusion with Actarus. in a more personal direction of their own accord. They expressed their relations with Goldorak in a fantasmagorical vein sometimes. The game then took the form of a semi-directive conversation, in which various questions were posed in a rather regular fashion: What is he like? Is he handsome? Are you fond of him? Is he real? Could he attack you? What does he do? Some of these questions were suggested by the children themselves and incorporated into the conversation.

198

Asking children to imagine the opinions of other children about Goldorak permits them to establish a certain detachment from the figure and, by making them leave the intimate relation they have with it, introduces the group through the bias of collective attitudes. The ideas held about Goldorak in the world of children are then revealed. Goldorak is regarded as a violent spectacle, intended to suit boys, who are supplied with subjects for identification and with games from it. The double aspect of Goldorak, being and thing, "real" and "unreal", fascinating and bad, forms an important factor in the attraction he exerts on the children. He creates a feeling of uneasiness, but he is sufficiently attractive to be integrated into the world of imagination of the playground. Goldorak is gifted with limitless powers. Both a being and a robot, he possesses extraordinary strength and powers that can be contained in the TV set. This issue did not emerge spontaneously from the remarks of the children. It was with the question of his existence that the problem of the place where Goldorak exercises his power was approached in the form of possible aggression. Goldorak's actions develop in places that are hard to identify by children, which contributes a lot to his fascination for them and to creating in them doubts as to his real existence: a "being" must " b e " somewhere. In accordance with his fictional nature, Goldorak cannot be localised. The space he occupies is that of the programme; his spatial existence is identified with switching on the TV set and is located inside the box that holds him prisoner. When the child passes fom the objective level to the imaginary, Goldorak's power can be exercised without spatial support. His fictional nature leads to his being localised in a mythical character; the space he occupies is then that of the planets in the science fiction comics. This planetary space is also the space of the dreams into which the child allows itself to be transported by the TV programme. Goldorak's motives for his actions often remain mysterious: he makes war on bad people, he protects the earth. But sometimes one gains the impression that his aggression is carried out gratuitously, as if it were an end in itself. His strength impresses, he always triumphs, which gives rise to unlimited admiration in some children and annoyence in others. The children may be fully immersed in the programme, but it is difficult for them to list the different protagonists in the planetary wars, the weapons used and the ways of using them. In their minds everything remains approximate. Frequent contact with this hero creates in the young audience a very profound relationship, in which are mixed fear, aggression, love and complicity, feelings situated on the level of fantasy in the dramatic experience of the programme. 199

When they claim not to be afraid, the children emphasise Goldorak's aggressiveness and strangeness. They are trying to reassure themselves, to prove to themselves that they are not really afraid. When they analyse the reasons for the fear that their hero inspires in them, they still do not want to stop the programme. Thus, as soon as one goes into the child's personal relationship with Goldorak, one finds oneself in the double dimension of childhood fantasising. The child makes itself afraid and at the same time reassures itself: afraid by imagining the worst, and reassured by remembering that it all does not really exist; reassured also by imagining mutual acquaintanceship between the hero and the child, implying magnanimity in exchange for love. In this case, love is identified with a form of submission. The children say they love Goldorak both for no reasons and for all sorts of them. In some cases this love proceeds from intimacy with the character, resulting from fascination. The programme can also produce pleasure, in this case they enjoy watching; but this type of love already assumes a minimum age. In this last case, breaking to some extent with the personalised relationship with the hero, love for Goldorak develops in the direction of love for the programme. When the children say they do not love Goldorak, this is for various reasons ranging from uneasiness to judgment of his actions. In expressing their complicity with their hero, the children show in what way they experience the fantasy world of the programme. When the impressions are sufficiently strong, they are persistent enough to have effects on the child's feelings and on imaginary interpretations that last long after the programme was watched. These interpretations are presented by the children in accordance with the degrees of a scale ranging from "pure fiction" to claims that it is "true". There are the fictional, the possible, the probable, and the true. Some effects, such as fear of Goldorak, result from experienced feelings. Some children are conscious of them and formulate them in trying to explain them. Their evidence presents a diversity ranging from (playing) with their own fear to a description of their own nightmare. " I was afraid, I saw Goldorak in my two eyes when I was asleep at night," says one of them. Goldorak's violence troubles the child in the same degree as it attacks the child watching the programme insufficiently detached from it, badly dividing up the area of fiction and the real places corresponding to its territory. Thus, some describe their fear of Goldorak imagining themselves attacked at home. In their description, these children pass from an attitude of self-analysis, brought into play in the description of their own emotions, to fantasising on the subject of imaginary aggression, revealing their intimacy with Goldorak, in his material conditions. Their relationship with television is then magic in nature, the screen being able to hold a special protective power. To protect 200

oneself from Goldorak's violence, one has to gain his favour by watching him, this is the idea relatively frequent in the mutual acquaintanceship through the screen. Goldorak acknowledges his faithful televiewers, especially those who possess his effigy. Intimacy with Goldorak sometimes goes beyond the strict duration of the episode to penetrate into the intimate life of the child, and can even continue into its nocturnal escapades. Then the children escape out of the territory of the house, towards strange and fascinating kingdoms, the planets of monsters and robots. To cause fear and at the same time to give reassurance is one of the effects of science fiction films. In contrast to the fairy tales in which the imaginary aggression of the bad wolf cannot be suggested, here real and fantastic elements are interwoven and the child would need help in its search to establish the difference between real and imaginary facts. Some children do not try to make this distinction. They abandon themselves to their love of Goldorak, a fascinating love in which the same reasons cause him to attract and repel. Joining forces with Goldorak causes the child to swing to the other side of the screen, and it becomes enclosed in the world of fiction. It no longer uses the conditional mood appropriate for imaginary statements and describes its journey with Goldorak in the past, present and future tenses. Intimacy with Goldorak then passes from the material level of the child alone in front of the TV set to the level of the imaginary world. Fiction takes the place of the real and overflows it. Simply recalling the episodes with an adult asking them questions leads some children to revive the film and to transform emotional and affective participation into material involvement. The "I have seen if'becomes "I was there". Conclusion Television forms the attitudes of young children and favours the creation of "the implicit", as emotional factors have a large place in it. The young televiewer of Goldorak is ready to love what is intended for him. He is in direct emotional communion with the hero. The relationship with the fictional character excludes the rational dimension, which, however, is not completely eliminated, but which needs to be explicitly invoked to appear. The dramatic experience of the televised programme favours participation that is sometimes complete and that ignores the distance between real and fictional. The children's imaginary world is peopled with science fiction figures which want to see themselves within the child's reach. 201

But contrary to the expectations of the adults, the children's interpretation : not always what is anticipated. It might have disappointed the authors of tb programme, if they were not content with the role of the Pied Piper c Hamlin: He does indeed carry them all away, but to where?

The clarity of the image and its metamorphoses It is characteristic of the image that it is analogous in appearance to the objec it represents. The young viewer has no a priori reason for questioning th reality of the scenes shown on the screen unless his experience and knowledg have already induced him to distinguish forms linked with representatio from the event or object thus evoked. The awareness of a duality betwee what is seen on television and the existence of real things develop progressively in the child and later in the adolescent. Young

children

When listening to the radio or watching television, most young children ar content to see and hear, without worrying about the manner in which th images and sounds are conveyed to them. Some may wonder about the scene of action but the replies can only be dependent upon a knowledg necessitating the intervention of adults desirous of furnishing them wit explanations. Of 42 children questioned about the sources of the programme, 31 showe evidence of some understanding (Lurcat, 1982). Above all the problems o localising what one sees and hears and that of the possibility of seeing wer broached. Amongst some children these questions evoke specific response through associations of ideas which are often tautological: television peopl are in the television. Mostly the area or space of the presentation is identifiei with that of the radio or television set. People are permanently present ii there, or their presence is dependent on the set being turned on. Sometimes the child senses the contradiction. Its difficulty then lies in makin the presence of the presentation in the set coincide with an origin in othe places. But even if it manages to dissociate the production sites from th reception site, localisation of the former remains vague. Often there is als< contamination between the shape of the television set and that of the place o production, which are assimilated into a big box. Some children express the idea that televised events are not real but are drawn whereas the people heard on the radio are indeed real and are to be found ii specific places. The transformation of people into images is sometime 202

suspected without being described, and some children know that devices such as the camera and the microphone are used. Some associate the act of filming with real places, whilst the connection remains imprecise for the radio. In short, children's ideas regarding the origin and manner in which televised programmes reach them are a mixture of magic and technology. This explains their particular experience of science fiction, which supports them in their magic theory. When the slow impregnation from the presentations converges with the information furnished by the child's environment, the child then elaborates its notions concerning the various programme sources made available to it. These simple explanations help to give the child the necessary distance, so that it is not completely at the mercy of its emotions. Children at the age of 10 to 11 Reflecting on the technical functioning of television is considered by children of this age group as a playful activity, which allows them to connect up the observations they make on their screen. But, although television production methods are the object of constant interest at this age, the techniques of transmitting the image appear to be equally beyond their scope of interest CPierre et al., 1982)*. Their type of approach to the medium is on the basis of observations, manipulative experiments, conducted on the technical objects of their daily environment. The vocabulary, the functional principles and methods of production are conveyed to them through the observation of certain programmes — above all the sports coverage, where the cameras can be seen, the animated cartoons, and science-fiction films. With all of these elements, whose informative value is very diverse, children construct logical explanations in order to answer questions concerning technical functions and the production of images. * This and the following as well as other references to Pierre et al. (1982) in this chapter relate to the evaluation of the educational project „Jeunes Téléspectateurs Actifs", carried out in 11 departments over two years. The experimental work was conducted by teams of teachers, research assistants and parents. The study was focussed on the child, the principal source of information being the child's own utterances as it expressed itself freely concerning its relationship with television. The field of investigation having been previously specified during an exploratory phase, a series of three conversations was conducted during the year with a representative sample of the school-child population. The technique employed was that of a free conversation with individuals or with a small group. In addition, meetings with the teachers and parents of the children interviewed were held in order to round off the information.

203

Their reasoning, which is based mainly on analogies, has two principal features: it concentrates entirely on what is visible and is centred on the spectator's point of view only. By taking what is visible into account, children are led to simplify phenomena. For them the professionals of television are only the actors or those who appear on the screen. To explain how the image is transmitted, they refer to telephone lines. The visible presence of the object is sufficient to explain the phenomenon. They therefore only see these phenomena from their own point of view. For them, television thus pursues educative goals just like the family sphere and school. By virtue of these intentions, which are the only motives the child acknowledges with regard to television, it can only tell the truth. This lack of critical distance with regard to television on the part of the child is reinforced by its realistic conception of the televised image. Television makes us see the real world. The image is the object itself seen via the camera, whose only task is to bring the object nearer to us. Thus, by showing the object, the screen proves its existence. However, children of this age begin to discover production features to the extent that their effects are visible on the screen. They thus observe the shifting of the camera as it pursues an action it is witnessing from place to place. A sequence is first of all this changing of location during which time continues to flow. But the child still experiences difficulties in understanding that time and space are not real but only represented, and it does not have sufficient command of the notion of ellipsis. The missing link which helps to maintain these difficulties is the notion of sound and video recording and reproduction, which is a prerequisite for the realisation of the programme and which permits montage activities. Amongst most 10- to 11-year-olds, this notion is in the process of being acquired, along with that of the active participation of the camera in the production of the spectacular, especially in trick sequences or when the camera adds a visual effect to the stuntman's exploits by taking up a vantage position. The camera thus progressively abandons its role of passive witness in order to reinforce the expressive quality of the scene being enacted. This age in which the child increases its observations concerning image production procedures is a sensitive period as far as the acquisition of the notion of represented space and time is concerned. But, intellectually secure in its family and school environment, the child adheres to the conviction that television is credible and that it demonstrates the reality of its assertions by means of the image.

204

Adolescents

of 14 to 15 years of age

For adolescents the world of television, which for 10- to 12-year-olds began and ended solely with the people appearing on the screen, becomes a walk of life where numerous professionals each have their own special field and their own type of work. But their favourite type of approach towards phenomena is from the human angle. Their interest is focussed less on actions than on those who effect them. The daily and personal life of the journalist or the famous reporter is of central importance for them. Knowledge of the technical functioning involved is no longer included amongst the adolescents'intellectual interests and they already adopt a passive attitude of acceptance towards this technological object just like adults. They are satisfied to use scientific vocabulary from school and certain concepts from the field of physics and leave this question up to the specialists, their interest being taken up more by the institutional and human aspects of programme production. In their explanations, the economic and above all the human context of production are incorporated into the description of phenomena and actions: an action has been finalised and its effects can be appreciated. They therefore take a plurality of points of view into account, and not just their own, as was the case with the 10- to 11-year-olds. The right to have one's own opinion and a realisation of circumstances differing from those of others can also be discerned. Television, like the family, the school, or other people, expresses a point of view towards which one should be vigilant and critical. This critical distance serves as a framework for their interpretation of the image. The image is created with a purpose. People behind the camera are trying to achieve effects on the televiewer. The world presented on the screen is thus a fictional world which is imposed on the viewer and influences him. Memory and learning Television disseminates a great wealth of information which children watch without any particular compulsion to learn. They know very well that they cannot retain everything and that the way in which things are retained is determined by certain constants which they have noticed on occasions {Pierre et al., 1982) Forgetting in the course of time Repeats are an opportunity for them to ascertain the reality of forgetting. Other circumstances in the course of which the distant televisional past is evoked are rare. Commercials, like films seen several times, are of course 205

retained. Children experience the fact that a discussion after viewing, like the reading of a text in connection with a broadcast, increases the chances of remembering. Forgetting and

interference

The very young are struck by the interference that takes place through programmes of the same kind; but they, too, experience the fact that the storing of new information involves the impression of partial or total loss of the information stored previously. The constant flow of images, the amount of information absorbed are thus recognised as factors favouring forgetting. Motivation,

emotion and

memory

Children are aware of the fact that selection plays a role in effective recall. For the youngest group (9 to 10 years of age) it is simply the pleasure principle which guides this selection, but it is also the affective and the emotional aspect on the one hand and the feeling of learning something on the other. The oldest group (13 to 15 years of age) say that they are impressed by a programme because it combines an informative or descriptive aspect, in keeping with the real world, with a profound emotional or affective quality. With them there is an underlying idea that selection of what is retained, which takes place when information is being stored, is a function of the individual's tastes and interests. But they rarely recall the mechanism by which interest comes to lend a helping hand in the memorising activity: increased vigilance. The image and

memory

Only the oldest recall the peculiar qualities of the image with regard to memorisation. Although aware of the appeal of the image compared with the text or the discussion, few take into account the fact that the image is also a great asset to memorisation. The majority only consider the direct acquisition of the information stored and have a rather diffuse awareness of how many memories may make their appearence on the occasion of a discussion or the repeating of a programme already viewed. They tend to be astonished by the contrast between the amount of images viewed and the proportion of memorised information immediately available for use. To resume, current televiewing habits give children opportunities to multiply their observations and reflexions on the way in which information is retained or forgotten. However, their conception of memory is not independent of their conception of learning and they do not fully realise the fact that acquired knowledge could be utilised in conditions other than the spontaneous reproduction of the information stored. 206

The model of the trained memory The conception of learning derived from the model proposed by the school system and the rote learning and verbal learning advocates, is a privileged one and impairs the judgments passed on television as a source of learning and knowledge. This conception, which hardly evolves with age, is characterised by great rigidity. Both at the level of transmission and conservation, the importance of the verbal aspect in learning as conceived by children is considerable. The model of knowledge is the encyclopaedia or the dictionary. The spoken word is one of the essential elements which permit understanding of what is seen on the screen. Television is often described as a means of enriching one's vocabulary, and one of the often cited obstacles to the proper comprehension of TV is the use of "difficult words" by those expressing themselves on the screen. The consequence is a tendency to see every message purely from the verbal aspect and to ignore the fact that subject matter can be retained in various ways, especially as the learning world tends to neglect the advantages of the image for memorisation and even to attribute strongly negative characteristics to it sometimes.

Concluding remarks These studies on the child's relationship with television in its daily life, from which we have just presented some important aspects, share the common feature of being orientated towards social practice. They attempt to shed light on the attitudes of parents and teachers towards the child viewer, in order to modify its attitude towards television in a positive way and at the same time to shed light on various aspects of child behaviour. The best thing that can be done at this stage is to recapitulate the conclusion of one of these studies which demonstrates the scope of interest offered by television for the education of children: "If we consider children's way of life and their needs, ... it would appear that television, rather than being a new problem to be solved, is a real chance to be seized. Due to its situation at the crossroads of multiple problem complexes, it permits enriched reflection concerning the world around us. This is due to the concrete, everyday character of the television experience. To explain something by referring to television is to proceed from a subject which is familiar to children and which they find interesting because it concerns their behaviour directly. In addition, a subject matter is being broached in this way which never fails to provoke questions or arouse explicit or latent interests ... 207

Finally, television is a complex object, which itself reflects a host of objects whose more or less transformed appearance it retransmits. The possibilities it offers are thus manifold and the range of questions to be treated covers a large share of what the child needs to know in order to understand the world" {Pierre et al., 1982; pp. 219-20). The child in front of the television may suddenly perceive that there is more to discover in this apparently familiar object than it thought. For the child to do so, it is sufficient that it should have been encouraged to make this type of reflection. The first prerequisite is that educators, instead of merely tolerating the existence of television in a more or less way, should take an interest in it. Such studies should be pursued and certain points examined thoroughly, such as the conceptualisation of television time and space, the notion of ellipsis, that of representation and everything touching on the structuring of the imaginary. However, the most essential research concerns the very young and those under nine years of age, a period during which numerous transformations take place. This knowledge concerning child behaviour cannot leave television programme producers indifferent. As the habits of young viewers become better known, the producers' responsability towards their young audience will become better defined. These studies can without doubt be of use to them for putting into operation the cultural objectives which they are obliged to serve and which link them with all those who are responsable for the education of young televiewers.

References Centre d'Etudes d'Opinion Audience de la Télévision. Présentation graphique des résultats du panel jeunes (population française âgée de 8 à 14 ans). Paris: Centre d'Etudes d'Opinion, 1982. Besenval, P. Télévision enjeu, en jeu, en jeux. Revue Autrement 1982, (36), 39-44. Blin-Basset, D. Les bébés aussi... Revue Autrement, 1982, (36), 83-87. Corset, P. Arrête la télévision, j'écoute les gosses, Revue Autrement, 1982, (36), 10-32. Lurcat, L. A cinq ans, seul avec Goldorak. Le jeune enfant et la télévision. Paris: Syros, Collection Contre-poisons, 1982. Pierre, E., Chaguiboff, J., and Chapelain, B. Les Nouveaux Téléspectateurs de 9 à 18 ans. Entretiens et Analyse. Paris: La Documentation Française, 1982. Souchon, M., Corset, P., and Gruau, M. Enquête auprès d'enfants des cours complémentaires 2ème année. Une étude sur la réception de la télévision par les enfants de la région parisienne, Paris: Service des Etudes d'Opinion, 1973. Sultan, J., and Satre, J. La Télévision à la Porte de l'Ecole. Paris: La Documentation Française, 1982.

208

Gavriel Salomon

Beyond the formats of television: The effects of student preconceptions on the experience of televiewing

Have you ever read a book without paying much attention to its details because you didn't think much of the author, only to learn later on from friends that you had missed a great book? Have you ever had a friend who for some strange reason read want ads in the newspapers with much interest and found, to your surprise, exciting information in them? Have you ever examined a painting by a great artist and found much depth in it and then discovered that others, who did not like that artist so much, could not possibly see the depth you saw in the painting? If you have, as I suspect you have, then you are familiar with the topic of this paper, namely the way preconceptions, a priori "theories" and attitudes, influence the way one treats new material. My argument is that children, much like adults, have preconceptions of different sources of information, television in particular, and apply these to the messages they encounter. By so doing they influence what it is they experience and the amount they learn. The last couple of years have seen a growing interest in the formats, formal features, and symbolic components of television. Thus attention has shifted from the contents of the medium, such as aggression or educational material, to the more inherent attributes of the medium. One can witness an impressive growth in the number and sophistication of studies which attempted to examine what sense children make out of the various formal features of television (e.g. Huston and Wright, 1983), how formal features affect children's attention and comprehension (e.g. Collins, 1979). Other research has shown that children are capable of learning to use some of the forms of TV for internal, cognitive representations, and what desirable educational effects certain features can be made to have (Salomon, 1979). Televiewing is, however, an activity that takes place in some specific contexts. One is not just exposed to a series of televised messages presented in a contextfree environment. Communicational contexts can be seen as external events, such as time, space, social milieu, and the like. Indeed, watching a movie in a movie theater is not the same as watching the same movie at home on a TV 209

screen. And watching a television show together with a couple of friends is not the same as watching it alone. But contexts need not be seen as only external factors; they can also be seen as internal, psychological events or factors, which are brought to bear on the encountered messages. In effect, as I have tried to explain elsewhere (Salomon, 1981), contexts are very much like mental frameworks, mental "containers", or schemata, into which a particular message is embedded. Schemata in general are viewed as internal structures of knowledge plus the operations that they prescribe for the handling of relevant new information. Stated simply, one has, for instance, a cognitive schema for restaurants that includes all that one knows about restaurants, believes about them and expects from them. That schema also includes operational guide lines as to the way one is to behave in a restaurant ("behavior scripts") as compared with a dinner at a friend's home. The "restaurant schema" is applied to the specific restaurant the person enters. It contextualizes the act by accomplishing two functions: It determines what is normative and what is not (you may take your shoes off at a friend's house but not in a restaurant), and it influences the meanings that specific events in the restaurant have. (Candles in the restaurant have a different meaning than candles in a church.) Children, similarly, have mental schemata pertaining to television and they apply them to the medium. We could say that their "television schemata" contextualize the televiewing process. Viewing the context of children's televiewing in this way may well reflect the point made by Schramm, Lyle, and Parker (1961) more than twenty years ago: "The chief part television plays in the lives of our children depends at least as much on what the child brings to television as on what television brings to the child" (p. 74). The argument presented in this paper is that exposure to television's contents and formats takes place within contexts of internal, psychological factors, particularly preconceptions concerning the nature of the medium and of televiewing. Such contextualizing of internal factors serves as an anticipatory mental schema which the child actively applies to his or her daily encounter with the screen. By applying his or her preconceptions, the child controls how she or he processes the messages and what it is that she or he chooses to process.

Preconceptions as contexts The internal context that the child brings to the screen is composed of a fair number of factors. First, children come to view television knowing something about the medium. By age six, as has been found by Gardner and his 210

colleagues, children are quite familiar with the language of the medium, its symbolic and grammatical forms, even though nobody has tutored them in this language (cf. Landry, Kelly, and Gardner, 1982; Morison, Kelly, and Gardner, 1981). Second, children come to the screen knowing something about the world. Without that knowledge they would be unable to comprehend much of what they encounter on the screen. Applying that knowledge they interpret the messages, make distinctions between fact and fiction, reality and fantasy, and the like. Third, children's cognitive development serves as a context. Depending on their general level of cognitive development, children pay more attention to some message components than to others, and understand messages in one way rather than in another. Thus, for example, Collins (1979) found that second- and third-grade schoolchildren show a far more limited and fragmentary comprehension of story material on television than eighth-graders and are hardly affected at all by the overall structure of the story. It makes little difference to the younger children whether the story is simple and straightforward or whether it is complex and jumbled. It does make a large difference to older children or to younger ones who have particularly well-developed abilities. Fourth, specific abilities, too, are contextual factors. We have found that children with poor mastery of particular skills (e.g. difficulty to focus on a single item embedded in a larger display), can, under certain circumstances, internalize a television form such as zooming in and out, and use it then as a mental skill (Salomon, 1974). Such contextualizing factors clearly influence the way children handle television messages, but they do not tell the whole story. Children bring to the screen not only knowledge and abilities: they also bring to the screen certain a priori perceptions of television and what it is all about, what should be expected of the medium, and what they can expect of themselves as televiewers. As with all schemata, the ones that pertain to television also include not only stored knowledge about the medium, its nature and attributes, but also so-caljed scripts that prescribe how television is to be handled. Such scripts may suggest how seriously television should be treated, how much mental effort should be invested in processing its material, and the like. Thus, we are not dealing just with "television schemata", but with "televiewing schemata". Equipped with such "televiewing schemata" that include preconceptions, expectations, and operational scripts, the children influence the way they experience the message. One may say that, to an extent, the children contextualize the medium to be experienced in one way or another much like readers influence the way they experience a novel. It is the reader who decides on the basis of his or her expectations, knowledge of the author's other 211

novels, mood, and attitudes toward that genre of literature, whether she or he will read the book carefully and in great depth, or whether she or he will read it shallowly and without the investment of much mental effort, whether she or he will not find in the novel deeper layers of meaning, and the like. The preconceptions with which one contextualizes televiewing can pertain to a number of related issues: — to the whole medium as a source of stimulation (Does the majority of children perceive television as a source of entertainment or as a source of information or learning material?); — to particular classes of content (Is Sesame Street differently perceived than the news?); — to the task of televiewing, that is to the functions it is expected to accomplish (Do you watch to learn or to "kill time"?); — to the viewer him- or herself (Does one perceive oneself as an able televiewer or as somebody who needs to work hard to get pleasure out of televiewing?). Such preconceptions are of course coupled with corresponding attitudes and "theories" concerning the nature of the medium. Indeed, we found that those who claim that television does not require much mental effort to understand because of its life-likeness are also the ones who attribute to the medium such qualities as "shallowness"; they do not tend to expect much learning to result from televiewing. Research on the way televiewers' preconceptions contextualize the television messages above and beyond the role played by specific contents and message formats is still in its infancy. But research in other fields, notably cognitive social psychology, is available to us and should be examined for its relevance to the present concerns. An important and repeated finding is that people's a priori perceptions of a message — that is, the way they contextualize it — influence the way they treat it and what they get out of it. Of particular interest are the studies of Langer and her associates at Harvard University (.Langer and Benevento, 1978; Langer and Imber, 1979; Langer, 1981). One of their major findings is that people who perceive a message on the basis of a few structural, surface features, as familiar and overlearned, treat it "mindlessly". They appear to forgo the careful examination of the information in the message and respond to it without much thought. They think that since the message is so familiar, at least in structure, that it is sufficient to employ processing skills which they have already mastered to automaticity. (Such processes, it should be noted, are not voluntary: they are controlled by stimulus attributes and can yield only that information which does not require "deeper" — i.e. voluntary — processing.) 212

When people behave "mindlessly", they do not bother to examine the material carefully nor do they bother to elaborate the material in their minds, that is to employ mental processes that are not automatically mastered and whose application to the message is voluntary and effort-demanding. Nisbett and Ross (1980) report findings showing that "once people have formed a strong theory or expectation of what the data will look like, they sometimes fail to examine the particulars of the case" (p. 71). The influence of preconceptions works, however, in two ways: The expectation that a message ought to be deep and effort-demanding, or that it is difficult, seems to call for a script that says: "Invest more mental effort in processing the message!" — even if the message does not really justify this added effort and could be handled "mindlessly" or shallowly. It works the other way round when people preconceive a message to be "shallow" and undemanding. This does not mean that such a priori "theories", or schemata, always preemt the close examination of data. We do not only assimilate and fit data into preexisting schemata: we also accommodate our schemata to new and unexpected information. Thus, even though we may contextualize a particular message in a schema, the operational script of which prescribes uninvolved treatment of the message, we may still discover that the message entails more than was expected and treat it accordingly. However, much may depend on such factors as the availability of alternative conceptions. (How many alternative conceptions of Bergman's films do you actually entertain? How many of the American TV series Dallas?) Motivational factors also play an important role in determining whether we will change the way we contextualize a message. There is the fear of committing an erroneous judgement, for example, that others may laugh if one misjudges Dallas as "deep". People often have a desire to stick to a particular "pet theory" to which they attach much importance. Another motivational factor is the extent to which a schema is frustrative or undesirable, thus ready to be replaced by another (Kruglanski, 1980). Preconceptions are more likely to prevent the close examination of inconsistent information when no alternative conceptions are available, when the available ones are deemed valid, and dear to one's heart, and when they satisfy their user. In such cases the specific demands of the stimulus material may exert little influence on the way the material is processed, relative to that exerted by preconceptions. The messages of television constitute a distinctive class, easily distinguished from those of other media, and children may have clear and strong preconceptions concerning the way these messages should be handled. Such 213

preconceptions could serve as internal contexts that prescribe the depth with which any particular class of programs should be processed. It can be hypothesized that to the extent to which television programs conform to the context of preconceptions and expectations and thus are perceived as structurally familiar, the preconceptions will affect the efforts with which the presented material is processed, even if it warrants otherwise.

Children's preconceptions of TV and the way they process programs We have carried out a series of studies to test some of the ideas presented above. Three of these studies are described below. Study 1: Preconceptions

of TV and print, effort and learning

In the first study (Salomon, in press), we wanted to learn something about children's preconceptions concerning TV relative to print. We asked ourselves, for example,whether children think that television is more real-tolife than print material with the same content. We know from the work of Gardner and his associates (Morison, Kelly, and Gardner, 1981: Landry, Kelly, and Gardner, 1982) that children make certain reality-fantasy distinctions earlier with print material than with television: considerations of quality, authenticity, and veracity are rarely expressed with respect to TV. We thus expected the children to regard TV as life-like, and therefore as "easy stuff". We also asked how efficaciously do the children learn from each medium. The work by Bandura (1982) suggests that the greater one perceives efficacy in performing a task, one's belief in one's own ability to carry it out, the greater the investment of sustained effort. But we had reason to suspect that when efficacy is relatively high, effort investment begins to decline. When a person believes that he is really good at something, than performing a task is easy for him, and then he is likely to invest less rather than more effort in performing it. Finally, we wanted to know how children explain success and failure in learning from each medium. Does success and failure mean the same in both media? We asked 124 sixth-graders about these issues, and then divided them into two random groups, one that saw a TV film and one that read the equivalent story in print form. We did everything we could to make sure the two versions were indeed comparable while retaining the uniqueness of each medium. We wanted to know: How much effort do the children report that they invest in comprehending the TV film and how much in the print story? 214

To what extent do their reports reflect their general preconceptions of each medium, and how much are the inferences they generate* related to their reports of effort investment? The results were very much as expected. Children reported perceiving TV to be far more realistic than print, and themselves to be more efficacious in learning from TV than from print. There were clear differences between the causes they chose to explain success and failure in TV and those which they chose for print. Success in comprehending TV was attributed by most children to the ease of the medium, whereas success in comprehending print was attributed mainly to the readers' smartness. On the other hand, failure to comprehend TV was attributed to the "dumbness' of the viewers, but failure in reading a text was explained as a function of its difficulty. Clearly, the two media are very differently perceived by children. The reports of effort investment in comprehending the two versions of the story conformed to preconceptions and correlated with them: Less effort investment was reported by the children who watched the TV film than by those who read the printed version. Perhaps most importantly, when learning scores were compared no differences were found with respect to simple memory, but a statistically significant difference was found with respect to inference generation. Not surprisingly, it was the print group, the group that reported investing more effort in comprehending the story, that generated on the average more inferences. Similar findings were reported by Meringoff (1980). Assuming that the two story versions were comparable, it turns out that the expenditure of more effort, a function of initial preconceptions, leads to the generation of more inferences. As TV is preconceived to be realistic and easy to comprehend (you must be dumb not to understand it), little effort is invested in processing it in actuality, leading to poor learning outcomes. Interestingly enough, as we expected, perceived self-efficacy in reading correlated positively with effort investment and with learning from the print version, but the correlations were negative in the case of TV! In other words, those who perceived themselves to be good readers, tried harder and learned more from print, but those who thought they were great televiewers, thus they thought TV was easy for them, did not try very hard and learned poorly. Only those who doubted their own ability to comprehend TV easily worked harder in understanding the film and thus learned more from the film. Could all this be a function of children's ability? * Inference generation was considered by us as a fair test of learning based on the employment of nonautomatic, effortful elaborations.

215

Study 2: Do abler children learn more from TV? The observation that children treat TV more mindlessly than print, apparently a function of their preconceptions concerning the nature of the medium, raises the possibility that ability may have an influence here. After all, one would expect abler children to do better despite their predispositions, thus to adjust the amount of effort they invest in processing a TV program to its demands even if their predispositions prescribe otherwise. Or are these children as captive of the context of predispositions as everybody else? The study was carried out by Leigh (in Salomon, 1982). We first tested 94 sixth-graders about the effort they generally expend in televiewing and in reading, and about the amount of effort expenditure in each medium which they think is worthwhile. We also asked them about their perceived selfefficacy in remembering and in generating inferences from TV and from print. Then we divided them into three groups, much as we did in the first study. One group watched a narrated TV program, another read the program's soundtrack (the story itself, as told by the narrator) with additional descriptions of the necessary scenes and actions. This yielded a rather difficult text, both in terms of the language and in terms of the grammar. The third group was given the same text, rewritten in daily language with simple, short and active sentences. The third group received then a relatively easy text. Following exposure to the TV story or the texts, the children were again given a questionnaire pertaining to the specific effort they invested in viewing or reading, followed by tests of inference generation and of recall. We had the general reading ability scores of all the children from their school records. The results concerning the differential perceptions of the two media were as expected, and replicated the ones obtained in the first study: The children reported investing in general more effort in reading than in televiewing. Texts make them think more and concentrate more. Low-ability children invest, on the average, more effort in both media than high-ability ones. However, we found no correlation between effort expenditure and ability: but effort investment correlated substantially (r = .55) with perceived worthwhileness of expending effort. Post-exposure measures of effort yielded, as expected, a clear differential picture. Much effort was invested in reading the difficult text and far less in reading the easy text and in viewing the TV program. Specific effort invested in reading the difficult text was only poorly related to initial perceptions of general effort investment (r= .22), but in the easy-text and in the TV groups the correlations were much higher (.65 and .57 respectively) suggesting that preconceptions play a much more important role in determining the way material is handled when an easy text or a TV show are encountered. On the 216

other hand, specific effort investment in the difficult-text group correlated highly with the perception of worthwhileness in expending effort (r = .66), but not so in the other two groups (r= .32 in the easy-text group, and r = .27 in the TV group). These correlation patterns suggest, consistently with our theory, that in the case of TV as in the case of an easy text where material can be handled without much effort, preconceptions play an important role. When the material encountered is particularly difficult, ability, rather than predispositions, plays the major role in processing the information. Indeed, ability correlated .47 with achievement in the difficult-text group, but only .20 in the easy-text and in the TV groups. Moreover, children's perceptions of how worthwhile it is for them to expend effort in processing material reflects their ability in the case of a difficult text (r= .35), but far less so in the case of an easy text (r = .18), and not at all in the case of TV (r= .008). In the latter two cases perceptions of worthwhileness reflect preconceptions (r = .45 in the case of easy text, and r = .32 in the case of TV). This then appears to suggest that subjective perceptions of how worthwhile it is to invest effort in a source of information are a function of ability in the case of a demanding text, and a function of the context of preconceptions, regardless of ability, in the case of an easy text or in the case of TV. More able students perceive the investment or effort in a difficult text to be worthwhile for them and indeed invest more effort in actuality, but they do not think it worthwhile for them to invest much effort in an easy text or in TV, and hence invest little effort in them. How are these differences reflected in their actual achievements? No clear differences in learning between the groups were observed. Does this then mean that despite differences in reported effort investment, a function of predispositions, or ability, learning remains unaffected? Further breakdown of the children in each group according to intelligence scores revealed the answer. As it turned out, and as can be seen in Figure 1, lowability children in the television group performed quite well: they performed less well in the easy-print group, and even less well in the difficult-text group. The high-intelligence children showed, however, the exact reversal: They performed well in the two print groups (as you would expect them to), but performed rather poorly in the television group. Now, why would highly intelligent children show poor inference generation when they are shown a television story, but not when they read it in a difficult text? Examination of these children's reports of effort investment tells the story. The better able children perceive television to be even less demanding than the less intelligent ones, and report the investment of less effort in 217

viewing than their less able peers. In other words, the abler children look down upon television in general. When faced with a show (a rather demanding one!), while being guided by their preconceptions of how easy TV is, they invest little effort in processing what they encounter on the TV screen, and learn relatively little from it. Figure 1: Interaction of learning with ability and media HIGH ABILITY

LOW ABILITY TV

PRINT

There was an additional finding of interest in this study. Reported investment of mental effort in television correlated negatively (r= .39) with children's belief in their ability to remember details. That is, the better they thought they were in remembering details the less effort they invested. Investment of effort in print, on the other hand, did not correlate with this self-reported ability; rather, it correlated positively (r- .62) with children's belief in their ability to generate inferences. In other words, the effort invested in televiewing is designed to combat memory difficulties, not to facilitate inference generation, whereas the effort invested in reading is designed, so to speak, to facilitate inference generation. It appears that the media serve for the children different functions, as indeed has already been shown by Katz, Blumler, and Gurevitch (1974); effort investment comes to serve different goals. All told, we find the better able children to have a rather detrimental view of TV: they think it is undemanding and not particularly worth their mental abilities. When then faced with a new TV show, which by all accounts is quite complex and demanding, they overlook its demands and instead of relying on their well-developed abilities (which they use so well when they are faced with the very same material in print), they rely on their preconceptions. And since these preconceptions prescribe shallow treatment of TV, their learning outcomes are far poorer than what they could have been. Could all this be changed if we changed children's perceptions of the medium, or at least changed their perceptions of what can be gotten out of it? 218

Study 3: Don't let them just watch; tell them what to do! The observation that children have certain preconceptions about television and that they tend to process televised material more shallowly than comparable print material does not yet constitute sufficient proof that these preconceptions serve, indeed, as contexts that influence the depth of processing. Could it be that not the a priori perceptions lead to shallow processing but rather the inherent nature of television? It has been claimed that television, for its fast pace, "crowdedness", pictoriality, and mosaic-like nature inhibits deeper processing (Singer, 1980). Could it then be that the shallow processing of televised material which we have observed in our studies is really not a function of perceptions but rather a simple function of the medium's formats? This argument is of course true to some extent. Television can be processed shallowly and still be enjoyed, something that cannot be said about print. Pictures may indeed be embedded in a wide network of associations without much effort; they do not require the generation of images, nor the decoding of basic elements perceived as not life-like. But this does not mean that television cannot be processed more deeply if that is what the context of preconceptions concerning the medium or the task prescribes. If television material does not require much effort for basic decoding, why don't the children invest it in subsequent mental elaborations of the material? Is it all a function of the medium? The third study (Halpern, in Salomon, 1982) was designed to test the hypothesis that television does not really inhibit deeper processing; thus, if we experimentally changed children's perceptions of the task to be accomplished by watching a television program we would be able to demonstrate a change in the way they process the television material. We did not aim at changing the children's perceptions of the medium but of the demand characteristics of the viewing/reading task. Eighty sixth-graders were divided into four groups: Two groups were exposed to a television show (the one used in the second study) and two were given the parallel text. In each case one group was shown or given the material with the pretext that it is "for f u n " (low demands), while the other group was told to watch the show or read the text "to see how much they can learn from it" (high demands). Thus, there were two television and two print groups; within each such pair one was the low-demands group and one was the high-demands group. Measures were similar to the ones used in the previous study. Results pertaining to the children's preconceptions about the two media were very much as in the previous two studies, so there is no need to repeat them again. The reports of effort investment in each medium-task condition and the amount of learning in each are of greater interest now. As we expected, 219

the high-demands television condition led to the reporting of significantly more effort investment than the low-demands television condition. Interestingly, such a difference was not observed between the two print groups. Children reported the investment of much effort in both regardless of the task requirements. Apparently, children invest much effort in reading anyway whether asked to do so or not. The learning outcomes reflected the differences in effort investment: While the low-demands television group learned relatively little, the high-demands television group showed a significant improvement (see Figure 2). Moreover, learning outcomes in the high-demands television group correlated positively with ability (r= .46) but not in the low-demands group (r= .20). Thus, the more intelligent children learned more than the less intelligent ones when told that we want to see how much they can learn from a television show. When they were not urged to learn they learned no more than their less able peers. Figure 2: Relations between ability and learning in the low and high demands

It appears then that typical televiewing, usually done for entertainment, fits well into children's preconceptions or internal contexts which prescribe the investment of little effort in processing. But when children are told to try harder they succeed in overcoming the dictates of their schemata as well as the alledged limitations of the medium and learn from it much more. Similar results are reported by Kwaitek and Watkins (1981) and by Kunkle (1981). Kunkle worked with university students to whom he showed a new television program. He told half of them that the program was designed for a commercial network and half of them that it was designed for public television (usually the more serious network). And voilà: The students in the second group reported the investment of more effort and showed more learning than the ones in the first group. There may be some limitations imposed by the medium on the effort we tend to expend in processing, but the ones imposed by our preconceptions concerning the nature of the medium 220

are of no lesser, even greater importance. Luckily enough, our preconceptions can be modified by instruction, and perhaps also by training. The three studies reported here are the first ones in a series. There are more studies carried out by us, but their results are not yet analysed. The available evidence suggests that children's preconceptions concerning television play an important role in influencing how deeply or mindfully they process the material and thus, how much they really learn from it beyond that learning which is made possible by the employment of automatic processes. It becomes evident that not the medium should be blamed for shallow processing, nor children's abilities. For the medium often carries material that could, even should be processed more deeply, and the children, as can be witnessed from their way of reading, can if they so desire, process material quite mindfully. Their typical way of handling television appears to be a matter of consensually held views ("television is an idiot box") as well as a function of the deceptively looking life-like appearance of the televised material.

The differential roles of children's preconceptions and of television's formats Television's formal features have been shown to influence the way in which programs are processed, and so have the contextual preconceptions that the children apply to the programs. These two sources of influence — one residing in the medium and the other in the viewer — may well accomplish complementary functions. There is, it appears, a certain division of labor between the functions of preconceptions concerning the nature of the medium, a class of programs, or the tasks to be accomplished, and the formats used in a particular program. Formal features and formats increase or decrease attention and they mark meaningful content in interaction with age (Wright and Huston, 1981). Formal features also activate certain mental processes (e.g. bridging a gap in response to cuts), the employment of which enhances the extraction and comprehension of the presented information (Salomon and Cohen, 1977). Some formal features, as Wright and Huston (1981) show, are not really informative; they command attention, particularly of younger children, by virtue of their salience (e.g. "hype"). Other features are more informative, thus are more attended to by older children who seek to comprehend the continuity of a plot. According to Anderson and Lorch (1983), children appear to attend to those parts of a program which are comprehensible to them. They decide on what is and what is not comprehensible partly on the basis of certain formal features. 221

However, as Wright and Huston (1981) argue, "informativeness (of formal features)... is a meaningful property only to the viewer who is actively seeking to understand the content at a level of processing beyond superficial enjoyment. To appreciate features for their informativeness, the viewer must be able to encode content, form hypotheses, and develop a context of expectations" (p. 77). They thus generate the developmental hypothesis according to which children's attention shifts from the salient to the more informative features as they grow older and as they accumulate viewing experience. I would argue that "the context of expectations" without which formal features remain meaningless pertains among other things to the mental effort one assumes is needed and is worthwhile for the comprehension of the presented information. Much depends on the level of comprehension one seeks, and this depends to a significant extent on "the context of expectations". One is likely to disregard potentially informative features or generate only limited meanings from them (as when one bridges spatial gaps but does not generate inferences about feelings), when one preconceives the program to be, say, "childish", or when one assumes the medium to be "shallow" in general. More generally stated, formal features can be meaningful, they can influence the kinds of mental operations that are to be activated for processing, and they can serve as enhancers of plot comprehension. But preconceptions concerning the nature of the medium or of a particular program determine the extent to which meanings are actually derived from the formal features, and how many of the feature-relevant operations (beyond the ones that are automatically evoked) will be employed in actuality. Certain mental activities are so well mastered that their activation in response to particular formats is not really a function of any preconceptions; as they are "automatic" one does not control them, regardless of the way one perceives the medium. For example, Carrol and Bever (1976) found that some formal features are used to "parse" a program. One may suspect that this use of formal features becomes relatively automatic over time, thus unaffected by one's preconceptions of the medium or expectations from it. Other, less well mastered and more effort demanding operations can also be aroused by program formats. Thus, for example, one can generate the inference that two scenes are causally related to each other although separated in time by other intervening scenes. But as such an inference is not automatic, its actual generation may depend on the internal context of preconceptions which the viewer brings to bear on the televiewing situation. Approaching a program with the knowledge that it is not particularly demanding, that it is just for entertainment, or that it is only "kids' stuff", 222

one is unlikely to employ to any great extent or intensity those effort demanding processes which the program's formats call for. Alternatively, when one approaches the program knowing that it is deep, important, or mentally demanding, one is more likely to fully employ the mental operations, effort demanding as they may be, as they are called for by the program's formats. There are a number of studies showing that children generate fewer inferences from a television program (or for a program the ending of which they have to provide) than from an aural or printed presentation of the same material (e.g. Meringoff, 1980; Salomon, 1983; Watkins and Coulombe, 1981). This relative failure to generate inferences from (or for) a television program is apparently not due to the children's difficulties, for they do it well with language material. Nor is it a function of the limiting powers of the medium, for they overcome these limitations easily when so instructed. Rather, they generate fewer and less elaborate inferences from television because they embed the medium in a context which seems to prescribe to them that investing mental effort is not worth their while.

Implications for production and education Research suggests that programs ought to be comprehensible to be attended to (Anderson and Lorch, 1983), and to sustain attention one does not need to employ violence, as simple "hype" and rapid action may suffice (Rice et al., 1982). But attention and comprehensibility are only part of the story. As the research reported here suggests, the depth of understanding a program hinges on the way it is approached, that is — on the context of preconceptions which is applied to it. Although that context exists prior to encountering a particular program on a given day, it is not totally insensitive to changes. One's expectations concerning the amount of mental effort needed to comprehend a program is potentially subject to immediate change as a function of what a program entails (it may be unexpectedly interesting and relevant), and it can undergo more lasting changes as a result of learning. One can learn that television is sometimes less "shallow" or less life-like than generally assumed. One implication then pertains to the way programs are produced, and another to the way televiewers are educated. The work by Langer (e.g. Langer, 1981) mentioned earlier, shows that the more a stimulus looks like it is expected to, thus fits well one's anticipatory schemata or preconceptions, the more are the latter applied to it mindlessly. If the material looks today very much the way it always looked, why not employ a well-rehearsed script which allows me to extract the desired 223

information without much effort? But when the appearance of the material deviates from expectations, these well-rehearsed scripts need to be employed more cautiously, or more mindfully. More processing attention needs to be devoted to the details of the material. R. Anderson et al. (1977) have pointed out that with respect to text reading more mental activity takes place when "... (a) text is incompletely specified, and the reader fills the gap ... when the set of relations expected on the basis of a schema is deliberately distorted by the author ... when the schemata employed by the reader are incongruent with the schemata of the author; or finally when the text is capable of being assimilated to more than one high-level schemata." There is no reason to believe that this general process does not apply to the specific process of televiewing. It follows from here that when a TV program, designed, say, for learning, has the appearance of an entertainment show, it may perhaps command much attention, but it is likely to be processed very much like all entertainment programs are, that is — effortlessly. For such a program would fit well one's preconceptions according to which TV is primarily an entertainment medium, hence "entertainment scripts" would be employed, leading to relatively little non-incidental learning. Our research bears this out. Consider also the learning outcomes of Sesame Street in the U.S.A. as reexamined by Cook et al. (1975). It turns out that children, particularly lower class ones, learn relatively little from that program. Could it be that the program's formal features which give it such a distinctive appearance of being a fun program, deceive the children to believe that it requires little investment of mental effort? The fact that these children were found to learn significantly more from the program when they co-oberserved it with their mothers (see also Salomon, 1977, for similar results obtained in Israel), suggests that learning is enhanced when the program is contextualized as less fun-like. The possible implication that could follow from the above is that programs in which one wants the viewers to invest more processing effort should not have the appearance of typical entertainment. However, this leads to a potential paradox. So-called serious programs have usually a different appearance than the ones designed to entertain. They are slower, poorer with formal features that excite and arouse, they are more didactic, include more monologues, and other elements which often suppress attention and involvement (Huston et al., 1981). This means that when a program includes formal features it appears as entertainment and may be treated effortlessly, and when it lacks these features, thus appearing as more serious, it suppresses attention and interest. Indeed, one would suspect that if Sesame Street had a less entertaining and pleasing format, it might have tought more but would have been watched by far fewer children. 224

The work carried out by Wright and Huston and their associates may suggest a way out of this paradox. The kinds of formal featues that are most frequently found in entertainment shows, the shows that one tends to see as prototypical to TV, are not the informative features but rather those of the "hype" kind (physical movement, rapid pace, sound effects, special visual effects, rapid cuts). Such features facilitate attention and comprehension of smaller children but not of older ones (Calvert et al., 1982). On the other hand, the formal features that attract attention and facilitate comprehension of older children are more informative, although less salient. They signal change of plot, the shift from one plot level (e.g. action) to another (e.g. dream), the expectation of something of importance, the existence or absence of continuity, the resolution of uncertainty, and the like. Programs designed for effortful processing can perhaps be with less "hype" and thus appear to be less entertaining, while still including many formal features of the more "reflective" and informative type. While production techniques may, to some extent, increase effort expenditure in a program, they are not likely to lead to great changes. As our research suggests, even high-ability children are led by their general preconceptions of the medium, overlooking the mental demands of a complex and relatively deep program. On the other hand, changes in children's overall preconceptions of the medium may exert a much stronger influence. For, as I have tried to show, much may depend on how one contextualizes the process of televiewing, what one aims at and how worthwhile one finds the investment of processing effort to be. Children can be taught that TV is not that real-life like (Dorr, et al., 1980), or that a program entails more than meets the eye (Singer and Singer, 1980). Children can be taught to make finer distinctions between kinds of programs, thus learn when the investment of effort is and is not worthwhile. Indeed, Weissberger (in Salomon, 1982) has found that the better university students discriminate between the different demands of TV programs, i.e. the less stereotyped their views the more effort they report investing in the medium. Trying to change children's conceptions of the medium need not be an impossible task. However, there is no assurance that whatever the children learn from even the most thoughtful and well-designed educational program will be used by them in actuality later on. Since TV can be viewed and enjoyed without much effort, why should a child, watching TV at home for fun, employ newly acquired preconceptions that prescribe more effortful processing? Why, for example, should they question the apparent life-likeness of TV's presentations, generate inferences, or connect the messages to their own personal experiences, if they can enjoy the programs without all this? Indeed, while there is no shortage of educational programs that try to educate 225

children and parents about TV, there is no evidence to the best of my knowledge to show that children actually make voluntary use of these programs when they are spread out on the floor at home in front of the TV screen. We are presently trying out an in-school program to train children to differentially invest more mental effort in TV programs by teaching them to raise questions, generate various kinds of inferences, and the like (Leigh and Ben-Moshe, in Salomon 1982). In addition, however, we are trying out a method which may increase the probability that the children will employ what they learn from this program at home on their own. This method is based on the assumption that people try to live up to their own self-images, particularly when that image is socially desirable and is reinforced. Miller, Brickman, and Bolen (1975) have shown that children will behave neater (pick up more litter), or work harder at solving math problems, when they are led to believe that they are neat or hard working. Following this line of reasoning, we are helping the children to build a self-image as "serious and smart televiewers", who can learn a lot from TV shows and who are good at this if they really try. We assume that the incentive for these children to actually engage in more effortful processing of TV on their own is the satisfaction they get from living up to their own image, as well as the knowledge of TV they can demonstrate socially. As the study is presently still in progress, no empirical data can be reported. In sum, it would appear that if programs could be made to appear less like the ones one expects to see, or — even more importantly — children could be taught to approach TV with more differential preconceptions such that they expect themselves to invest more mental effort in particular programs, more processing effort would be invested and more would be learned from the medium.

Summary The argument presented here is that televiewing takes place in a context of particular preconceptions concerning the nature of the medium, its demands, and the kinds of activities that are to be performed with it. We have found that children perceive TV to be "easy stuff", mentally undemanding, and themselves to be extremely good televiewers who do not have to invest much mental effort in comprehending its programs. We have found that equipped with such preconceptions, children indeed fail to invest processing effort in televiewing even when this is called for by a demanding program. Paradoxically, they invest more processing effort, thus generate for example 226

more inferences when given a text, although it is evident that the text demands more effort for basic decoding. Still, and despite the fact that the basic decoding of TV messages is apparently easier, children fail to elaborate TV mesages in their minds, but do it with print material. Somehow the children come to believe that if not much effort is needed for basic decoding, the material does not deserve to be elaborated upon at later phases of processing. The context of preconceptions may affect the way formal and symbolic features on the medium are being handled. Some formal features are handled automatically, regardless of one's perceptions of the medium or attitudes toward it, as a function of long experience of televiewing. However, the extent to which other, less automatically responded to features activate mental processes and lead to better comprehension, may greatly depend on one's preconceptions, for it is them that prescribe in their function as context, with how much depth a program will be processed, thus — the extent to which informative formal features are allowed to realize their potential. It is possible to suggest the hypothesis that the depth with which a program is processed is a function of formal features only when viewers contextualize a program as deserving deeper processing. When they contextualize it as "easy stuff", there will be little depth to their processing and the presence or absence of informative (non-automatically responded to) features will make little difference. Children's preconceptions of the medium can be modified by educational programs such that they will treat the medium less shallowly. The investment in educational programs aimed at such a goal appears to be of greater promise than changing the typical formal features of programs. However, I have reason to suspect that improved understanding of the medium, even when accompanied by repeated training of how to selectively process it more effortfully, is not going to have much of an effect on the way children process TV when they are on their own and watching a program voluntarily. There is nothing in TV that compells the viewer to choose the investment of more mental effort in processing when less can suffice for enjoyable viewing. Apparently, children must come to perceive themselves as "serious televiewers" and try to live up to that image before lasting changes can be achieved. Clearly, though, even this may not be enough as long as TV is perceived by parents, teachers, and just everybody, as a fun devil to be treated as moving wallpaper.

227

References Anderson, D. R., and Lorch, E. P. Looking at television: Action or reaction? In J. Bryant and D. R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Anderson, R. C., Reynolds, R. E., Schallert, D. L., and Goetz, E. T. Frameworks for comprehending discourse. American Educational Research Journal, 1977,14, 367-381. Bandura, A. Self efficacy mechanism in human agency. American Psychologist, 1982, 37, 122-147. Calvert, S. L., Huston, A. C., Watkins, Β. Α., and Wright, J. C. The effects of selective attention to television forms on children's comprehension of content. Child Development, 1982, 53, 601-610. Carrol, J. M., and Bever, T. G. Segmentation in cinema perception. Science, 1976, 191, 1053-1055. Collins, W. A. Children's comprehension of television. In E. Wartella (Ed.), Children communicating: Media and development of thought, speech, understanding. Beverly Hills: Sage, 1979. Cook, T. D., et al. Sesame Street revisited. A study in evaluation research. New York: Russel-Sage, 1975. Dorr, Α., Graves, S. B., and Phelps, E. Television literacy for young children. Journal of Communication, 1980, 30, 71-83. Huston, A. C., and Wright, J. C. Children's processing of television: The informative functions of formal features. In J. Bryant and D. R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Huston, A. C., Wright, J. C., Wartella, E., Rice, M., Watkins, Β. Α., Campbell, T., and Potts, R. Communicating more than content: Formal features of children's television programs. Journal of Communication, 1981, 31, 32-48. Katz, E., Blumler, J. G., and Gurevitch, M. Utilization of mass communication by the individual. In J. Blumler and E. Katz (Eds.), The uses of mass communication. Beverly Hills: Sage, 1974. Kruglanski, A. W. Lay epistemo-logic process and contents. Psychological Review, 1980, 87, 70-87. Kunkle, D. Assessing AIME: An empirical analysis of the relationship between mental effort and learning from TV. Unpublished paper, USC, 1981. Kwaitek, K., and Watkins, B. The systematic viewer: An inquiry into the grammar of television. Children Media project. The University of Michigan. First annual report to the Spencer Foundation, July, 1981. Landry, M. O., Kelly, H., and Gardner, H. Reality-fantasy discriminations in literature: A developmental study. Research in the Teaching of English, 1982,16, 39-52. Langer, E.J. Playing the middle against both ends: The usefulness of adult cognitive activity as a model for cognitive activity in childhood and old age. In S. R. Yussen (Ed.), The development of reflection. New York: Academic Press, 1981.

228

Langer, E., and Benevento, A. Self-induced dependence. Journal of Personality and Social Psychology, 1978, 36, 886-893. Langer, E. J., and Imber, L. G. When practice makes imperfect: Debilitating effects of overlearning. Journal of Personality and Social Psychology, 1979, 27, 2014-2024. Meringoff, L. K. Influence of the medium on children's story apprehension. Journal of Educational Psychology, 1980, 72, 240-249. Miller, R. L., Brickman, P., and Bolen, D. Attribution vs. persuasion as a means for modifying behavior. Journal of Personality and Social Psychology, 1975, 31, 430-441. Morison, P., Kelly, H., and Gardner, H. Reasoning about the realities on television: A developmental study. Journal of Broadcasting, 1981, 25 (3), 229-242. Nisbett, R., and Ross, L. Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, N. J.: Prentice-Hall, 1980. Rice, M. L., Huston, A. C., and Wright, J. C. The forms of television: Effects on children's attention, comprehension, and social behavior. In D. Pearl, L. Bouthilet, and J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties. (Vol. 2). Rockville, Md.: National Institute of Mental Health, 1982. Salomon, G. Internalization of filmic schematic operations in interaction with learners' aptitudes. Journal of Educational Psychology, 1974, 66, 499-511. Salomon, G. Effects of encouraging Israeli mothers to coobserve "Sesame Street" with their five year olds. Child Development, 1977, 48, 1146-1151. Salomon, G. Interaction of a media, cognition, and learning. San Francisco: Jossey-Bass, 1979. Salomon, G. Communication and education: Social and psychological interactions. Beverly Hills, Cal.: Sage, 1981. Salomon, G. Reading and Television. Second Annual Report to the John and Mary R. Markle Foundation. Jerusalem: The Hebrew University, 1982. Salomon, G. Television watching and mental effort: A social-psychological review. In J. Bryant and D. R. Anderson (Eds.), Children's understanding of television: Research on attention and comprehension. New York: Academic Press, 1983. Salomon, G. TV is "easy" and print is "tough": The role of perceptions and attributions in the processing of material. Paper submitted for publication, in press. Salomon, G., and Cohen, A. A. Television formats, mastery of mental skills, and the acquisition of knowledge. Journal of Educational Psychology, 1977, 69, 612-619. Schramm, W., Lyle, J., and Parker, E. B. Television in the lives of our children. Stanford, Cal.: Stanford University Press, 1961. Singer, J. L. The power and limitations of television: A cognitive-affective analysis. In P. H. Tannenbaum (Ed.), The entertainment functions of television. Hillsdale, N. J.: Erlbaum, 1980. Wright, J. C., and Huston, A. C. Children's understanding of the forms of television. In H. Kelly and H. Gardner (Eds.), Viewing children through television. San Francisco: Jossey-Bass, 1981.

229

Part II

Principles, methods and problems of formative research

Keith W. Mielke

The educational use of production variables and formative research in programming

Constructive uses of television No small part of the general concern with television as a teacher stems from the ubiquity of television across America, the huge amounts of time that children and youth devote to watching television, and a growing body of research-based concern about the effects of televised violence and other programming content. What can be done? Institutionalized actions in this arena can be directed to the prevention of harm, to more precise specification and understanding of the problem, and/or to making positive use of the television medium itself. The vehicles for these interrelated actions are: (1) imposed regulations, voluntary codes, and public education campaigns; (2) basic or applied research; and (3) production of constructive programming. The emphasis in this chapter is on the last alternative: turning the power of television to constructive ends. As Schramm (1976) pointed out, the idea of using television in positive ways is not new. He cites testimony given by Paul Lazarsfeld to the U.S. Senate Subcommittee on juvenile delinquency back in 1955: "Let me draw your attention to the fact that everyone talks about bad television programs and the effects which they have; but actually it would be much more constructive and enlightening to experiment with good programs. Why shouldn't it be possible to get reformers and writers together and have them devise programs which everyone thinks would be desirable and beneficial?... You have to get psychologists and writers to meet and work together. You have to have funds to provide programs for experimental programs, regardless of whether a station or network is willing to put them on the air. But the aridity and negativism of much of the discussion which takes place today can be overcome only if it is shown that there is something like a good program, that there are people who can be trained to write and produce them, and that children are willing to listen to them." I think one would have to agree that, in the 28 years following Lazarsfeld's insightful testimony, the positive uses of television have not received the 233

majority of attention, whether indexed by the focus of policy deliberations, by the volume of research, or by programming efforts. Part of the shortfall is no doubt caused by the multiple difficulties in actually harnessing television to constructive ends. In the same sense that it is easier to assess the effects of drought than it is to make rain, the path to constructive uses of television is a difficult one. With a few notable exceptions, the purposeful use of television to produce intended effects has been pursued more (and more effectively) by advertisers than by educators. Goals are specified; the demographic and psychographic nature of the desired audience is researched; the message is designed with great care; various features of the television medium that can enhance the message are fully employed; the message is pretested and revised for maximum appeal and impact; the diffusion of the message is carefully controlled; reinforcing messages are disseminated through other channels such as stores; multiple barriers to the advocated behavior are analyzed and addressed; packaging is made very attractive; and so on throughout the various elements of a complex system that must operate systematically before the intended effect is achieved. It is not a simple process, nor is it uniformly effective, but it is carefully planned with reference to a wide range of variables. When similarly specific goals are sought in the constructive uses of television for children and youth, a similarly complex process is confronted. Instead of firing the gun and hoping that a clay pigeon might be sailing overhead, a level of system-wide planning is required that differs in type and amount from that typically associated with institutionalized education. Two elements of this complex process have been singled out as topics for this chapter: production factors and formative research. The discussion of production factors stresses message design which makes full and intentional use of the television medium's attributes. The linkage between message design and formative research is basic. The more innovative the message design in its exploitation of production factors, the greater the need for trial and revision through formative research. The systemic nature of the process of constructive education through mass communication media should be kept firmly in mind as the two elements of that system are discussed. I am aware that a disproportionate number of examples in this chapter are drawn from the Children's Television Workshop (CTW). There are two reasons for this: (1) CTW programs reflect an above-average interest in the pedagogical use of production factors, as well as a commitment to formative research; and (2) I am most familiar with the CTW examples. The issues, however, are general. 234

Production factors There are points of view in which the significance of television production factors is played down. Proponents here argue that production factors are virtually impotent in terms of communication effects. The term "fancy production techniques" here would not be a compliment, but a pejorative. The imagery would connote sugar coating and slick packaging, all to no constructive end. Although not an exact counterpart, there is an opposing view. The tendency here, on the other side of the fence, is not only to attribute great importance to production factors but to be willing to generalize from one unique blend of production elements to other, dissimilar settings. Research efforts in this camp would pursue questions such as: — What are the generalizable effects of rapid cutting or various camera angles? — Can the documentary format be useful in information gain and attitude change? — What proportion of entertainment is necessary to keep an audience interested in an instructional program? I feel that television production is underplayed in the first instance and overplayed in the second. In a middle zone, production factors can be dealt with constructively. Message designers can call selectively on the vast array of production techniques, when it makes pedagogical sense to do so. Whatever the resultant blend of concept and execution in one task, the message designer will probably need to start dissimilar assignments with a fresh analysis of the new goals, the new target audience, the new content, etc. In this middle zone, production factors are neither to be ignored nor put into a formula for mechanical replications. Research in instructional television (ITV) during the 1950s and early 1960s did not make a strong case for the significance of production factors, at least as far as measurable learning was concerned. This body of research was reviewed and synthesized by Chu and Schramm (1968).They opened each section of their report with a proposition that seemed to be warranted by the available research evidence and then went on to summarize the research base of support for the proposition. Their research base is not resummarized or critiqued here, but their introductory propositions that pertain to manipulable production factors are extracted and listed to give an idea of how production factors in ITY were perceived a decade ago: 235

— There is no evidence to suggest that either visual magnification or large-size screen will improve learning from television in general. — There is insufficient evidence to suggest that color will improve learning from film or television. — Where learning of perceptual motor skills is required, a subjective angle presentation on television will tend to be more effective than an objective angle presentation. — There is no clear evidence on the kind of variations in production techniques that significantly contribute to learning from instructional television. However, students will learn better when the visuals are presented in a continuous order and carefully planned both by the television team and the studio teacher. — Attention-gaining cues that are irrelevant to the subject matter will most probably have a negative effect on learning from instructional television. — There is no consistent evidence to suggest that either humor or animation significantly contributes to learning from instructional television. — Subtitles tend to improve learning from instructional television, particularly when the original program is not well organized. — There is insufficient evidence to suggest that dramatic presentation will result in more learning than will expository presentation in instructional television. — Inserting questions in a television program does not seem to improve learning, but giving the students a rest pause does. — The students are likely to acquire the same amount of learning from instructional television whether the materials are presented as a lecture, or in an interview, or in a panel discussion. Similar sentiments were expressed later by Kincaid and others (1974), although with a recognition that production factors may have something to do with reaching and holding a noncaptive audience: "There is no consistent evidence that making the media more sophisticated enhances effectiveness. For instance, for televised instruction, there is no consistent evidence that color, animation, humor, or dramatic rather than expository presentation improves effectiveness, although it does attract a larger audience — Sesame Street is a prime illustration of this" (p. 8). One of the conceptual blind spots of instructional television in the 1950s and early 1960s resulted from the tendency to look at television as an extension cord for the eyes and ears, bringing viewers the sights and sounds they could 236

have gotten had they physically been somewhere else. In the thinking of that time, classroom "master teachers" would be made universally available via television. Subject matters that made the transition to television with relative grace would be identified as being especially appropriate for television. Quite aside from the implied insult felt by many teachers — if television is bringing in the "master" teacher, what does that say about my teaching? — was the insensitivity to exploiting what television could do best. In a McLuhanistic fashion, the content of early ITY was traditional classroom instruction, reminiscent of early cinematographers planting their cameras in front of the stage to film plays as the content of early movies, giving everyone a front-rowcenter seat. In neither case was the potential of the medium exploited. Along these lines, Salomon (1977) makes a distinction between a televised lesson and a television lesson: "The former is a lesson transmitted via TV. The latter is a lesson which makes use of whatever uniqueness is offered by the medium" (p. 2). Analyses of production variables are thus messagé-design tools working in the service of producing television lessons. My own thoughts about television production factors started to jell in the late 1960s. I wrote the following: "'Fancy production techniques' has almost become an uncomplimentary phrase in ETV, although the poverty of research dealing with TV production strategies stands in striking contrast to the plethora of gross media effectiveness comparisons that leave message treatment variables unanalyzed. The measurement of 'effects' in these comparisons typically utilizes paperpencil tests. The typical baseline of comparison is face-to-face (verbal) instruction. Television will surely have difficulty finding its stride in such a setting....Suppose we looked at potential ETV messages through the eyes of a combination film maker, educator, and communication theorist. We might see ourselves as a packaging operation, processing visual and aural information, with superb control over what got into the package and how it was treated. We would see ourselves as dealing with many encoding systems, not just verbal systems. We would be encoding a wide variety of information, not just factual information for a traditional information-gain test" (Mielke, 1968). This same line of thought was continued later: "At the highest level of message-medium interaction, the raw material would be expanded far beyond the confines of the classroom..., but there would also be extensive and conscious manipulation of the production variables. Careful consideration would be given to the presentational style, the televisual 'statements', the conscious structuring and control of the stimulus to achieve a desired effect. There would be awareness of what might be called the 237

'language' of television, as well as an aesthetic or pedagogical rationale for using one treatment instead of another" (Mielke, 1970). I feel that television production variables seemed to register little effect in the early ITY studies because of a lack of pedagogical rationale for using these variables in the first place. Without pedagogical rationale, unmotivated cutting adds confusion, not excitement. It is only commonsense to surmise that animation will be most useful when its specific attributes that can be helpful are used, and so on. Use of pedagogical rationales to gain benefits from production variables is the domain of message design, when the message is designed for television from the beginning. Lazarus (1977) offers this view: "There are...problems in reconciling content and production. The two overlap the most in questions of pedagogy — not which content to teach, and not the technical details of production, but matters like the sequence and pace of presentation, the best level of detail in scripts and close-ups, appropriate degrees of realism, and so on...." The point of view that production variables are insignificant rests on the deficiency of not exploring pedagogical rationales for harnessing those production factors that can improve effectiveness. On the other hand, the attempt to explain everything in terms of production "formulas" rests on yet another deficiency: not grasping the inherent interaction between content and production. It is possible to envision a continuum that runs from a pole of "high content/low production effects" to the opposite pole of "low content/high production effects." Consideration of the polar positions may shed light on the middle ground of content/treatment interaction, where virtually all television programming for the in-home audience takes place. When television is used merely as an extension cord for the eyes, as in a monitoring situation, very little imposition of a production nature takes place. The flight schedule in the airport, seen live, is rather similar to the televised version of that same flight schedule, seen down the hall. Such effects as can be expected from that message can be attributed to content and not presentational style. Leaping across to the other end of the continuum, one can encounter television programming with no content. Examples include experimental excursions into "television art," where kaleidoscopic-like images can be generated and manipulated electronically by individuals (or even by ensembles working in concert in real time visual approximation to a jazz ensemble). Neither extreme characterizes the television programming that children spend so much time viewing. Here, in the middle ground, the interaction between content and presentation style is typical. Effects of televised content depend 238

in part on how it is produced. Effects of production variables depend in part on the content being presented. Just as it was counterproductive to ignore production variables, it is counterproductive, I feel, to establish prescriptive production formulas that go very much beyond general guidelines. The formula approach is deficient because it is simplistic and mechanical; it lacks a rationale or a theory. For example, to say that animation is appealing or easily comprehended is to stop short of considering why this might be so, and what the relevant attributes of animation are, if any, that could enhance appeal and comprehension. Without such reasoning it is of course possible to produce animation that would not get the anticipated results or to overlook the same conceptual benefits in other formats. Rationales that get into the analysis of animation, such as noting that the ratio of relevant to irrelevant visual cues is usually very high, suggest the conditions under which animation will and will not work efficiently in learning. The typical paucity of irrelevant cues in animation is traceable to economics, where the incentive is to communicate the most intentional meaning with the least possible complexity of drawings. If some figure or animal is irrelevant to a scene, the economic motive is to not put it in. It would add to the cost without adding to the payoff, so it tends to be avoided. This results in a typical style that may have important implications for nonanimated programing as well, but it takes an analytical message-design approach to be able to extract the implication. To make a production "recipe," one has to be able to describe the ingredients. One of the reasons it would be so difficult to predict effects of production variables from a formula is that very subtle differences in a production element can make a lot of difference in the reaction to the program. Artistry lies in manipulating those subtle differences. In taping sessions for television, for example, it is not uncommon to record several "takes" until the director determines that one particular take has a satisfactory blend of attributes. If one would overlay a set of descriptive or analytic production categories on the various takes, it is doubtful that they would be distinguishable. I can't conceive of a set of production descriptors that would be precise and specific enough to minimize subjective judgments, yet of sufficient generality to be useful in describing more than one idiosyncratic stimulus, that would differentiate among most takes in any given group of takes. The basis of discrimination inside the head of the director will elude a content-analysis approach to categorization. To generalize from one production to another via formula, one must know which, from among the multitude of idiosyncratic attributes in a program, are the critical attributes. No matter how exhaustive the list of descriptive 239

attributes, there can be other attributes not taken into account, and these can cause the generalization to fail. Stanford (1974) describes such a failure to replicate, telling of a researcher who"... found that animation and puppets did not appear to be as effective as indicated in the Sesame Street research, but [who] noted that this may have been due to the fact that Sesame Street used professional puppeteers and animators." It is of little help to state a generalization that puppets, generically, either are or are not effective, and it provides little analytic insight to advocate the use of professional puppeteers. The search for a production formula, therefore, may involve more problems than meet the eye. Opposite the "good" that could come from exporting and replicating success stories, or avoiding the replication of failures, stands the potential for " h a r m " that could come from the uncritical imitation of the easily perceived attributes of successful programming or the harm that could come from concluding that it is impossible for some blend of production factors to work, therefore cutting off exploration. Producers, told that the formula says something is impossible, have a way of producing the exception to the rule. The same holds for content. For instance, it would be difficult to imagine a less promising subject matter for a voluntary children's television audience than reading. If dictated by a formula for feasible content, the conclusion would surely be to stay away from it entirely. Yet, The Electric Company series transformed the seemingly unpromising content into excellent programming. Critical to this transformation was a message-design procedure that exploited the various potentialities of production with pedagogical rationales. I feel that most production manipulations are product specific. That is, the effect of one production variable will be influenced by the entire context in which the one variable takes place. Given that different situations by definition have different contexts, the effect of the same production manipulation may well have a different effect or no apparent effect at all. For example, an interesting study, unpublished, was done with a laugh track in the course of the formative research for Feeling Good, a CTW experimental series in health education. A sample reel of nine test segments was produced originally without a laugh track, although most of the segments were humorous by intent. Later, a professional laugh track was added, giving an opportunity to test the effects of the laugh track under conditions of control unusual for formative research. After matching test respondents to the two versions on such factors as ethnicity, sex, age, and education, we compared measures of program appeal, cast appeal, and comprehension. For most of the nine segments, the laugh track increased the appeal of the program and cast. Surprisingly, however, the laugh track version was comprehended at a 240

lower level than was the original version. A simple and general conclusion would be tempting: laugh tracks increase appeal but decrease comprehension. I would be most reluctant, however, to generalize beyond that particular set of segments. Even there, no effect held constant across all nine segments. An enormously complex interaction was taking place, insight into which demanded detailed familiarity with those particular segments. Even the term "laugh track" does not refer to any constant quality, but varies from execution to execution. In terms of message design work for those segments, however, the information did help to understand the various reactions to the two versions. To the extent that some of the enormous variations that occur in children's television programming can be pulled into a tighter circle in which production and audience elements can still vary, but within very restricted categories, it may be possible to venture forth toward some principles of message design that will have predictive validity. Bernstein (1978) has completed a study that goes right to the cutting edge of this kind of effort, being careful to avoid the error of overgeneralization: "The design of children's television programming, both before and after the study, is and will remain very much more art than science. No mysterious attribute formula for the design of appealing programs was sought which would substitute research rules for creativity and artistry. Nor was one found" (p. 4). Using 33 Sesame Street segments as stimulus materials (thus restricting the quantity and variance of a large number of known and unknown production factors) and preschool test audiences in the lower socioeconomic range (thus restricting the variance in audience and audience/program interaction factors), Bernstein explored the relationships between (a) a highly selective list of programming attributes (thus restricting the production dimensions to be considered in the first place) and (b) various measures of visual attention (thus greatly restricting the range of effects under examination). Even under those tight constraints some ambiguities emerged, such as a strong method variance in the measures of attention, and in findings that differed from those of previous research findings (e.g., in effects of programming context). This is not to suggest that this study has more constraints than in the "typical" study of production factors (if there be such a thing); if anything, this study has fewer constraints than typical. The difference is that, in this study, the constraints were recognized and made explicit from the beginning. The nature of Bernstein's findings, within the constraints described above, is suggested in the following quotation: "The segment attribute predicator which best explained variance in preschool visual attention was the kind of storyline used. Affective storylines, which 241

were characterized by dense, active audio tracks and relatively static visual tracks, were related to low viewer attention. Lesson or demonstration storylines, which were characterized by sparse, inactive audio tracks and visual action highly functional to the understanding of segment messages, were related to high viewer attention. Storylines wherein songs were central were negatively related to viewer attention. Additional segment design attributes of Sesame Street's audio track also explained variance in preschool children's visual attention. The use of puns, a highly sophisticated verbal form of humor, was related to low viewer attention, as measured by the slide distractor method. The segment design attribute, adult and muppet principal speakers, which was characterized by much verbal humor, affective storylines, and little visual movement, was related to low viewer attention. Segments with principal speakers who were on screen throughout were negatively related to high viewer attention. Segments characterized by verbal rhythm and rhyme were related to high visual attention" {Bernstein, 1978; p. 94). What kinds of things are known (or can be known) about production and format variables? Do they "behave" with the regularity and predictability that will allow understanding of their effects? If understood, can the effects be controlled? Can we generalize propositions about effects of production variables to the level of policy? What types or forms of production should be encouraged, or financed, or regulated? Such questions have been asked by researchers and government personnel for some time, at least since we became sensitized to the influence television was having on children. It is clear that not very much can be chiseled in granite at this point. If there is a production formula that will reliably deliver certain audience effects, it seems to be a wellkept secret. Once the commitment is made to search for pedagogical rationales to employ all useful production variables, and once the difficulties of trying to establish generalizable production formulas are recognized, it is possible to do a great deal of constructive work in message design for goal-directed television programming. For example, imaginative production work, incorporating analyses of production factors into message design, has been done over the years by the Children's Television Workshop in its Sesame Street series. Lesser (1972) discusses this in his article "Assumptions Behind the Production and Writing Methods in Sesame Street", from which the following examples are selected: — To suggest to children that they can mentally simulate various scenarios - before taking action, a character's thoughts (thinking through alternatives) would be acted out above his head prior to taking any action. 242

— Television's ability to focus attention to certain cues and away from others can be used to help young children differentiate relevant and irrelevant information. — Cross-modal reinforcement (video with audio) can be carefully structured in television production. — Imitation behavior can be enhanced by producing clear and interesting models of target behavior that audience members can actually imitate on their own. — A variety of production devices can be instrumental in directing attention, such as use of surprise, incongruity, animation, pixilation. — Production devices that incorporate a wide variety of visual and verbal humor can be used to sustain voluntary attention. Attention is also facilitated by diversity in characters, content, style, and pace. These insights are not sure-fire formulas; they are guidelines that relate various options in the production process to educational goals for the Sesame Street series. The previous examples of production factors have mostly had their real or potential application in the design of goal-directed television materials. The analysis of production factors also has importance in such areas as regulatory policy and theoretical research, as illustrated in the following two paragraphs. Policy use of production analyses is illustrated by the work of Adler et al. (1977) who have recently completed for the National Science Foundation a summary of research that pertains to television advertising directed to children. The following examples show how production factors relate to such a policy context: — One important issue is the extent to which children can differentiate between program content and commercials. The three commercial networks employ different production devices with the intent of achieving this separation in the minds of the viewing children, but the effectiveness of these devices is unknown. — The manifest accuracy of product attributes suggested by various production devices is an issue, as is the accuracy of product attributes perceived by children. — Disclaimers and caveats may be missed or miscomprehended, and such (mis)perception and (mis)comprehension can be affected by production factors, as can issues of host selling, commercials for foods and drugs, portrayals of violence and unsafe acts, use of premiums in commercials, and repetition of commercials. 243

There is also a theoretical research dimension to analysis of production factors, typified by the work of Salomon, who poses such intriguing questions as this: "If language, a communicational symbol system, can be internalized to become a 'tool of thought', and if media, other than language, have their own symbol systems, could they not be internalized as well? Moreover, could they not come to serve mediational (i.e., covert representational) functions similar to those accomplished by internalized language?" (Salomon, 1977; p. 4). Thus it may be, according to this line of thought, that television's zoom-in feature could stimulate internal mental processes of relating a part to the whole. Work continues in this area, but one need not wait until the theoretical issues are resolved to use the analysis of production factors in message design. The innovative use of production factors in message design is made much more manageable with the use of formative research, which is discussed next.

Formative research* In children's television programming, formative research is, in general, designed to provide diagnostic feedback into the decision-making process for production, so that programming improvements, if needed, can be made before broadcast. The typical contrast is with summative research, which assesses the extent to which goals were achieved by the program or series as a whole, among the intended target audience (Palmer, 1974). The former is analogous to mid-course tutoring; the latter is analogous to a final report card. The type and amount of formative research needed are related to the type and amount of accountability for the programming. Commercial entertainment programming is not held accountable for learning effects. On the other hand, most in-school ITV programming is not held accountable for attracting a voluntary audience. The curriculum-based or goal-directed programming that competes for the attention of the in-home viewer, however, faces dual accountability; it is required to deliver voluntary audiences and to have educational effects. Formative research can be useful in all categories. Particularly in the most difficult combination, the role of formative research * I have come to prefer the term "formative research" to "formative evaluation" in the sense that "evaluation" is sometimes too restrictive. "Evaluation" means judging the worthwhileness of something, i.e., placing value upon it. This does characterize much of the activity of formative researchers, as when they "evaluate" test segments of programming. Formative researchers do more, however, at least at CTW. They help develop the curriculum, and they do background research on the target audience, for example, and these activities are not entirely evaluative. My interest, however, is in the activities, not what they are called.

244

can be critical. The blending of entertainment with carefully planned educational content tends to push production people into uncharted waters, where any reasonable navigational aids that formative research can provide will fall on receptive ears. Some approaches to formative research are based directly on the terminal goals for the project. For example, if an educational film was supposed to teach one how to use a new machine, a formative research team could show it to a test audience, who would then be tested directly on their ability to use the machine. The simplicity of this approach is appealing: If a program is supposed to do X, let formative research see if it does indeed do X. There are real limitations, however, in actually applying this "test the terminal goal" approach to formative research in television, particularly television designed for the voluntary, in-home audience. To distinguish them from the terminal goals, let us refer to prerequisite steps to the terminal goals as instrumental goals. The instrumental goals must be reached before the terminal goals can be reached. Among the first and certainly among the most difficult of the instrumental goals in television designed for the home audience is attracting the voluntary audience in the first place. (The saying goes that "you have to get them into the church before the preaching does any good".) Since programmers cannot control the competitive attractions for the in-home audience, whether the competition is other television programming or something else entirely, the line of formative research suggested by the instrumental goal of attracting the audience is usually addressed to measures of appeal of the test programming. Measures of appeal can take many different operational forms. The distractor method has been used with success at CTW among children too young for responding to questionnaires or manipulation of research apparatuses. Test program material is shown along with some distraction, such as frequently changing slides, or other distractions, such as toys to play with. An observer notes when and for how long the child pays attention to the test program versus the distraction. The instrumental goal issue is not only to attract the attention of the target audience but also to hold it, since in the home setting there is no external compulsion to stay with the program. The distractor method also indexes that. When it can be shown that some parts of the test program attract and sustain attention, while other parts do not, that becomes useful diagnostic information for program revision. None of this essential information on instrumental goals would come from going directly to measurement of terminal goals. Among older children, say, in the 8- to 12-year range, we have successfully used a modified version of the Stanton-Lazarsfeld program analyzer. This is 245

a research apparatus that allows test audience members to register votes (evaluative reactions about program appeal) on a continuous basis, while the test program is being watched. A moving paper scroll on which votes are permanently recorded allows the research staff later to match exact program content to the responses, which provide diagnostic feedback to production. Another method is to photograph or make a video tape of children watching a test show, then analyzing manifest visual attention to the television screen. Among older children, postviewing questionnaires can also be employed for measures of appeal, for the program as a whole, for specific segments with the program, and for various cast members. Yet another method is to play excerpts from a variety of programming (which includes the test program) and then take a vote on which program the children would like to see in its entirety. When the question is what part of the screen is the focus of attention, as was the case for The Electric Company when print would be on the screen, more elaborate research apparatus is needed, such as eye-movement cameras. A variety of softer methods is also employed to test appeal. Foremost among these is simply talking with test audiences about what they liked or didn't like and why. Formative measures of credibility fall somewhere between appeal and comprehension. When message designers are exploiting what television can do uniquely well, they get away from such formats as lectures and into such more-difficult-to-evaluate areas as role modeling, extension and compression of sensory data, setting up vicarious experiences that could never be experienced first hand, and so on. Consider only role modeling. It cannot be simply assumed that an intended role model will actually function as such. The attractiveness and credibility of the intended model need to be ascertained through formative research. When something doesn't "ring true," such as an amateurish performance or production, or anything suspected of being "phony," children are quick and unforgiving in their evaluative reactions. Another important instrumental goal is comprehension. This refers to such obvious factors as vocabulary level but, in television, comprehension covers much more than that. If an informational point is being developed, say, in three linear, cumulative segments, and if the final segment is not comprehended, the question is raised as to whether the test audiences understood the lead-in steps or whether the problem lies solely in the final segment. Therefore, assessing the cognitive prerequisites to later understanding within the development of a message is important. Formative research is also used to see if the audience has the necessary background to comprehend the material in the first place. Also, to the extent that a full range of production techniques is employed with pedagogical rationale to convey 246

the intended message, the issue of comprehension extends to comprehension of the production techniques themselves. Whether conscious of it or not, adults are familiar with many production conventions. If we see a person boarding a plane and then see a cut to the plane in the air, we naturally assume that the person we saw boarding a moment ago is now aboard the plane in the air. This may or may not be the case with young children; it needs to be ascertained with formative research. There are several methods for testing comprehension. Since it is very difficult to recall after the fact what was being thought about at various times during a test program, there are certain advantages to obtaining comprehension measures as the program is being presented. One way to do this is to stop the tape at critical junctures and ask the child to explain what is going on at that point in the program. A variation on this is to first view the program in its entirety, then replay the program with the sound turned off, asking the child to provide play-by-play narration. Again, with older children, postviewing interviews and questionnaires can be employed. Therefore, one major reason for not testing terminal goals directly in formative research is that the instrumental goals of attention (appeal) and comprehension cannot be merely assumed. Another reason is that it is frequently unrealistic to expect terminal-goal behavior as the result of a single exposure. Also, it is frequently valuable to test parts of programs before the entire program is assembled. Unlike draft versions of books which can be revised easily, revision of completed television programs is usually prohibitively expensive. The bulk of formative research in children's television, therefore, is devoted to measures of appeal and comprehension. There is no uniform set of criteria by which to evaluate research that is called formative because the label in practice has been used to refer to different functions. Research called formative that is used in-house is different in important ways from research called formative that is used out-of-house. (The previous discussion of formative research has been of the in-house variety.) When used for the purpose of informing the ongoing, in-house decision needs of the production staff, the formative research itself takes on a definite inhouse character. In this context, it is most natural for the research staff to be in-house personnel. It requires an insider's access to the real production problems, the real production decisions, and the real set of feasible options in order to produce research that is useful in actual, in-house decisionmaking. In the in-house setting, formative researchers are not disinterested third parties who pass judgment on processes and products; they are rightfully 247

expected to be as committed to the success of the project as are other members of the team. Their in-house role as formative researchers is not only to spot problems but to go on to suggest improvements and solutions. For a variety of reasons, such formative research tends to stay in-house rather than go into public dissemination; formative research reports are not usually published in scholarly journals. Being working documents, timed by a complex flow of internal decision-making, they tend to vary in completeness and frequently lack the context and background necessary for an outside reader. Being oriented to spotting problems, generating insights, and suggesting solutions, they often fail to meet criteria for design and rigor associated with conclusion-oriented research which is done to test hypotheses and build theory. Conclusion-oriented research seeks to establish generalizable propositions; decision-oriented research tends to be productspecific (i.e., descriptive of one particular stimulus and/or its effects). Product-specific formative data tend to be of interest and value only to the in-house staff. Such data as scene-by-scene attention profiles obviously have greatest relevance to those primarily involved in the design and production of those particular scenes. Even other formative researchers tend not to be interested in their colleagues' product-specific data; what they are more interested in are the formative research techniques and procedures, because these have more general applicability. The major criterion by which in-house formative research is to be judged, therefore, is its actual utility in reaching informed decisions in the design and production of the television materials. This utility is indeed affected by methodological factors; issues of internal and external validity are relevant to all data-based research. Perhaps even more important to actual utility, however, are such factors as addressing research questions that feed real production decisions, feeding data and interpretation back in time to actually affect the decision, putting results in a nonjargonized form that is easily comprehended by nonresearchers, and skill in going beyond the data to extract message design implications in discussion with the production staff. The utility of formative research is thus heavily dependent on good interpersonal relationships between researchers and producers. This cooperative linkage is not necessarily a natural one or one that can be legislated into existence; it must be nurtured carefully both by research and production people. External (out-of-house) decision-makers have information needs related to the project, but they tend to be for summative-type decisions, such as continuation of funding or distribution and utilization decisions. The external sources, presumably, are not involved directly in ongoing production 248

decisions, which are the primary target for in-house formative research. Therefore, when access to formative research is required by external sources, such as funders, and the research is used to make a summative-type decision, the research, by way of its functional use, is then summative research, regardless of what it is called. That is to say, the ultimate designation of a datum as formative or summative depends in large part on how it is used. A datum used internally to modify and improve the production of materials is formative; a datum used externally to reach decisions based on demonstrated or predicted degree of success in achieving goals is summative. Summative data can be used formatively, as when an ongoing series incorporates into its revision plans the summative outcomes from a previous season. This causes no special problems. The reverse however, using (internal) formative data for (external) summative-type decisions, can be problematic. The pressure on formative researchers to provide both formative and summative input places researchers in a difficult situation, raising the probabilities that the research will serve neither function well. Crane and others (1977) have recently studied this issue. These researchers conducted confidential staff interviews at eight projects, all of which produced regional or national television programs for children. In each of the eight projects, one evaluator, one producer, and one administrator were interviewed to obtain three distinct perspectives on the impact of evaluation on the production of the children's programming. Pertaining to this issue, the authors concluded: "Perhaps the most difficult problem encountered by evaluators arose in situations where both formative and summative evaluation activities were demanded of the evaluator. It is difficult to cultivate a relationship of cooperation and trust necessary for formative evaluation when the summative evaluation report can potentially destroy production" (p. 27). In a policy research study for the U. S. Office of Education, Mielke et al. (1975) also addressed this issue of confusing or mixing the formative and summative research functions: "Evaluation is important in program development and production, where pretesting can provide corrective feedback to the production staff before the product is finalized. This function of formative evaluation is most appropriately conducted by in-house evaluation staff communicating directly with in-house producers. Summative evaluation compares goals with achievement and is used by out-of-house decision-makers for decisions such as refunding, distribution, and adoption. Dysfunctional pressures are brought to bear on the formative evaluation process when in-house formative data are utilized by out-of-house decision-makers for summative-type decisions, such as in the review of pilot-testing data for go/no go decisions on an entire series. 249

We recommend that summative evaluation procedures be utilized for summative types of decisions. The summative procedures we recommend include employment of external evaluation agencies that work in cooperation with the production agency, subject to third-party review" (p. xi). In my opinion, the formative research role is too important in goal-directed children's programing to be pulled off course, away from its major objective of helping producers make better programs. If half-heartedly pursued merely to satisfy a funding requirement, formative research can easily become ritualized and meaningless. If forced into functions for which it was not intended, such as summative research or even public relations, the contamination of its original function is a likely outcome. In considering various forms of decision-oriented research for in-school television programming a few years ago, I recommended: "The top priority for research resources should go to formative research. There should always be a summative research component of some dimensions, but without an active formative research component, isolated summative research may be little more than an autopsy" (Mielke, 1974). The case for a high priority on formative research, that truly functions as formative research, is equally valid for all goal-directed children's programming. When clearly locked into teamwork with production, formative research objectives and procedures evolve as the project itself evolves. Long before actual production has started, formative researchers can be contributing to the curriculum and developing goals, possible measures for those goals, and even general kinds of production treatments that have high promise of addressing those goals programmatically. Background research on the target audience can be usefully incorporated at these early stages of a project, for example: — What is a typical information level on the program topic among the target audience? — What facilitative or hindering attitudes toward the proposed content are likely to be encountered? — What research techniques and procedures are appropriate for this target audience and this topic? — According to the literature and expert advisors, what are the predominant needs in the proposed content area? (Which of these needs are more and less amenable to television treatment?)

250

Existing materials on related topics, produced by others, can be evaluated among the target audience to find programming attributes that are appealing and comprehensible. Then, in early stages of production, scripts and storyboards can be evaluated, feeding forward into actual production. Segments of unfinished programs can be evaluated, feeding forward into final editing. Pilot or test shows can be evaluated with greater care and rigor, feeding forward into final series decisions. Early series programs can be evaluated, feeding forward into production or editing decisions for later programs in the series. It can be seen that formative research is not a static enterprise but one that conforms to the changing needs for input into important managerial decisions. It is difficult to extract actual production guidance or message-design principles f r o m much of the basic research on the effects of television, but it might be a worthwhile effort to review that literature f r o m the viewpoint of message-design implications. Some work in progress at the Indiana University Institute for Communication Research, for example, holds promise in this regard. The research, funded by the National Science Foundation and being conducted by Dr. Dolf Zillmann and colleagues, deals experimentally with various effects of humor in educational television programming. Drawing guidance f r o m all possible sources, the conscious, pedagogical use of the full range of production variables, harnessed to constructive ends and ascertained with formative research, would clearly raise the probabilities for significant impact of television programming in any educational area.

References Adler, R. P., Friedlander, Β. Ζ., Lesser, G. S., Meringoff, L., Robertson, T. S., Rossiter, J. R., and Ward, S. Research on the effects of television advertising on children. Washington, D. C.: United States Government Printing Office, 1977. Bernstein, L. J. Design attributes of Sesame Street and the visual attention of preschool children. Unpublished doctoral dissertation, Columbia University, 1978. Chu, G. C., and Schramm, W. Learning from television: What the research says. Washington, D. C.: National Association of Educational Broadcasters, 1968. Crane, V., Quiroga, B., and Linowes, J. The impact of evaluation on the production of children's television programs. Paper presented at the annual convention of the 251

National Association of Educational Broadcasters. Washington, D. C., November, 1977. Kincaid, H. V., McEachron, N. B., and McKinney, D. Technology in public elementary and secondary education: A policy analysis perspective. Menlo Park, Cal.: Stanford Research Institute, April 1974. Lazarus, M. Education and television: Making the marriage work. The National Elementary Principal, 1977, 56 (3), 32. Lesser, G. S. Assumptions behind the production and writing methods in Sesame Street. In W. Schramm (Ed.), Quality in instructional television. Honolulu: The University Press of Hawaii, 1972. Mielke, K. W. Asking the right ETV research questions. Educational Broadcasting Review, 1968, 2 (6), 54-61. Mielke, K. W. Media-message interactions in TV. In G. Salomon and R. E. Snow (Eds.), Commentaries on research in instructional media. Viewpoints, 1970, 46, 15-33. Mielke, K. W. Decision-oriented research in school television. Public Telecommunications Review, 1974, 2 (3), 31-39. Mielke, K. W., Johnson, R. C., and Cole, B. G. The federal role in funding children's television programming. Volumes I and II. Bloomington: Institute for Communciation Research, Department of Telecommunciations, Indiana University, 1975. Palmer, E. L. Formative research in the production of television for children. In D. E. Olson (Ed.), Media and symbols: The forms of expression, communication and education. Chicago: University of Chicago Press, 1974. Salomon, G. The language of media and the cultivation of mental skills. Jerusalem: The Hebrew University, 1977. Schramm, W. The second harvest of two research-producing events: The Surgeon General's inquiry and Sesame Street. Proceedings of the National Academy of Education, 1976, 3, 151-219. Stanford, M. C. A review of recent research on instructional television and radio. In J. P. Witherspoon (Ed.), State of the Art: A study of current practices and trends in educational uses of public radio and television. Washington: Advisory Council of National Organizations and the Corporation for Public Broadcasting, 1974.

252

Edward L. Palmer

Formative research in the production of television for children * The beginnings The creation of Children's Television Workshop (CTW) in 1968 launched a new era in the use of broadcast television to help meet urgent national educational priorities. Innovation in effective research-producer collaboration has been singled out consistently by CTW's president and founder, Joan Ganz Cooney, as both key and the unique element in the success of the Workshop's productions. CTW was formed as an independent producing organization with initial grants of $ 8 million, half contributed by Federal agencies and half by foundations. Out of these grants came the Sesame Street series for preschoolers (Cooney, 1967), now in its fifteenth continuous broadcast season. This series, and its companion, The Electric Company reading series for children, also created by CTW, became cornerstones in the children's afterschool program block on U.S. public television, and gave rise to what is indisputably the world's leading model for systematic educational television planning and research. Moreover, virtually throughout today's emerging strong academic tradition of television with children, these series are credited as a key source of researcher inspiration. On the applied side of the movement, direct production outgrowths include an $ 80 million program of Federally sponsored children's educational television series, made primarily for home viewing, and supported by direct Congressional mandate through the Emergency School Assistance Act; contribution of funding support for 15 seasons of Sesame Street and 6 of The Electric Company; funding for 2 seasons of a major children's science series titled 3-2-1 Contact; and creation of Sesame Street adaptations in 11 other languages and 3-2-1 Contact adaptations in 3. In addition, variations on the model have been applied by CTW in creating the 24-part Feeling Good television series on health for adults (Mielke and Swinehart, 1976) and 50 oneminute television messages on family health for distribution in Latin America. * Author's note: This chapter is based on a merging and slight updating of two previous book chapters on the same subject (Palmer, 1973; 1974).

253

An important feature of this movement is its emphasis on "formative" planning and resèarch. First, important objectives are clearly identified, then systematic audience tests are carried out in order to evaluate progress toward their achievement during the actual course of a program's production. Formative research is typically contrasted with summative research, which is concerned with follow-up testing to determine the educational effect of new products and practices when actually put into use (Scriven, 1967). What follows is a description of the approaches to formative planning and research taken by the Children's Television Workshop in the production of Sesame Street and The Electric Company. The focus is entirely on the early period. Later and continuing developments in formative research are beyond the scope of this chapter, as is a review of the extensive recent literature on children's attention, comprehension and learning with respect to television. Although an effort is made throughout to discuss our approaches to formative research in ways that will suggest their potential usefulness in the development of other new educational products and practices, this paper is not a compendium of research for the general guidance of producers of educational materials; it is, rather, a case study of the use of a powerful new technology for the achievement of planned educational effects. As a case study, it focuses on the overall operational framework within which CTW's formative research proceeds, on the strategies and rationale for the design of formative field research methods, on organizational and interpersonal conditions, and on similarities and contrasts between the functions and the methods of formative research on the one hand and those of more traditional research approaches on the other. At the beginning of the Sesame Street project, the functions formative research could serve and the field methods it could apply were not at all clear. There were no precedents of sufficient scope and generality from either the field of educational technology in particular or the field of educational planning and research in general to provide clear guidelines. What has been learned about the formative planning and research process at CTW has come about under quite unusual circumstances, and since it is unlikely that these conditions will ever be duplicated in substantial detail, it remains to be seen what sorts of new or modified approaches will be required in different situations. What is presented here certainly cannot be construed as a dependable recipe that will assure the success of other like ventures. Among the unusual circumstances associated with the Workshop's productions, some, no doubt, had quite a direct bearing on the effectiveness of the formative research. For instance, the two Workshop projects were well funded; each was budgeted in its first season at upwards of seven million 254

dollars for production, research, and related activities. This level of support made it possible to utilize high-level production talent and resources and to make extensive use of expert educational advisers and consultants. In addition, both projects enjoyed unusually long periods of time — in each case, approximately eighteen months — for prebroadcast planning and research. Time and resources were available to plan their curricula carefully and to state their educational objectives in very explicit terms. This meant that producers and researchers alike, as well as the independent evaluators who were carrying out pre-season and post-season achievement testing projects, could proceed without ambiguity of purpose and in a coordinated fashion. Had there been ambiguity in terms of either the particular objectives to be addressed or the commitment of the producers to direct each segment toward the achievement of one or more of those objectives, the formative research could not have been useful, for there would have been no clear criteria for evaluating a program segment's effectiveness. Also unusual in the CTW case was the policy followed in production recruiting and the organizational and interpersonal relationships between the in-house research and production staffs. All of the key producers came from commercial production backgrounds. None had formal professional training in education or experience in educational television production. Yet they were given the responsibility for final production decisions. They did not work under the researchers, nor did the researchers work under them. The intended function of the formative research was to provide information which the producers would find useful in making program-design decisions relating to both appeal and educational effect. To the extent that the formative research worked, it worked in large measure because of the attitudes taken toward it by producers and researchers alike. The producers were committed to experimenting with the cyclic process of empirical evaluation and production revision and tended to have the ability not only to see the implications of the research, but also to carry these implications through into the form of new and revised production approaches. Accordingly, the usefulness of CTW's formative research has depended not only on the qualities of the research itself, but also on the talents of those who put its results to use. Moreover, the producers never expected the research to yield full-blown decisions; they recognized that its function was to provide one more source of information among many. From the research side, because the responsibility for final production decisions resided with the producers, it was necessary to develop and apply only methods which provided information useful to the producers. Accordingly, the producers were involved from the 255

outset in all research planning. No observational method was ever persistently applied, and no specific study was ever taken into the field without their participation. At this time, there was no tradition of accumulated knowledge in the area of formative research practice. This was partly because so little research of this type had been done, but it was even more a result of the fact that it had only recently come to be recognized as a distinct and systematic field of endeavor. A limited number of articles had contributed to the early general conceptualization of the field (see Cronbach, 1963; Hastings, 1966; Scriven, 1967). In addition, formative research studies associated with specific product improvements were reported for various media, content areas, and student levels: for example, for televised instruction (Gropper and Lumsdaine, 1961); for programmed instruction {Dick, 1968); and for kindergarten instruction in conceptual skills (Scott, 1970a) and in art (Scott, 1970b). However, the scope and depth of the formative research literature at this time was in no way commensurate with its promise for education. The promise of the approach is that it will provide designers of educational products and practices with empirical data far more directly pertinent to their respective media, materials, and learning conditions than are the results of traditional, more basic, research. For the field of educational television in particular, it offers ways to help bring about planned effects. With mass broadcast distribution making it possible to reach hundreds of thousands or even millions of viewers (in the case of Sesame Street, the weekly audience is estimated to include more than eight million children), it behooves the producers to employ every reasonable means for ensuring in advance that the programs will achieve their objectives. As experience with formative research procedure begins to accumulate and the conditions for success or failure become more clear, many of the factors considered significant in the work carried out at CTW must almost certainly receive prominent continuing attention. Those considered most significant here are the subjects of the sections that follow. In brief overview, these include the overall operational framework within which CTW's formative research proceeds, the strategies and rationale for the design of formative field research methods, organizational and interpersonal conditions, and similarities and contrasts between the functions and the methods of formative research on the one hand and those of more traditional research approaches on the other.

256

The CTW operational model The principal activities undertaken in the production of Sesame Street have come to be viewed by CTW as a model, and this model was again applied in the production of The Electric Company. If there is a single, most critical condition for rendering such a model of researcher-producer cooperation effective, it is that the researchers and the producers cannot be marching to different drummers. The model is essentially a model for production planning. More specifically, it is a model for planning the educational (as opposed to the dramatic) aspects of the production — and formative research is an integral part of that process. In the case of Sesame Street and The Electric Company, at least, it is hard to imagine that the formative research and curriculum planning could have been effective if they had been carried out apart from overall production planning, either as a priori processes or as independent but simultaneous functions. The activities included in the model are presented below in their approximate chronological order of occurrence. Behavioral

goals

As the initial step toward establishing its educational goals, CTW, in the summer of 1968, conducted a series of five three-day seminars dealing with the following topics: 1. 2. 3. 4. 5.

Social, moral, and affective development Language and reading Mathematical and numerical skills Reasoning and problem-solving Perception

The seminars, organized and directed by Gerald S. Lesser, Bigelow Professor of Education and Developmental Psychology at Harvard University, were attended by more than a hundred expert advisers, including psychologists, psychiatrists, teachers, sociologists, film-makers, television producers, writers of children's books, and creative advertising personnel. Each seminar group was asked to suggest educational goals for the prospective series and to discuss ways of realizing the goals on television. The deliberations of the seminar participants and the recommendations of the CTW Board of Advisors were reviewed in a series of staff meetings from which a list of instructional goals for the program emerged. These goals were grouped under the following major headings:

257

I.

Symbolic representation A. Letters B. Numbers C. Geometric forms II. Cognitive organization A. Perceptual discrimination and organization B. Relational concepts C. Classification III. Reasoning and problem solving A. Problem sensitivity and attitudes toward inquiry B. Inferences and causality C. Generating and evaluating explanations and solutions IV. The child and his world A. Self B. Social units C. Social interactions D. The man-made environment E. The natural environment Specific goals under each of these broad headings were stated, insofar as possible, in behavioral terms, so that they might serve as a common reference for the program producers and the designers of the achievement tests. Appropriate coordination of production and evaluation thus was assured. Existing competence

of target

audience

While the statement of goals specified the behavioral outcomes which the program hoped to achieve, it was necessary to ascertain the target audience's existing range of competence in the chosen goal areas. In its initial formative research effort, the Workshop research staff therefore compiled data provided in the literature and did some further testing of its own to determine the competence range. The resulting information helped guide the producers in allocating program time and budget among the goal categories and in selecting specific learning instances in each goal area. Appeal of existing

materials

To be successful, CTW had to capture its intended audience with an educational show whose highly attractive competition was only a flick of the dial away. Unlike the classroom teacher, the Workshop had to win its audience, and it had to hold their attention from moment to moment and from day to day. At stake was a potential variation in daily attendance which could run into millions. Measuring the preference of the target audience for 258

existing television and film materials was therefore crucial in the design of the new series. Experimental

production

Seminar participants and CTW advisers had urged the use of a variety of production styles to achieve the curriculum goals adopted. Research had confirmed the appetite of the target audience for fast pace and variety. Accordingly, the CTW production staff invited a number of live-action and animation film production companies to submit ideas. The first season of Sesame Street eventually included the work of thirty-two different film companies. Prototype units of all film series produced by or for the Workshop were subjected to rigorous preliminary scrutiny and empirical field evaluation. Scripts and storyboards were revised by the Workshop producers on the basis of recommendations from the research staff, further revisions were made after they had been reviewed by educational consultants and advisers, and finished films were tested by the research department with sample audiences. Some material never survived the process. Four pilot episodes were produced for a live-action film adventure series entitled The Man from Alphabet, but when the films were shown to children they failed to measure up either in appeal or educational effect, and the series was dropped. Sample video-taped material went through the same process of evaluation, revision, and occasional elimination. By July of 1969 a format for the program had been devised, a title had been selected, a cast had been tentatively assigned, and a week of full-length trial programs had been taped. Completed prototype production elements were tested by the research staff in two ways: (a) the appeal of the CTW material was measured against the appeal of previously tested films and television shows, and (b) the CTW material was tested for its educational impact under a number of conditions. For instance, field studies were conducted to determine the effect of various schedules of repetition and spacing, of providing the child with preliminary or follow-up explanation, of presenting different approaches to a given goal separately or in combination, and the relative effectiveness of adult versus child voice-over narration. Extensive observation of children watching the shows provided information about the child's understanding of various conventions of film and television techniques. When each research study concluded, the results were reported to the producers for their use in modifying the show components tested and for their guidance in producing subsequent elements. It should be noted that this progress-testing also served 259

a formative research function for the Educational Testing Service staff by field-testing the instruments and administration procedures that were to be employed in later summative evaluation. The progress

testing

The evolution of Sesame Street did not end with the first national broadcast on November 10, 1969. Formative research studies conducted throughout the six-month broadcast period continued to guide the development of new production techniques, format elements, and teaching strategies. As before, these studies had two foci: (1) the holding power of entertainment techniques, and (2) the effectiveness of educational content. Earlier and continuing studies of individual program segments, while useful, were necessarily limited in scope. With the onset of the broadcast season, it was possible to examine the impact of continuous viewing of entire shows over a period of time. Accordingly, the research staff instituted a program of testing the show's effectiveness, using the summative evaluation instruments designed by the Educational Testing Service (ETS) of Princeton, New Jersey. A sample of day-care children, predominantly four- and five-year-olds, were pretested prior to the first national telecast. One-third were tested again after three weeks of viewing the show, the same one-third and an additional onethird were tested after six weeks of viewing, and the entire group was tested after three months of viewing. Comparisons between experimental (viewing) and control (nonviewing) groups at each stage of the testing gave indications of strengths and weaknesses in both the execution of the curriculum and the production design. Other independent formative studies of program appeal and of the responses of viewing children also influenced production decisions during this period. Summative

evaluation

The summative research and evaluation carried out by Educational Testing Service (see Ball and Bogatz, 1970a,b; Bogatz and Ball, 1971a,b) followed a plan developed in consultation with CTW staff and advisors. Participation of ETS representatives in all main phases of prebroadcast planning helped to ensure coordination between program development and follow-up testing. ETS developed and administered a special battery of 11 tests covering the major CTW goal areas to a sample of children from Boston, Philadelphia, Durham, and Phoenix. The groups included three-, four-, and five-year-olds in urban and rural settings, from middle- and lower-income families, in both home and day-care situations. A special side study related to children from Spanish-speaking homes. The 11 tests were as follows: 260

7. Classification Test. 1. Body Parts Test. 8. Puzzles Test. 2. Letters Test. 9. What Comes First Test. 3. Numbers Test. 10. Embedded Figures Test. 4. Shapes and Forms Test. 11. Sesame Street Test. 5. Relational Terms Test. 6. Sorting Test. Other measures assessed home conditions, parental expectations for the children, and the like. In instances where the results of the first season's summative research were fed into production decisions for the second season, they took on a formative function. For example, the summative data indicated that the children's knowledge and skills before viewing the programs had been underestimated in some goal areas and overestimated in others. This was taken into account in programming the second season of Sesame Street. The Writer's

Notebook

As the producers and writers began to develop scripts, animations, and liveaction films addressed to particular behaviorally stated goals, it became apparent that the goal statement was not a wholly adequate reference. After having been given several successive assignments in the same goal area, they began to express the need for extended and enriched definitions which would provide creative stimulation. Gradually, through trial and error, a format for the Writer's Notebook was developed which the producers and writers found useful. The Notebook emphasized four criteria: 1. To focus on the psychological processes involved in a particular form of behavior; 2. to exploit and extend the child's own experiential referents for such behavior; 3. to prompt the creation of various similar approaches by the producers and writers themselves by presenting them with highly divergent examples; 4. to provide suggestions free of any reference to particular characters or contexts from the television program, so that the ways in which the suggestions could be implemented would be left as open and flexible as possible. These features of the Writer's Notebook may be highlighted through an example. In the broad area of "symbolic representation", the word-matching objective was stated as follows: "Given a printed word, the child can select an identical printed word from a set of printed words." To implement this objective, the Notebook encouraged the producers to use words with different numbers of letters, to vary the location within the word of the letter or letters 261

which fail to "match," and to present various matching strategies — such as comparing two given words letter by letter, moving words which were initially separated into physical superimposition, and spelling out each of two given words and comparing to see if one has made the same sounds both times. Another recommended approach was to make use of the "sorting" format, already familiar to viewers, wherein three identical things (in this case, words) and one odd thing are presented simultaneously along with a standard song which invites the viewer to find the one which is different. Still another was to construct a letter-by-letter match for a given word by choosing from a large pool of letters. To encourage still other approaches, another recommendation was to present pairs of words which matched in one sense but not in another — for example, pairs in which the same word is presented in different type faces, or in which one of the pair is the upper- and the other lower-case version. The producers and writers asked that similar suggestions be developed for other goal areas. Again, suggestions were solicited from advisers and consultants. In addition, the Notebook provided a place and a format for collecting the ideas of the in-house research staff and helped to ensure that these ideas would be seen and used.

A model for research on presentational learning Learning by way of televised presentations does occur, and the objective of the formative research at CTW is to discover principles of program design by which this type of learning can be improved. In the specific case of Sesame Street and The Electric Company, the research seeks principles of presentational learning appropriate to their particular educational goals, audiences, and production techniques. However, there are more basic objectives also, one of which is to create generalizable formative research methods and practices. Another is to discover generalizable principles of presentational learning. In pursuit of these objectives, a model for research on presentational learning is being developed. A summary outline of the model is given in Figure 1. One point highlighted by the model is the need to identify and rigorously define features of program design (independent variables) which are reliable predictors of learning and learning-related outcomes among viewers (dependent variables). The statements which relate program-design features to empirical outcomes are principles of presentational learning, which are potentially generalizable to other televised presentations and to other media. 262

Figure 1: A model for research on presentational learning. Major attribute categories are very general dimensions of the televised presentation. Viewer outcomes represent the effects of the presentation on viewers. Within each attribute category, any number of specific program features (variables) may operate to affect viewer outcomes. Principles of program design are hypothesized or well-validated relationships between program features and specific viewer outcomes.

MAJOR PROGRAM ATTRIBUTES (Categories of independent variables)

VIEWER OUTCOMES (Dependent variables)

Appeal

Visual orientation, attention, attitude, channel selection, etc.

Comprehensibility

Comprehension

PRINCIPLES OF PROGRAM DESIGN (Statements linking specific independent and dependent variables)

Activity-Eliciting Potential a) Potential to elicit motor and psychomotor activity b) Potential to elicit emotive activity c) Potential to elicit intellectual activity

Internal Compatibility of Elements

Verbalization, gross physical acts, imitation, directionfollowing, etc. Arousal, attitude, etc.

Within each major attribute category, any number of specific features may operate to affect viewer outcomes. Statements linking specific and well-defined program Synthesizing or integrating, forming concepts or principles, features to learning outcomes generalizing, comparing, eval- belong in this column. uating, predicting, etc. Attention to signal vs. noise, integration of elements (visualvisual, auditory-visual, auditory-auditory), etc.

Moreover, they are seen as principles which will have to be taken into account within any comprehensive theory of presentational learning. The model is tentative, intended more as a point of departure for further exploration than as a fixed scheme. However, as it stands, it plays a central role in guiding CTW's formative research studies and in organizing their results. It features four main categories of program attributes, including Appeal, Comprehensibility, Activity-Eliciting Potential, and Internal Compatibility of Elements. Activity-Eliciting Potential presently has three main subcategories: (a) motor and psychomotor activities, including imitation, (b) emotive activity, and (c) intellectual activity. Each subcategory in turn encompasses a number of 263

program attributes. The subcategory of intellectual activity, for example, includes those program attributes responsible for such viewer outcomes as integrating (putting together the audio and visual portions of the presentation, or the successive elements of an unfolding plot, etc.), generalizing (relating the presentation to past experience or to future possibilities), anticipating (predicting possible upcoming events in the presentational sequence), to forming new concepts and principles, imputing the motives and intentions of characters or the program's producers, and evaluating - as in assessing the credibility of the plot, message, or character premises, judging the quality of the performance or the technical quality of the production, and the like. The main attribute categories contained in the model and some of the research methods employed within each are discussed more extensively later. But first there needs to be a clarification of the intended role of the model in the formative research process. The development of this model is not so much an outgrowth of the formative research process as an integral part of it. It was developed in its present form with a number of specific objectives in mind. First, it is intended by virtue of its highly simplified form to serve as a convenient checklist for both producers and researchers, suggesting program attributes which they need to take into account in creating new segments or designing new formative field studies. It is also intended to serve a number of useful organizing functions. For example, the various field research methods are organized according to the categories of attributes to which their results relate. This is important in that the various methods within a given attribute category tend to elucidate complementary sets of program-design features and viewer outcomes, and their organization within broad attribute categories tends to highlight this complementarity. Still another intended function of the model is to bring together within a small and therefore convenient number of attribute categories the great number of hypothesized program-design principles growing out of the formative research. Finally, the model is open-ended in that it necessarily falls short of presenting an exhaustive list of program attributes and audience outcomes. This openendedness gives rise to another of the model's potentially valuable qualities: it invites researchers and producers to identify other potentially significant program attributes, audience outcomes, research methods, and principles of program design. To cite an example of this effect, former CTW researcher Joyce Weil, while participating in research planning for a new adult television series on health, extended the categories of the model to encompass two variables considered important in developing such a series, namely, 264

credibility, on the independent variable side, which relates to belief, and memorability, which relates to long-term recall and to factual distortion. The principles of presentational learning which would appropriately appear in the third column of Figure 1 were characterized earlier as statements which relate features of the presentation to empirical outcomes. One important consideration in establishing these principles is the need to define program variables with enough precision that they yield high interrater reliability, i.e., two or more raters ranking the same program segments according to a given definition will produce the same or nearly the same rankings. Another consideration is that the principles of presentational learning based on these definitions need to possess predictive validity, i.e., the rankings assigned to segments in terms of a given program variable should yield better than chance predictions of their effectiveness. CTW is exploring a number of procedures for identifying program-design variables and linking them empirically to their effects on viewers. A procedure developed by Rust has been applied to data on program appeal (see Rust, 1972). In this procedure, empirically derived scores on a program's appeal are determined for each 7.5-second interval throughout its presentation. The scores for the various intervals are converted to unit normal (z-score) form, then lists are made so that highest and lowest scoring segments are displayed separately. These are then scanned by researchers and producers in an attempt to define the program-design variables which appear to differentiate the high from the low. The definitions of variables so derived are then tested for interrater reliability and improved if necessary. Finally, the definitions are applied a priori to new segments and evaluated for their power in predicting the appeal measured for those segments. The procedure may be applied with other program attributes as well. Another procedure has been explored for CTW by Gavriel Salomon of the Hebrew University in Jerusalem (Salomon, 1974). In Salomon's procedure, each segment was scored according to a number of different program-design variables which may contribute to the measured outcomes. As with Rust's procedure, it was important that the definitions of these program-design variables possess high interrater reliability. Through the application of multiple and partial regression analyses, Salomon then could evaluate both the individual and interactive contributions of the predictor variables (the program-design variables) toward producing the measured outcomes.

Formative research on program appeal Appeal research bears on a wide range of program decisions. It reveals the effects of various forms and applications of music, e.g., not surprisingly, that 265

segments with music tend generally to be high in appeal, that a sudden transition into loud, lively music will usually recapture attention when it has strayed, and that music can help sustain the appeal of program segments designed to be repeated many times, such as those designed to give repeated drill and practice. It can be useful in the auditioning process, not only for helping to identify most liked characters, but also most liked activities in which characters can engage, such as one person's guiding another through a difficult task in a supportive versus demeaning manner; in the resolution of interpersonal conflict through the arbitrary use of power versus through cooperation; and in the portrayal of an individual's personal struggle to achieve a goal or to improve on his own past performance versus receiving al lecture, to mention a few. It also indicates the relative appeal, in general, of films, animations, and live performances, as well as the most and least popular types of each. Animation, of course, is nearly always popular, but so are other forms which contain some of its features, such as fantasy, ludicrous events, noncluttered backgrounds, an economy of words and actions in playing out the plot, "banana-peel" humor, and a high ratio of visual to verbal communication. Appeal research also helps to indicate how long attention can be maintained under the various conditions, the optimum amount of variety, the optimum pacing of events (with the Sesame Street audience it was exceedingly rare for attention to remain high for more than three or four minutes without a change in format), and the ability of a segment to bear up under exact repetition (in this respect, segments vary widely according to a number of different factors, but they usually hold up if they are high in technical quality from a production standpoint, or if they contain music and jingles, and fall down if the interest resides exclusively in the punch line, if the segment is too long, if it is moralistic, or has too much verbiage and too little visual action). In addition, research on appeal can show the effectiveness of special technical effects, such as fast and slow motion and unusual camera angles. One such effect, pixilation, which deletes selected frames of a film and thus portrays the antics of characters in a rapid, jerky fashion, appears, surprisingly, to be both as high in appeal and as reliably appealing to young children as animation. Monologues and dialogues rarely sustain attention when presented in the form of talking heads. However, voice-over narration by unseen commentators describing or asking questions about a visually presented event are frequently effective. This tends to hold true particularly when the verbalizations are sparse and pointed rather than sustained. There is also a clear and general preference for incongruity, surprisingness, and fantasy, as compared with straightforwardness, predictability, and 266

realism; for episodic versus linear styles of continuity; and for familiar (but not necessarily realistic) versus unfamiliar conventions and symbols dealing with time, sequence, and interpersonal relationships. Finally, this type of research can be used to investigate characteristic individual or group preferences vis-à-vis such program design features. Since both Sesame Street and The Electric Company were designed according to a magazine format, with successive brief segments addressed to very explicit educational objectives, it was important to maintain high program appeal on a moment-to-moment basis. Accordingly, a method (referred to as the Palmer distractor method) was introduced which yields appeal data throughout the course of a program. The Palmer distractor method consists of placing one child at a time in a simulated home-viewing circumstance. A black-and-white video-taped recording of a television program is presented and at the same time a color-slide show is flashed at 7.5-second intervals on a rear-projection screen equipped with an eighty-slide carrousel. The rearprojection screen, which is approximately the same size and height from the floor as the television screen, is placed at about a forty-five degree angle from the child's line of vision to the television screen. The child is seated in a chair three to four feet away from and facing the television but is free to move about within the confines of the room at any time. A continuous record is kept noting when the child's eyes are directed toward or away from the set. For each viewer, the eighty-slide carrousel is started at a different slide so that the stimulus competing with a given 7.5-second interval of televised presentation is different for each viewer. Composite graphs of the results help the researchers and producers identify the elements of program content responsible for high and low appeal. A frequently used complementary form of appeal testing consists of taking observations on successive groups of viewers, where each group typically contains from three to five viewers. Usually, four to six such groups are observed in testing a program. A detailed record is kept according to predefined categories of visual, verbal, and motor behaviors. The visual behavior of children in viewing groups provides a cross-check on the distractor results. The record of verbal and motor responses, in addition to reflecting upon program appeal, helps to identify the program approaches which are most and least effective in eliciting active participation. The fruitfulness of this particular approach is very much a function of the training and the creative interpretive skills of the researcher. Audience surveys can provide much additional material on program appeal, as can structured interviews. Both enable researchers to determine the salient and lasting as opposed to the immediate appeal of various program features. 267

The lasting appeal of an element is particularly important where it is necessary to attract a voluntary audience day after day. Formative research on program

comprehensibility

Once attention is assured it becomes critical to assess the information extracted. What did the children grasp of the intended instructional points? Can they interpret the motives or intentions of the characters? Comprehensibility testing, while useful in evaluating a viewer's understanding of the dramatic action, is undertaken primarily to point up program-design features important in producing instructional effects. It gives producers an empirical check on their assumptions about the comprehensibility of the program-design features they are employing, and even limited amounts of field research can help them to maintain a generalized sensitivity to his important attribute. The ultimate objective of CTW's research in the area of comprehensibility is to identify and set down specific program-design principles. These principles include the use of production approaches which can help to clarify the relationship between an event occurring on the screen and the theme, the plot line, or the logical progression of the dramatic component, or between the instances and noninstances of a concept, the referents and nonreferents of a term, or the most and least effective of a set of proposed solutions to a problem. The unique conventions and capabilities of the televised medium are frequently used to convey special meanings. The manner in which these conventions are used determines their comprehensibility to the viewer and thus their effectiveness in communicating the meanings intended. These conventions include the flashback technique, special lighting effects, combinations of music and lighting, various camera perspectives, fast or slow motion, pixilation, and the matched dissolve between objects. They also include the juxtaposition of events to establish a metaphoric or analogic relationship and elements of fantasy, such as presenting puppets and cartoon characters who move and talk like humans. Still other conventions include the creation of "magical" effects, such as making an object suddenly appear or disappear from a scene or grow smaller or larger, and exaggerating motions and consequences, as with slapstick and "banana-peel" humor. Among other conventions which can be used in more or less comprehensible ways are the speech balloon and the rules for playing of games and for reading, spelling, performing mathematical operations, and interpreting maps. Still others include timing, sequencing, and the use of redundancy, as in repeating an event exactly or with an illuminating variation, restating a point from 268

alternative perspectives, and making use of introductions or reviews. The list could go on indefinitely, a fact which helps to suggest the significane of this attribute in educational television research. Again, as with appeal testing, testing of comprehensibility employs not one but a family of complementary research methods. One very useful approach is to present a program via a portable video playback system to an audience of one or more children, to stop the presentation at predetermined points so as to "freeze the frame," and then ask the viewers what events led up to or are likely to follow from the pictured situation. If the research concern has to do with a premise about a character or his motivation, the viewer might be asked: "What kind of person is he?"; or "Why did he do (say) that?"; or "What do you think he will do next?"; "Why do you think that?"; and so on. In one segment designed for The Electric Company, a musical rock group made up of children - the Short Circus - sang a song which contained the letter combination ow dozens of times. Every time the group sang a word containing the letter combination ow, a printed ow was superimposed on the screen. The intention was to provide repetitive practice in associating the spoken and printed forms of this particular letter combination. By using the method of freezing a single frame, it was possible to evaluate the extent to which members of the target audience actually perceived the speech-to-print correspondence. In this case, the letter combination was frozen on the screen at a point late in the song and, as the experimenter pointed to the printed letter, the subjects were asked such questions as "Why is that there?" or "What does it mean?" In a related method, a program or segment is played through once or twice. It is then presented again but this time without the sound (or, in a variation upon the method, with the sound but without the picture), and the viewer is asked either to give a running account of what is happening or to respond to specific questions. Other methods useful for evaluating comprehensibility include observing the spontaneous responses of children watching a program, testing them for achievement gains following their exposure to a program or program segment, and having them role-play scenes they have viewed. A strength of comprehensibility testing relative to traditional forms of summative evaluation is the opportunity it provides for discriminating between the most and least effective of the many segments devoted to a particular achievement objective. A possible but largely surmountable limitation is the tendency for these methods to produce biased results. Because this kind of testing is done while the program is being viewed and the viewer 269

knows he will be questioned, there is typically an overestimation of a segment's effectiveness. While this bias can be subjectively discounted, it must further be weighed against the possibility that segments which produce no measurable learning when presented in isolation may be effective in combination or when presented along with an appropriate introduction or review. However, this limitation does not detract seriously from the usefulness of such methods. The bias can in fact be turned to an asset, as when it can be shown that a segment of questionable value fails to make its point even when it is evaluated by means of a liberally biased method. Formative research on activity-eliciting

potential

A frequently expressed point of view about the potential of television for instructional purposes is that due to passivity of the viewer, the medium is virtually powerless to produce learning. However, since it is patently obvious that television does teach, it seems desirable to explore how this capability comes about and in what proper and constructive ways it can be exploited. The position taken here is that in spite of the apparent passivity of television viewing, the medium's activity-eliciting potential is perhaps the chief basis for whatever instructional value it possesses. One significant form of activity television can elicit is intellectual activity. It can also elicit verbal behavior and gross physical acts, from modifying the viewer's performance on tests of attitudes and achievements to encouraging him to imitate televised models. It is important to note that the concern of the medium can be either to exploit these effects as instruments of instruction or to foster them as instructional objectives in their own right. Some examples of intellectual activities include integrating separately presented items of information, anticipating upcoming events, forming new concepts, imputing motives and intentions to characters, following progressively developed dramatic and instructional presentations, and guessing answers to questions. The viewer may also take an active role in evaluating relationships between premises and conclusions, between information given and interpretations made of it, and between behavioral ideals and the actual behaviors carried out by the performers, or he may relate televised information to his own previous experiences and his future plans. Tentative indices from formative research on the activity-eliciting capabilities of the medium suggest that many of its assumed limitations may be at least partially surmountable. For example, on the premise that one-way televised presentations permit neither reinforcement nor feedback that can be tied to the learner's actions, it is often assumed that learning through trial and error or through trial and reinforcement cannot occur through this medium. This 270

is not a trivial issue from a practical standpoint, since vast amounts of money may be spent studying the use of two-way communication systems for this purpose; yet - at least conceptually - it is also possible to effect this kind of learning through one-way television, simply by the use of "if" statements. That is, the viewer may be offered a choice among alternatives, given time to make his choice (his point of most active involvement), and given reinforcement — an accuracy check of the form: "If you chose thus and so, you were correct (incorrect)". Empirical studies may or may not support the viability of such an approach, but it certainly deserves further investigation. The notion that certain activities containing a motoric component can be learned only through direct experience is also questionable. For example, learning how to construct alphabetical characters may be more dependent on practice in scanning over the configuration of the letter, on extended or repeated exposure to the letter, and on having an occasion to make and correct the more common errors than on motor activity itself. But all of these are features one-way television can either duplicate or simulate. We need to know more about the possibilities television offers for simulation of learning conditions in which direct, "hands-on" experience has traditionally been considered essential. We also need to know more about the entry skills required for learning to occur and about the possibilities for employing simulation to facilitate subsequent performance in hands-on learning context. All this is not an argument in favor of a widespread substitution of television for physical activity among children, by the way, nor is it intended to deny the great importance of extensive direct experience in learning, especially in early learning. It is intended, rather, to urge open and positive consideration of some of the possible but not yet systematically explored capabilities of the television medium. Formative research on internal

compatibility

Internal compatibility is a program attribute which has to do with the relationship of different elements appearing within the same segment. The basic strategy underlying both Sesame Street and The Electric Company is the attempt to effect instruction through the use of television's most popular entertainment forms. To this end it is essential that the entertainment and educational elements work well together. Without the entertainment, attention strays, and without the education, the whole point of the presentation is lost. In segments where these elements are mutually compatible, the educational point is an inherent part of the dramatic action and is often actually enhanced in its salience as a consequence. In other segments where the entertaining elements predomi271

nate, they may override and thereby actually compete with the educational message. The relationship of elements is also of concern in areas which have to do with auditory-visual, auditory-auditory, and visual-visual compatibilities. The objective of formative research on internal compatibility is to shed light on the program-design features which make for a high or low degree of compatibility. In one method used for assessing compatibility, a panel of judges, using a predetermined set of categories, is asked to rate each segment of a program according to the extent to which the entertainment element either facilitates or competes with the instructional content. Working from each segment's compatibility score, which is a composite of the ratings given by the various judges, it is possible to identify sets of high-rated and low-rated segments and to present the producers with an interpreted list of each type. The interpretations identify program-design features to be emulated, revised, or avoided. Another method involves eye-movement research, which has proved specially useful in the case of The Electric Company because of the extensive presentation of print on the screen and the desire to find ways of motivating the child to read it. In most segments, the print appears on the screen along with competing stimuli. By reflecting a beam of light from the cornea of the viewer's eye and recording the result on a photographic device for later interpretation, it is possible to identify the conditions under which the viewer reads or does not read the print. Once again, the results indicate programdesign features worth emulating and approaches which need to be revised or avoided. Among the important program features focused on by this method are the location of the print on the screen, the effect of various ways of animating print and of the exact repetition of segments upon the elements attended to, and the usefulness of special motivational devices, such as telling all but the punch line of a joke and then presenting that in print. Methods for measuring eye movement obviously have implications also for assessing other program attributes, such as appeal and activity-eliciting potential. Extending

the methods and findings

of formative

research

The search for ways to improve the contributions of formative research is a continuing process at CTW. Suggestions for new research methods and critical appraisals of current methods have been published not only by the CTW research staff, but also by a wide variety of outside consultants (e.g. Rust 1971a, b, c; 1974; Mielke and Bryant 1972a,b; O'Bryan and Silverman, 1972). Although it is not a primary mission of CTW, the staff also has a high 272

interest in relating the ever expanding body of formative research findings to more general theories of learning. An excellent contribution has been made by Gerald Lesser (1974)*.

Organizational and interpersonal factors As technologically sophisticated forms of instruction come into increasing prominence, it will be necessary to make increased use of production teams whose members possess a diversity of highly specialized talents. In anticipation of this trend, we need to know more about related organizational and interpersonal conditions. These conditions deserve attention in any attempt to establish a working partnership between television-research and production groups, and they play a role which is even more prominent in the formative research context than in the context of more traditional approaches to educational research. An important factor in CTW's case has been the opportunity for the members of the two groups to learn about each other's areas of specialization during an eighteen-month prebroadcast period. Furthermore, every new formative research approach is treated as an experiment to be continued or discontinued depending on its evaluation by the producers themselves. The fact that CTW's researchers and producers possess not the same but complementary skills is also significant, largely because it makes for clear and distinct functions on the part of each group. Still another factor is that the producers, before joining the project, made the commitment to try to work with formative research. This advance commitment helped to support the cooperative spirit through the early, more tentative period of the effort. Also, research never takes on the role of adversary to be used against the producers in winning a point or pressing for a particular decision. On some matters, the producers must hold the final power of decision and be free to ignore research suggestions if production constraints require it. In all, the factors consciously dealt with in the interests of researcher-producer cooperation have ranged from the careful division of labor and responsibility to housing the two staffs in adjacent offices and from patience and diplomacy to occasional retreat.

* A comprehensive CTW Research Bibliography (1979; revised edition) is available from the Research Division, Children's Television Workshop, 1, Lincoln Plaza, New York, N.Y. 10023.

273

The distinctive role and functions of formative research Formative research is distinguished primarily by its role as an integral part of the creative production process. It is important to maintain a clear distinction between this type of research and summative research — that is, research undertaken to test the validity of a theory or the measurable impact of an educational product or practice. Research undertaken in the context of scientific validation is concerned with effects which have been hypothesized a priori within the framework of a broader deductive system, with the use of empirical and statistical procedures well enough defined as to be strictly replicable (at least in principle), and with the highest possible degree of generalizability across situations. While research carried out within the formative context can possess these same characteristics, it need not and typically does not. The main criterion for formative research recommendations is that they appear likely to contribute to the effectiveness of the product or procedure being developed. It is neither expected nor required that they be validated by the research out of which they grew. Establishing their validity is the function of summative research. As this view implies, to achieve the objectives of formative research it is often necessary to depart from traditional research practices and perspectives. This is not to say that experimental rigor has no place in the formative context. However, even where strict experimental and control conditions have been maintained, there is seldom anything to be gained by using tests of statistical significance. The creative producers often prefer to work directly with information about means, dispersions, and sample size. Also, whereas matching of experimental and control groups on the basis of pretest scores is discouraged where inferential statistics are to be used, because of the conservative effect on the usual tests for the significance of the results, such matching can be very useful in the interests of efficiency and the maximization of the reliability of information based on small samples. In the area of sample selection, it also can be useful to depart from the traditional practice of including all age and socioeconomic groups for which the educational materials are intended. Time and effort may often be saved by selecting a sample of average performers, of performers from the high and low extremes, or, where the intent is mainly to upgrade the lowest performers, a sample only of those. In general, where biased methods of sampling and biased methods of testing are more efficient than unbiased methods, and where the objective is not to make accurate population estimates, it is often useful to exploit the very biases which quite properly would be avoided in other research situations. 274

In practice, it tends to be difficult for researchers trained and experienced in traditional approaches to adopt an appropriate formative research point of view. In the formative situation, their first responsibility is to improve a specific product or practice and not to contribute to a general body of knowledge (though the two objectives certainly are not incompatible). In such instances, studies must first address the information needs of the product designers and not the special theoretical interests of the researchers. Where it is economically impossible to cover a wide range of empirical questions and to rigorously report or establish careful experimental conditions, and where the usefulness of the results is not unduly compromised as a consequence, the former course may deserve priority. Quantitative indices, such as percentages, and highly detailed item-level data, if they communicate most effectively with the creative producers, are to be preferred over those which conform to standard practice for research reports. Broad, speculative interpretations of empirical results are typically more useful than interpretations limited to the more strict implications of a study. And, as indicated earlier, biased methods of sample selection and testing often can be employed to good advantage. However, in following these departures from standard research practice, there is a risk of producing misleading results. Accordingly, it is essential that resulting production recommendations be appropriately qualified. Formative research, in my view, is properly eclectic and pragmatic. In these respects, it is highly compatible with the current trend toward defining instructional objectives very explicitly, then developing through systematic trial and revision instructional systems for achieving them. This approach, incidentally, in no way diminishes the traditional role of the behavioral sciences in education or the usefulness of existing theory and knowledge. Rather, it holds that a useful step between basic research and educational practice is additional research of a formative sort, directly concerned with specific combinations of educational objectives, instructional media, learners, and learning situations. This is not to say that formative research is exclusively concerned with putting theory into practice. It has the equally valid function of starting with practice and transforming it into improved practice. It also has the function of providing hypotheses for further research and theoretical development. This is, incidentally, what is coming to be the dominant conception of the technology of education — a commitment not to teaching in the older audiovisual tradition, but to achieving a planned educational effect. One long-standing point of view in education holds that theories and results growing out of the "mother" disciplines of psychology, sociology, anthropology, and the like, will filter into effective educational practice if 275

enough educators have been trained in these basic disciplines. While this approach has been useful to a degree, it has not produced broadly satisfactory results. Meanwhile, creators of new educational products and practices have proceeded largely without the benefits of measurement and research. This is partly because skill and training in these areas have been linked to the process of theory construction and validation and partly because there has been an inappropriately rigid adherence to traditional research practice within the product development context. Formative research procedure promises to help in creating a mutually constructive relationship between these two overly isolated realms — the science and the technology of learning.

Outlook The field of formative research in educational television has been both advanced and set back by recent events. In academic centers all around the U.S., a new and vital tradition of basic social science research on televison's positive effects has emerged, much of it concerned with variables and constructs of interest to educational television practitioners. Computer data collection and processing has made turn-around of results far more rapid, and thus far more useful. Moreover, computers have given rise to the creation of new and effective forms for graphic display of results. Case studies in formative television research have burgeoned in the U.S., largely due to the fact that formative planning and research, following the CTW model, has come to be required in nearly all Federally funded educational television ventures. A major setback has been brought about by the unstable and, of late, greatly curtailed amount of Federal funding support for all areas of social need, including that for public television programming. One very specific result has been to reduce drastically the ranks of formative researchers actively involved in program planning and research, and in contributing to an exchange and accumulation of practices in the field. A magnitude of this problem is reflected in recent trends in children's educational television production. In 1970, there were two new or substantially renewed children's series on U.S. public television. By 1976, the number had risen to ten. In 1979, there was but one. This one, Sesame Street, continues in its fifteenth season to make use of formative research on a regular and ongoing basis, as its subject matter is expanded and its production approaches modified. The research group for this series also maintains an ongoing exchange with researchers involved in Sesame Street adaptations in such countries as Kuwait, Mexico, Israel and the Philippines. 276

The recent formative research on 3-2-1 Contact has made a fine and durable contribution in both methodology and guidelines for program design for older children (see Mielke and Chen, 1981; Mielke, 1983). Fluctuations in program funding over future years may see formative television research advance by spurts, but there can be no doubt that it will continue to play a key role wherever serious attempts are made to educate through television.

References Ball, S., and Bogatz, G. A. The first year of Sesame Street: An evaluation. Princeton, N. J.: Educational Testing Service, 1970. (ERIC Document Reproduction Service No. ED 047 823). Ball, S., and Bogatz, G. A. A summary of the major findings in "The first year of Sesame Street: an evaluation". Princeton, N. J.: Educational Testing Service, 1970. (ERIC Document Reproduction Service No. ED 122 799). Bogatz, G. Α., and Ball, S. The second year of Sesame Street: A continuing evaluation (Vols. I & II). Princeton, N. J.: Educational Testing Service, 1971. (ERIC Document Reproduction Service Nos. ED 122 800, ED 122 801). Bogatz, G. Α., and Ball, S. A summary of the major findings in "The second year of Sesame Street: a continuing evaluation". Princeton, J. J.: Educational Testing Service, 1971. (ERIC Document Reproduction Service No. ED 122 802). Cooney, J. G. The potential uses of television in preschool education: A report to Carnegie Corporation of New York. New York: Children's Television Workshop, 1967. (ERIC Document Reproduction Service No. ED 122 803). Cronbach, L. J. Course improvement through education. Teachers College Record, 1963 , 64, 672-83. Dick, W. A methodology for the formative evaluation of instructional materials. Journal of Educational Measurement, 1968, 5, 2. Grooper, G. L., and Lumsdaine, A. A. Studies of televised instruction. Pittsburgh: Metropolitan Pittsburgh Educational TV Stations WQED-WQEX and American Institute for Research, 1961. Hastings, J. T. Curriculum evaluation: The why of the outcomes. Journal of Educational Measurement, 1966, 3 (1), 27-32. Lesser, G. S. Children and television: Lessons from Sesame Street. New York: Random House, 1974. Mielke, K. W. Formative research on appeal and comprehension in 3-2-1 Contact. In Bryant and Anderson (Eds.), Children's Understanding of Television: Research on Attention and Comprehension. New York: Academic Press, Inc., 1983. Mielke, K. W., and Bryant, J., Jr. Formative research in attention and appeal: a series of proposals. New York: Children's Television Workshop, 1972a. (ERIC Document Reproduction Service No. ED 126 867). 277

Mielke, Κ. W., and Bryant, J., Jr. Formative research in comprehension of CTW programs: a series of proposals. New York: Children's Television Workshop, 1972b. (ERIC Document Reproduction Service No. ED 126 868). Mielke, K. W. and Chen, M. Children, television and science: an overview of the formative research for 3-2-1 Contact. New York: CTW, 1981. O'Bryan, K.G., and Silverman, H. Report on children's television viewing strategies. New York: Children's Television Workshop, 1972. (ERIC Document Reproduction Service No. ED 126 871). O'Bryan, K. G., and Silverman, H. Research report: Experimental program eye movement study. New York: Children's Television Workshop, 1973. (ERIC Document Reproduction Service No. ED 126 870). Palmer, E. L., et. al. A comparative study of current educational television programs for preschool children. Monmouth: Oregon State System of Higher Education, 1968. (ERIC Document Reproduction Service No. ED 032 123). Palmer, E. L. Formative research in educational television production: the experience of the Children's Television Workshop. In W. Schramm (Ed.), Quality in instructional television. Honolulu: University Press of Hawaii, 1972. Palmer, E. L. Formative research in the production of television for children. In G. Gerbner, L. P. Gross, W. H. Melody (Eds.), Communications technology and social policy. New York: Wiley, 1973. Palmer, E. L. Formative research in the production of television for children. In D. R. Olson (Ed.), Media and symbols: The forms of expression, communication and education. Chicago: University of Chicago Press, 1974. Reeves, B. F. The first year of Sesame Street: The formative research. New York: Children's Television Workshop, 1970. (ERIC Document Reproduction Service No. ED 047 822). Rust, L. W. Attributes of Sesame Street that influence preschoolers' attention to the TV screen. New York: Children's Television Workshop, 1971a. Rust, L. W. Attributes of The Electric Company pilot shows that produced high and low visual attention in 2nd and 3rd graders. New York: Children's Television Workshop, 1971b. (ERIC Document Reproduction Service No. ED 126 872). Rust, L. W. Attributes of The Electric Company that influence children's attention to the television screen. In-house research report. New York: Children's Television Workshop, 1972. Salomon, G. Sesame Street in Israel: Its instructional and psychological effects on children. Jerusalem: Hebrew University of Jerusalem, 1974. (ERIC Document Reproduction Service No. ED 122 814). Scott, R. O., and Martin, M. F. The 1969-70 classroom tryout of the SWRL Instructional Concepts Program. SWRL Technical Memorandum, 1970 (a). Scott, R. O., Castrup, J., and Ain, E. The SWRL Kindergarten Art Program. SWRL Technical Memorandum, 1970 (b). Scriven, M. The methodology of education. In R. W. Tyler, R. M. Gagne, and M. Scriven (Eds.), Perspectives of curriculum evaluation. AERA Monograph Series on Curriculum Evaluation. Chicago: Rand McNally, 1967.

278

Saul Rockman

Formative research and evaluation of instructional television programs The Agency for Instructional Televison (AIT) is not widely known in the field of children's television, but annually it provides almost one-half of the school television programs broadcast in the United States. AIT develops and acquires from stations and independent producers television series for distribution to educational television organizations and school systems. They, in turn, transmit the programs to the elementary and secondary classrooms in their jurisdiction. Instructional television and education in general is, in the U.S. and Canada, a state/provincial, and local endeavor. At the national level there are no curriculum requirements and few restraints on curriculum. The instructional television series AIT develops are in conjunction with consortia of state (U.S.) and provincial (Canada) educational agencies. These consortia are cooperative ventures in which each state, province, or local agency wishing to participate contributes a small portion of the total cost and, in turn, receives a product of high quality it could not afford to create alone. Twenty-five to forty agencies generally participate. Over the past decade these efforts have been unusually successful in creating fourteen school television series meeting the highest educational standards and often attaining awardwinning production quality. AIT's efforts in bringing together state and provincial education agencies to thoughtfully consider common productions have made it a leader in the field of school television. AIT has introduced innovations in distribution, off-air recording rights, instructional design systems, and evaluation. It is, however, the consortium funding arrangement and the development process that stand out as AIT's major contribution. This chapter will focus on AIT's research and evaluation activities as they relate to program production. Producing school television programs is the fundamental reason for creating a consortium and is the major activity in a consortium project. However, it is not the entirety of the consortium process. Before I can detail some of the production activity, I need to outline the consortium process, illustrate some of the AIT programming, and describe how it is used in the classroom.

279

The AIT consortium development process The consortium development process is a means of fostering the production of widely useful instructional television series to meet significant educational needs. This consortium process has three stages: initiation, production, and assimilation. Research and evaluation play a role in each of these stages. Part of the effort is formative evaluation in the production stage, but there are other research initiatives of interest to producers of children's television. Let me take the broadest research perspective as I describe the entire development process. Initiation The first step in a consortium project is to identify a significant, widely-shared curriculum. Since the potential members of any consortium are from states and provinces in the U.S. and Canada, each with their own separate education systems, the subject cannot be a parochial one, such as regional geography. Since a consortium project aggregates large amounts of money to use a mass medium, the subject cannot be highly specialized, reaching only a few students in each jurisdiction. It should take advantage of the economics scale. Since agencies must elect financial participation in a project, the content must be important to education. The importance comes not from universality — spelling, for example — but from instructional significance — basic economics, for example. As a result of education reform, research in child development, and the widespread marketing of standard textbooks, apparently independent state and provincial education systems have curricula with much in common. Classrooms across North America usually teach the same ideas in the same grades. There is often more disparity among classrooms in the same state or province than between average classrooms in different states or provinces. There is no reason to create a television series if it is merely a way to perpetuate mediocre teaching practices of the present. Educational experience over the past twenty years has also shown that introducing instructional innovations that differ enormously from current teaching practices is dooming them to failure. Curriculum and teaching practices can be changed gradually, slowly; thus, the pedagogical approach to the series content must be carefully identified and verified. If it is too far ahead of what teachers are doing, they are not likely to use it; if it is only doing what they are already doing, they don't need it. Television, when used to advance instruction and curriculum, can be a powerful tool, especially when its function and content are carefuly selected to meet the expressed needs of education. 280

Evaluation at this stage is a verification that sets out to confirm or refute the curriculum need and its proposed treatment. In discussions with leaders in the content area, with curriculum supervisors at state and provincial education agencies, and with classroom teachers of the subject, the ideas take shape and are modified. As in formative evaluation, the goal is to continue to revise. At this stage the revisions are undertaken to develop an idea that is widely acceptable and that will contribute in an important way to moderate curriculum reform. Among the issues reviewed are the need for new television materials in the subject area, the compatibility of the proposed treatment of the content with existing teaching practices, and specific topics to be included in the television series. The information collected during this needs determination and verification process becomes the basis for a proposal that is then submitted to the fifty state education and television agencies in the U.S. and the ten provincial ministries of education in Canada for funding. At its most fundamental level, the verification process is the market place of ideas, and each television project must compete for relatively scarce resources with many other worthwhile educational activities. An agency's decision to participate financially is an indication of the importance of the project to the agency. Production While formative evaluation and the application of evaluation findings to the production of television programs will be covered in detail below, certain aspects of the production phase are worth mentioning at this time. The production phase of a consortium project includes the design and scripting of the television programs, the production and evaluation process, and the preparation of print materials to accompany the television programs. These activities are managed by AIT, and some components — most prominently the production itself — are contracted services. Overseeing AIT's management of the project is the consortium, the funding agencies. Beyond its financial contributions, the consortium participates in the development of instructional materials by discussing and reviewing the project's progress and the materials it produces. Usually twice during the production phase, the consortium meets to look over the project. AIT brings representatives together at first to discuss the instructional design for the programs and, later in the process, to view and discuss the initial programs in the series. Discussions at these meetings often lead to suggestions for modification and refinement. By participating at these meetings, consortium agencies help shape the project and make it their own. And by doing so, they increase their knowledge of the project and their commitment to use it. 281

The creation of classroom television materials flows from an instructional design model. It begins with a design team that creates the "prescription" for the individual lessons, including objectives and suggested content ideas. The design team consists of content specialists, television producers, teachers of the designated student audience, instructional designers, developmental psychologists, researchers, and others whose expertise will contribute to a useful instructional design. Evaluation plays a role here also. The instructional design is reviewed by the consortium and is also verified with classroom teachers, with subject matter specialists, and occasionally with students. The issues raised at this stage of development include an assessment of the objectives for their instructional validity, age appropriateness, and fit with current classroom teaching. The evaluation seeks recommendations for the incorporation of specific content ideas and suggestions for how topics should be introduced and treated within the programs. Suggested script topics are also commonly evaluated and new script ideas solicited. Reactions lead to revisions of the design report, as a set of prescriptions to guide the production of the television programs. From the design document will emerge several draft versions of treatments (two- to three-page narratives of what the program might be like), multiple drafts of scrips, and film or video programs. At each stage, evaluation activities help shape the instructional materials. AIT's belief has been that it is less expensive to modify instructional design, treatment, or script — than to change film or tape. The efficiencies gained at this stage are extremely important. Funds for revision of the programs are built into the project's budget, but these funds, like all funds, are limited. We quickly run into the Catch-22 of evaluation: the more you evaluate, the more you revise; the more you revise, the less money you have for additional revisions. At some point you stop evaluating because you can no longer afford to revise your programs. You can stretch your evaluation and revision money by catching the big errors in the paper stage and by making many more smaller revisions at the rough-cut stage. As this belief was developing from experience, AIT began to shift its evaluation efforts from rough-cuts to earlier stages and has developed evaluation techniques to help improve paper products as well as film and video products. After a script has been evaluated, revised, and approved, it is sent to the production agency and the film or tape shot and assembled into a rough-cut. But formative evaluation continues. The products are tried, not in a laboratory setting, but in representative classrooms, a setting in which they will be used. The distractions in the classroom are the same ones that will be present when the program is eventually used; the variety of teachers that will use the finished program can also use it in evaluation studies; the diversity of student abilities and interests can be well represented. 282

Revisions in the programs are suggested and, if substantial changes are made, the program may be taken out for further evaluation. At some point a program and its related print are considered finished and prepared for distribution to the consortium agencies. Implementation

and

assimilation

One benefit of belonging to the consortium is that the initial investment of risk capital permits each consortium member to duplicate and distribute the series to any and all schools in their jurisdiction. They are, in a sense, the owners of the television series. By participating in the consortium meetings and in the evaluation activities, consortium agencies develop a sense of ownership. State and provincial educational agencies are willing to put substantial effort into promoting and helping teachers use a series that belongs to them. AIT, on behalf of the consortium, will undertake nationwide promotion and will design teacher training components. Nevertheless, the responsibility for successfully implementing a new television series rests with state, provincial and local educators. They must introduce the series to their teachers; they must schedule the series to fit their local needs; and they must distribute teacher guides to their schools. Implementation begins even before the programs are completed. Planning and conducting in-service training for teachers is started well before the initial broadcasts of the series. Implementation continues even after the first year of use. Teachers already using the series prepare others to use it. Users learn more about the series each time they view it with their class. We have learned that in the first year of a series, teachers are trying to understand what it is all about; their use of a series is almost mechanical. They are trying to master the tasks of using the materials and relying heavily on the teacher's guide to get them through. During the second and possibly the third use of the series, activities stabilize; then comes refinement, in which teachers modify their use of the series to increase its effects on students. In some cases, summative evaluation or impact research might be conducted. However, I feel strongly that the results are worth the effort only if they can be used to plan future series or implementation activities. The limited resources for research are much better spent on formative evaluation and program revision. While summative studies of the series' effects in the classroom have political utility, insuring each lesson's appeal, usefulness, and in some limited fashion, its success in reaching objectives, makes formative evaluation a far better investment. While the funders are asking "How do I 283

improve it?", this latter question is the more interesting one. Robert Stake differentiates between formative and summative evaluations by noting that "when the cook tastes the soup, that's formative; when the guests taste the soup, that's summative." The value of both kinds of evaluation is not to prove, but to improve. In formative evaluation, one improves while the product is being created. In summative evaluation one learns from one product and plans for improvement in the next set of materials. For the producing agency, summative evaluation has usefulness for formative purposes. By learning about the overall effectiveness and eventual adoption and use by the audience, the producer increases his understanding of what works and what does not. The next series will likely profit from it. My reason for discussing the consortium development process in such detail is not that any individual component is necessarily unique to AIT, but rather to point out the comprehensive nature of the process and to note the exceptional care taken in the development of a television series. It is not a rule that all evaluation and verification activities are given equal emphasis or that each project can afford to take advantage of all possible evaluation opportunities. Each project has unique needs and each its own budget. Each project staff has its own concerns and tolerance for information. The consortium development process, with all its various components, has been elaborated in detail by John Middleton (1979). He has attempted to capture the set of activities as shown in Figure 1. In recent years AIT has attempted to compress the time line shown in this figure, especially to shorten the period necessary to obtain the financial commitment of consortium agencies.

Formats and programs The consortium projects undertaken by AIT have covered such topics as career development, applied mathematics, social and emotional growth, the metric system, affective health issues, early childhood education, visual arts, and problem solving. The grade levels for these series range from kindergarten through high school. The individual programs tend to be 15 to 20 minutes in length and the series usually have either 12 to 15 or 30 units, providing schools one program per week for a semester or for a school year. Almost all of these school television series have a similar format. They are short dramatic narratives, illustrating and applying the concepts to be taught. In contrast to the work of the Children's Television Workshop — whose 284

285

programs are composed of a large number of very short segments or vignettes — AIT's approach has been to develop dramatic, story-line programs that present both information tied to the curriculum and a social model for how to deal with that information. The dramatic format is undoubtedly expensive. Next to full animation, onlocation drama is probably the most expensive film or tape format in use. However, it is also, next to full animation, the most appealing, especially to children and youths. While I have been talking about 15- to 20-minute storyline programs, they are often less than fully plotted dramas. Our own research and that of others have indicated that young children often have difficulty following a long series of plot events. Even children 10 to 12 years of age do not always comprehend the twists of a complex plot, especially when contemporary cinematic conventions are used. So for younger students, the dramatic narrative may simply cover a single incident with its effects having a linear beginning, middle, and end. Subplots, flashbacks, and dream sequences are used with extreme care. This particular format was not chosen solely on appeal. Undergirding its selection is a belief of the role of television in the classroom. Most directly it can be stated as: "Teachers teach; television helps them do a better job". Instructional television is not designed to replace a teacher, rather to do what a teacher is often unable to do. Television can motivate, show applications in the real world, create an appealing context for further teaching, and so forth. Television is not at its best in doing direct teaching, something a teacher can usually do better. Television can effectively teach; many examples from Sesame Street come to mind immediately. But that series and other similar ones were designed to be viewed at home, without a teacher present. (Even studies of the impact of Sesame Street show greater learning with an adult "teacher" present during viewing.) Television can teach better with a teacher present and teachers can be more effective when using television to help them teach. Teachers can introduce ideas; television can show extensions and applications of those ideas, and teachers can then use the television examples to reinforce and complete the teaching. Back to the dramatic format. Let me illustrate how it is used to convey information and present social models by describing several programs from different school television series. Self Incorporated is a series of fifteen 15-minute programs designed to help eleven- to thirteen-year-olds cope with the problems that arise as a result of the physical, emotional, and social changes they are experiencing. One 286

program, Pressure Makes Perfect, deals with the pressure to achieve felt so strongly by young people: Nan is quite a talented musician. With a lot of practice and the desire to achieve, this thirteen-year-old could become an outstanding pianist. Nan's parents want her to succeed very much and have provided her with a good piano and excellent instruction. But in their wish to help their daughter they unwittingly increase the pressure that Nan feels. As she tries to prepare a difficult piece for her recital, her parents' own hopes blind them to Nan's feelings of pressure, and they respond by demanding more of her. Nan deals with the tension by rebelling. At the recital she plays a short, simple piece instead of the more difficult one, bangs her fists on the keyboard, and shouts defiantly to the stunned audience: "And that is my recital!" With that, to the consternation of her parents and her music teacher, she stalks off the platform. Trade-Offs is a series of fifteen 20-minute programs designed to help nine- to thirtheen-year-old students think their way through economic problems and increase their understanding of economic principles. We Decide is a lesson about social decision-making which applies a five-step economic decisionmaking model, emphasizing the steps of generating alternatives and establishing criteria. The first day back at school, Donna and her seventh-grade friends realize that there are not enough bike racks for everyone who brings a bike and decide to find a solution to the problem. At first, they think of different ways to raise money to purchase more bike racks, but school policy does not permit this. The students then identify several alternative ways of assigning the limited number of spaces. They could: take turns using the spaces; allocate by drawing; assign to those using them the longest; do nothing and have a first-come, first-served system; or assign them to those who live furthest from school. The class discusses and evaluates the different alternatives according to the criteria of freedom, equal chance, and certainty. Limited animation is used to review all five steps of the decision-making process and uses a decisionmaking grid to evaluate each alternative in terms of the criteria. At the end of the program, viewers are asked to vote for the alternative they think best. ThinkAbout is a special series of sixty 15-minute programs designed to strengthen the reasoning skills of ten- and eleven-year-olds and to reinforce their language arts, mathematics, and study skills; it focuses on the common 287

denominator of the basic skills: thinking. The program Get Ahead Goals, is a general aid to independent learning and problem solving:

with

Leon, an average twelve-year-old, sets a great many goals but never does anything about reaching them. This time it's going to be different, he tells his sister Susan. He really wants to try bicycle racing. His friend Ed, who is getting ready for the regional bike racing championship in six months promises to help him. During a planning session with Susan, Leon sets an overall goal — to be a good bike racer and go to the regional races. Then he lists three steps he will have to take in achieving this goal: 1) buy a new bike, 2) train, 3) learn bike racing strategy. After reading library books about bicycles and consulting with Ed, Leon decides what bike he wants. He earns the money by sweeping leaves and enlisting new customers on his paper route. The training is a slower process. With advice from Ed, he conditions his body by jogging, working out with weights, riding his bike, and eating nutritious meals. Learning the racing strategy is more difficult still. After seeing three races, Leon wisely decides he is not ready for the regionale. He modifies his original goal and plans to make his debut in a local race instead. His realistic goal-setting, careful planning, and willingness to followthrough pay off. In the local race Leon finishes respectably. What these programs have in common are appealing events appropriate to the age and experience of the viewers. These events present a problem to be solved, a problem that relates to the curriculum, and that calls on the teacher to participate with his or her students in finding a solution. There are some subject areas for which this approach is not appropriate — in teaching the fundamentals of division for instance — but also allied tasks for which it is — showing when to divide instead of subtract by examples of everyday usage and common mistakes in applications. Television of this nature is often used to show the "why" and the "when" rather than the "what" or the "how". This kind of television program, because of its relationship to the curriculum and of its open-ended nature (the problem set up in the program might not be solved or completed satisfactorily), needs to have a teacher working with it. Rather than being "teacher-proof", these programs set out deliberately to engage both students and teachers simultaneously and have them work through the issues together. The events depicted in the program are designed to be a common stimulus for teacher-student interaction. They may be likened to "trigger" films, presentations meant to stimulate or initiate certain kinds of behaviors, in this 288

case, discussions. These discussions are not random, as our observations indicate, but begin with the events depicted in the program and move away from them into the course of study planned by the teachers. The goals of specific programs reflect this process. Objectives often refer to "recognition" and "awareness" rather than to the teaching of specific knowledge. The story-line, the basis of the discussions, is selected and written to be "real" to the viewers. There is no expectation that each child will have experienced the events depicted in the program. The events, and the age-appropriate actors portraying them, are meant to be plausible in the lives of the children themselves or people they might know. The underlying themes of many of the emotional health and effective programs for example, are universal (e.g. pressure to achieve) and the specific portrayal builds on the theme as well as setting up the story-line events. Thus children recognize the major issue for discussion and build on the actions of the characters to identify behaviors associated with that theme. These programs take advantage of common themes and experiences that children in the target audience might have observed or experienced themselves. They build on social reality while adding behavioral models with whom the viewer can readily identify. Television's influence on children's learning is not based on mere viewing of the program but on the context for viewing and discussion provided by the teachers in the classroom. The teacher plays a central role in helping the students interpret the events on the screen and, when necessary, providing an explanation or clarification. By leading a post-viewing discussion, the teacher controls its direction and its relationship to classwork. The construction of social reality takes place within the complex structure of the classroom. The nature of the discussion process is fairly similar from class to class within grade levels and, given certain latitude, across grades. Our observations have shown that discussions almost always begin with a review of the events in the program. This permits clarification of any confusion and a confirmation that each student is working from the same set of understandings about the program. The teacher and the students then commonly move away from the program to events or applications of the concept in their lives or in the experiences of their friends. This change carries them from the concrete "reality" of the program to the actual social reality of their own lives. From this point the discussion begins to generalize to the theme or issue of the program or can be directed to some specific class event or lesson. Under the control of the teacher, the classroom interaction will lead the students to engage the program's instructional content and its potential applications to classwork or to actions outside of class. 289

Figure 2: Student Classroom Discussion Profile BUL PROGRRMS. ALL VERS IONS...COLUMN VRR1 OBLES

We have used a structured observation system to record and analyze the classroom discussions of many classes, and for several different series. We backed up the structured observations with running narratives of the discussions taken by another observer also present in the classroom. The categories we have established clearly illustrate the progression from talking about the program to talking about the ideas and concepts in their own experiences (self/others) and to talking about the ideas in the abstract (others/concept). As Figures 2 and 3 show, the change over time is substantial and make changes in discussion patterns quite obvious. These data are summarized for eight programs in the Inside/Out series, an affective/emotional health series. More than 150 third- and fourth-grade classes were observed. The parallels in the discussion patterns between teachers and students suggest that teachers move away from talking about the program and attempt to steer the discussion toward the instructional topic. At the same time, students want to deal with their own experiences, but also seem 290

Figure 3: Teacher Classroom Discussion Profile RLL PROGRAMS. ALL VERSIONS...COLUMN VARIABLES

willing to engage the instructional topic. Teachers gradually begin to relate their own experiences too. In similar studies with students in the sixth and seventh grades, we note a much greater reluctance to talk about themselves; but students are still willing to talk about their friends and the experiences of others they know. Teachers of these students do not talk very much about themselves or about their experiences, while they stress the instructional message. The discussions are relatively short for these classes, about 10-15 minutes. When we get to high school, grades ten and eleven, the pattern reverts to one similar to the youngest students. Both teachers and students seem willing to personalize and engage the instructional theme of the program. These discussions, like the third and fourth graders, last about 20 minutes. These studies point out the manner in which teacher and students interact after viewing a program. The programs serve as the starting point for the discussions and lead the class into the topic. 291

Instructional television programs are used in a social context. Teachers and students interact with each other and with the program. The value and impact of a specific program is influenced by what the teacher does prior to showing it, what the students bring to the program, the teacher's ability to incorporate the program into the instruction, the pedagogy of the program, and the effects of the educational setting in which all of this takes place. The program does not stand alone, it becomes part of a lesson, one in a sequence of many lessons during the course of a day or week.

Production and evaluation Now that the reader has a sense of how AIT develops projects, the nature of the programs, and the way they are used in the classroom, we can turn to specifics of the production process and the evaluation data that affect it. Evaluation is not an independent venture. It is closely associated with the product (school television programs) and the people who create it. Much of what is sought is at the behest of those creating the program — designers, producers, writers — and much is dependent on the stage of the product's creation. The development of a children's television program is a series of approximations, each closer to what will finally appear in the classroom. One can move through content design, treatments, scripts, and rough-cuts to the finished program. At each stage are numerous opportunies for evaluation. At each stage questions can be raised and answers provided to improve the product, and both questions and the answers may change from stage to stage. Figure 4 illustrates the design and production of one consortium project — ThinkAbout — along with the verification and evaluation systems. The ThinkAbout project was more comprehensive than most and therefore demonstrates the range of possibilities. The verification and evaluation activities are designed to work in concert and contribute information and ideas to each other. In the earliest stages of the project, the verification system was part of the evaluation, in the later stages it became part of the project director's set of political checks and balances. But if evaluation of one sort or another can be undertaken at any stage in the development of the program, what is worth examining when? And with what effect? Issues for

evaluation

Imagine, if you will, a three-dimensional matrix. On one axis are the successive stages of a program's creation from treatment through finished 292

Full Production/May 1978-February 1979

κ β co S o

bo

O

8 ÉL

a o •O κ ß o í-t o M S 13 a

α> a £α "ω > ω Ο

s SI £

e o ·δ c ä s JZ « "Ό υ -Ό o uO HtbD.

' • H l ·

O U

«

o Ζ

293

product. On the second axis are listed various kinds of evaluation data, such as attention, appeal comprehension, content accuracy, and so on. On the third axis are the sources from which information can be obtained, including teachers, students, and content specialists. This three-dimensional matrix is a structural model from which evaluation activities can be selected. The criteria for selection are determined by those involved in the production process in conjunction with the evaluator. They most likely will include the utility of the information, the time, staff, and fiscal resources available, and the specific evaluation techniques known by people making the decisions. Let us ignore for the moment the decision makers for whom evaluation data are being collected. Assume, quite reasonably, that various participants in the production process — designers, writers, producers, and the consortium agencies — are all interested parties in evaluation, stake-holders, if you will. What information is of interest in various stages of production? What data can be collected and be productively used? Collected from whom? And in what manner? I want to walk through some of the cells in the matrix, using the stages of a program's development as stepping stones and using examples of AIT's evaluations as illustrations at each step. First, however, a few definitions might be useful: Appeal: The extent to which the audience(s) likes the characters, story events, content, and cast. Associated with appeal are also issues such as credibility, or realism, age-appropriateness, and message relevance. Attention: In what parts of the program and to what degree does the audience attend to the screen, to the audio portion or to the interaction of audio and video? Attention can include measures of eyes on the screen, appropriate responses to humorous or sad segments, or talking back to the set. Comprehension: Comprehension of story-line is the audience's understanding of the intent or action in a scene or program. Comprehension of the objective is the extent to which the audience grasps the point(s) of the program. Effectiveness or Learning: The extent to which the appropriate audience achieves the objectives, whether those objectives focus on values, attitudes, interests, or performance. There are other issues, such as content accuracy, realism, and salience, whose meanings are self-evident or that can be defined in context. Treatments The earliest approximation of a program is the two- to three-page narrative treatment, an attempt by the writer to place the goals and objectives in a dramatic framework. It might include a description of the characters, the 294

story-line, and sample dialogues. The instructional designer would look at it to see if, on the surface, it met the specifications he set down. The producer would examine it to see if it could be accomplished within the budget. The content consultant would check for content accuracy and suggest ways of incorporating more (and more precise) information. Beyond these staff reviews, there are other opportunities for collecting useful information. At the treatment stage, evaluation can explore potential. Can the ideas become an interesting and instructionally effective program? If the approach taken by the treatment cannot carry the instructional load or if the ideas have little appeal for the teachers and students, then it is better to scrap the treatment than face the same problems at the script stage. It is easier to reject the treatment because writers and producers have little invested in it. The evaluator is the primary link between the project and the target audience. By taking the treatment out to classrooms, the evaluator can quickly obtain reactions and suggestions so that a writer is not stalled for too long at a time. At AIT we plan for less than ten working days between the submission of a treatment or script and a telephone report of student and teacher data. When necessary it can be less, but the vagueries of the mails and the habits of script writers often prevent the scheduling of class visits much before the script or treatment is in hand. When we take a treatment to a classroom, we usually read it to the class as a narrative without any embellishment. After listening to the presentation the students fill out a questionnaire about their immediate reactions to the story and the characters and respond to questions about instructional content. A class discussion ensues, led by the evaluator. He or she probes to see if the ideas were appealing and relevant to students, if they could identify with the characters, and if they saw the events as realistic. The evaluator also asks about other ideas that might fit into the program, and about which parts were successful and which unsuccessful. In a series on economics for students in the eighth through tenth grades, a program on competition and monopoly business practices went through some interesting changes in the treatment phase. The first treatment used electronic games as the focus of attention and was rejected by the project staff due to its oversimplification of the monopoly issue and the possible faddish nature of electronic games. The second treatment was sent to teachers for subject matter reviews and was taken out to classrooms. In this version a school photographer taking yearbook pictures started to receive some competition from students dissatisfied with his work. The content specialists found a few correctable errors, and thought the idea relatively weak. Students, however, found the entire topic trivial; the school photographer came but once a year 295

and existed on the periphery of the students' world. The story lacked relevance. Because of these findings, a new treatment — based on ideas discussed in the evaluation classrooms about food concessions at a school basketball game — was commissioned. This was a more relevant setting in which to deal with competitive and non-competitive markets. It was a regular part of school life and had the potential to carry the instructional load. At the script stage, discussions with students led to the incorporation of realistic details which made the program more believable. For instance, they suggested that students were more likely to bring popcorn then sandwiches to a game. Treatment evaluations help the writer exploit the strengths and overcome the weaknesses of an idea to produce a good script. The responses from teachers and students can confirm successes, identify problems, and suggest alternatives. Scripts By their greater richness, scripts offer more possibilities for evaluation and for change based on audience reaction. In fact, evaluation can often help a writer write a script. For a program on failure and self-concept for children eleven to thirteen years of age, an evaluator had about 200 students complete a rather open-ended questionnaire about what they did when experiencing failure and than write a story (fiction or fact) about a person going through a failure or disappointment and how they felt. Most of the students quickly completed the questionnaire and also worked on the story with haste. Their efforts were rather predictable. Many students started with the hope of victory and saved the failure for the end, while others started with an apparent failure but ended up winners at the end. One young girl, seemingly more intense than the others, asked to take the paper home to finish it. When she returned to school the next morning, she proudly handed in her paper. It was the story of her tryout to be a cheerleader at school the previous year, her failure, and her disappointment. There was no success, other than the fact that after going through failure she survived with a cheerful outlook and expectations that the next time would be better. It became the basis for the script. The script writer took the idea and molded it to carry the instructional content. The girl was given an honorarium for her idea and served as a consultant as the script was being written. The program was filmed in her school and she tried out for " h e r " part — she was not chosen. She did, however, make the cheerleading squad that year. 296

For another series, the evaluation staff — as the link between writers and the target audience — helped find two groups of students assembled from several area high schools. One was a representative sample of the potential audience. The other had practiced and performed together as an improvisational acting group. The first was used to help develop and confirm script ideas, and test writers' perceptions of what was real for students. The second group was given situations to improvise and thus helped the writers develop ideas and dialogue. These two groups met regularly during the major scripting period and contributed significantly to the scripting process. One of the conclusions reached through these efforts and others at the script stage is that the more common the event or theme, the more likely it is to be salient; it is easier to recognize yourself and your friends than someone or something that is quite different. Familiarity breeds appeal; the novel results in ambiguity and confusion. The script must carry a double load — the story and the instruction. The instructional goals are paramount and take precedence over story-line. If a script seems to be teaching the ideas or seems to have the potential to teach, then effort can be put into improving a weak story-line. However, if the storyline is so strong that it overwhelms the educational points of the program, then substantial changes might be necessary. It is much better to find these things out while working with paper than after the film has been shot and edited. Over the years the script stage has become more important for evaluation because more problems are resolved most efficiently there. As more evidence and ideas accumulate from earlier stages, evaluation can be more precisely targeted; initial answers lead quickly to additional, more carefully framed questions. Production is then guided by the combination of student reactions and reflected knowledge gained from initial evaluations. There are several ways AIT has gone about evaluating scripts. A reading of a narrative based on the script followed by questionnaires and discussion has proven fruitful. Older children can read the script individually, but the range of reading skills within a class militates against this; they can read along with the evaluator, however, and have the script in front of them for discussions. We have tried several techniques for visualizing a script prior to its being made into a film. There is always the assumption that the closer we can get to the look of a finished product, the better the data will be. One attempt was a slide-tape presentation with AIT staff providing a dramatic reading of the script at a recording studio with an organist playing background and transition music. Children of staff members and staff 297

Figure 5: Story board and still taken from the finished program relating to it (ThinkAbout series)

members themselves acted out the parts at an approximation of the exterior and interior locations while a photographer took slides. This proved quite effective, but was relatively expensive; the finished slide-tape presentation looked more like a finished program than a hasty visualization of a script. The product was taken into schools and tested with very positive results. Another try at visualizing a script was the story board. A graphic artist hurriedly sketched 40-60 drawings of the critical events for a 15-minute script. The evaluator could use the story board as illustrations for his or her story presentation in classrooms. Going beyond that, we recorded the story-line narrative while shooting an inexpensive black-and-white videotape of that story board. Thus we could go into the classroom with a television program, although it was a far cry from the finished product. For a series in the visual arts for grades four through seven, the artist/writer/designer created a script by using pictures cut from art magazines and books and sketched in the items he couldn't find. Dialogue and 298

299

narrative were handwritten notes in various locations surrounding the pictures. The script was laid out in a long horizontal scroll. We taped the scroll to the wall (actually, 2-3 walls in a room), at child height, and asked children to talk and walk through the script, recording their comments on tape. As the students elaborated- and made connections among pictures and illustrations, they provided information about their comprehension and of the script ideas and the appeal of certain concepts. The tapes were summarized and the summary provided to the writers for appropriate revisions. The most common approach to script evaluation still remains story-line narrative and discussion. Using such an approach the evaluator can explore some of the subtleties of the content, to see if the ideas are recognized by both students and teachers. Often it is not readily identified in the first script; sometimes the ideas are not communicated, sometimes they are miscommunicated. This last problem is often present where content specialists attempt to communicate definitions. To an expert in the field the terms are practically self-evident; to a classroom teacher, they can be confusing and lack all meaning. In a recent economics series, it became evident at the script stage that the definition of some important term was inadequate to express what the content expert wanted. The problem was that he couldn't get the writer to understand it to the point where she could express it in the script. In addition to content, the script stage permits some study of the nuances of plot, character development, and dialogue. "He wouldn't say (do) that!" — this is a common expression in classroom discussions during script evaluations. It is an opportunity to fine-tune the dialogue and bring in details that make the program appear more realistic. Once production begins, the dialogue will be modified on location to seem more natural, the real objects found in the locations will be used to create realism, but it is still important that these details are incorporated at the script stage. Rough-cut The "rough-cut" is a roughly assembled version of the actual program and usually does not include music, special visual effects, or opening and closing credits. It is a bare bones approximation of the finished program. Many revisions occur during treatment and script development and thus by the rough-cut stage, some programs require little or no revision whatsoever. Other programs field-tested at the rough-cut stage require revisions ranging from very minor to quite extensive. It is not unknown to have an entire program scrapped at this stage. More common revisions are additional voiceovers to emphasize the program's message and increase attention, changes in scenes and dialogue to make certain events and characters more realistic, 300

scenes deleted because they were not appealing, and new scenes produced that provide the target audience with effective models of behavior. The evaluation of rough-cuts usually takes place in classrooms that are representative of the classrooms that will eventually use the programs. There are two goals: one is to explore the program and how to improve it, the other is to examine the lesson, the overall context in which the program is used. An effective lesson can be described as a teacher introducing a program, the class viewing it, and then the teacher using the teacher's guide to conduct a followup discussion of the salient issues. If the evaluation was limited to the program, the information would certainly help the production team make revisions when necessary. But by broadening the evaluation to include an approximation of its eventual use, we gain not only program data, but also suggestions for revising the printed materials that accompany it, and information that can be useful in promotion and utilization activities. Since the rough-cuts have profited from the accumulated data gathered at previous stages of evaluation, certain issues hold less interest than they might have at an earlier stage. There are generally few content comprehension problems, although it is important to make certain that the story-line does not — because of the elaboration that normally takes place in creating moving images from a script — overwhelm the content nor lead the students in unintended directions. By using questionnaires, interviews, or analyses of the post-viewing discussion, evaluators can find out whether the students are left with questions about the story-line or are ready to focus on the content issues. Sometimes the tone of a program will mislead the viewers. In one lesson on how to deal with stress in everyday life, the program took a farcical approach to a rather serious topic and, as a result, the students were unsure of what the program was trying to accomplish. In another program we found a very positive, yet unanticipated, outcome. When asked the purposes of the program, almost half the students thought that, in addition to the indended outcomes, it was promoting remaining in school to complete the degree and not dropping out prior to graduation. Since this was a desirable end, we added what the students saw to the list of outcomes included in the teacher's guide. While comprehension of content issues is less of a problem at the rough-cut stage, comprehension of story-line remains a focus. Ideas that fall logically together in a narrative rendition of a script do not always look the same on film or tape. How a program is shot and edited can change the students' perception of what it is attempting to communicate; the conception of the director is not always the same as the writer's or the designer's. Verifying that students still follow the presentation is thus an important activity at the rough301

cut stage and is undertaken by a variety of methods including questionnaires, interviews, game-like tasks, and observations. What we have found, especially with children younger than 14, is that television techniques used in popular commercial television are more comprehensible than filmic techniques often used by sophisticated film makers. While contemporary film will often move between fantasy and present reality or from the present to either future or past, by intercutting events, television productions will use slow dissolves, filters, or other changes in focus. These television techniques permit a more gradual shift in cognition, help viewers identify that a transition is taking place, and lead them to anticipate the nature of the change. This is not to say that younger children are unable to comprehend changes in the program's time elements or its relationship to present reality, merely that such shifts should be treated with exceptional care, using the anticipatory signals available to the medium and should be used sparingly. With young children, a transition might be made with the actor saying: " I wonder what it might be like if..."; with older viewers, less obvious cues might be used. Comprehension issues also arise when what is visually obvious to the production staff is less so to the naive viewer. Subtlety in presenting important details is often a problem when adults make programs for children. We have found that if significant details are passed over lightly, often students may not pick them up. The camera needs to focus on these details; the audio needs to point them out; they may need to be repeated. This doesn't mean that loud noises or flashing visuals are needed when an important cue is presented, but many programs fail to convey their message and have to add voice-overs or additional footage. The important ideas need to be made evident, but not overdone. In the treatment and script stages we are at least a step away from the reality portrayed in the program. We ask questions about what the program "would be" rather than what it "is". At the rough-cut stage we get a much better idea of appeal, salience, and identification with the characters. At the script or treatment stage, we ask students if they would like a television program made from the story they just heard or read. At the rough-cut stage, we ask how much they actually liked the program they just viewed. Appeal is often difficult to assess, especially at the rough-cut stage. One must use care in isolating the issues contributing to the overall appeal of a program. Are viewers responding to the characters, the issues, the way the story-line developed, or the way it handled the issues? Students seem to identify with characters or social role models who are slightly older and able to do things that the viewer sees as soon within his or her grasp. However, role models who 302

are too competent or bright are not as appealing; their perfection cannot be matched by the viewers. Setting goals or standards well beyond what students can reasonably expect to achieve often results in the role model being perceived as unrealistic. Conversely, students feel themselves superior to characters who illustrate behavior they have already mastered. Programs featuring such characters, viewers feel, are for younger students. Thus, they pay less attention to the information in the program and its implication for them. Age-related behaviors that are obviously well beyond what students can see themselves doing also tend not to be useful. To a twelve-year-old, learning to drive is not an immediate concern and he or she will lose interest in a program featuring that activity. Social role models need to be slightly ahead of the viewers, but at the same time they need not be perfect. It is easier to identify with a character that stumbles over words and cracks in the sidewalk — "That person is more like me." Role models need to do things that students want to do and expect themselves to be doing shortly; but at the same time, characters in the programs can illustrate mistakes that the viewers are likely to make and show that corrections can be made. Through the behaviors of the characters in the program the students learn the patterns of behavior associated with the pogram's goals. Character development is often merely a change in behavior. Related in many ways to students' identification with the characters is the issue of realism. As stated earlier, the realism of the program reflects the plausible, but not necessarily the literal truth. The people need to be people like the ones students know from their experience, participating in actions that are possible in the students' world. It is often more effective to have a highly realistic person in a less realistic situation than the reverse. The importance of the issue covered by the program, or its salience, often affects how the program will be received and understood. Issues of importance to the viewer increase his or her motivation to watch and learn from the program; less important issues produce boredom and reduced educational effectiveness. There is a surprising disparity between what adults think students find important and what students actually say. A program topic such as "privacy" was highly important to eleven- and twelve-year-olds, much to the surprise of the script writer and producer. Similarly, a program on how members of a family change their family roles as children grow up was not as important to high school students as many adults originally expected.

303

This inability to always pedici accurately is one reason why evaluation and field testing are so important to developing useful and appealing programs. Even extensive evaluation may not show a simple relationship among variables. In Table 1 (p. 305), summary data aggregated from several questions on each variable from seven programs in a high school series on social and emotional growth are displayed. Examining the data on appeal, realism, objectives, and importance shows little pattern in the relationship among these variables. For example, the program Accepting Feelings rated higher than Career Aspirations on three of the criteria but the same on appeal. One program factor that cannot be studied with a treatment or script is attention; for that we need a film or television program. Attention to a program, in all its manifestations, is especially important for school television. If a student is not attending to the program, he or she is likely to be doing something else — and that something else is likely to be disruptive to the rest of the class. Television programs for the classroom need to be as attention holding as possible, or teachers will find them difficult to use. Normal classroom patterns are disrupted by bringing a television into the classroom. If further breaks from the instructional norm or with traditional decorum occur while viewing, teachers will relegate the TV sets to the closet. We have collected data on attention to television for many years and they seem to support the hypothesis proposed by Dan Anderson, that comprehensibility of the material leads to attention (cf. Anderson, this volume). In the classroom, watching television is a purposeful activity similar to information gathering. Students attend to materials that further the instructional task in which they are engaged. They ignore material that confuses them or that is inconsistent with their goal. When this is made more difficult by production techniques, attention suffers. Taking a look at attention data gathered for the program Pressure Makes Perfect (described above) from the Self Incorporated series, one can see a relatively high level of attention, reduced at certain points. In this project attention was defined as eyes on the screen as determined by observers present in the classrooms. About 25 classrooms were observed. The opening sequence was somewhat novel and used the techniques of a fisheye lens and rack focus to move in and out of the setting. It was hard for students to identify quickly what was happening so they turned to one another quizzically. It took almost one minute for attention to settle and have the class watching closely. Beginning at about five minutes into the program, a fairly long sequence of classical piano music occurs. It is slow paced, with little camera movement and little reason to attend visually. However, as soon as something happens to further the drama, students return to viewing intently. 304

c t. o

o

r-

σ\

ο α) r tí tí οο ο Ο M

•S ma

g

ε

g: M-

g:

S .h il c§ w

Brt—

a

"

o

7

r-

«ι σ\

(S

U

bj) ω ^Η e £ ο •a » 7 O J3 II υ S c £ w

O tí ce "ce a ce rtí "on O ίS o Β ο 3 C/} •β σ I-H *+-> I—1 Ö D •J α> O ti i I 0) C < — α, w si

tí

cu

Λ«

θ\

^ > O M U

— ι I i> S . o. < D O ¡-H Hi 'cî < * > H J3 & ö U ~ ω « 8'S d> H '-> J7 m < o

e < ·ο S υ Cd Lh Cl tí !? o O o.

305

Attention data of this kind are useful in diagnosing potential problems at the rough-cut stage. By themselves, attention graphs illustrate only part of the viewing process, but when inattention is associated with the lack of learning or with learning the wrong things, it necessarily leads to modifications. Some inattention, for instance, seems naturally associated with transitions from one scene to another. Thus important content material should not follow immediately after a highly emotional scene. Students need a moment to adjust and the editing of the program should take this into account. What happens in the classroom before the program has an important effect on attention? When teachers prepare students to view, they set up expectations for viewing behaviors. If viewing is an assignment, due all the respect of regular classwork, then attention is more likely to be high. If less credibility is given to the program, then less respect will be shown. In preparing school television programs, great effort is placed on seeking and holding viewers' attention. If the class is quiet and attentive, responding appropriately to the events of the drama, then learning from the program is more likely than if these conditions do not hold. Teachers are willing to try television series and continue to use television materials that engage and teach their classes. Attention may be a necessary but not sufficient condition to learning from instructional television programs. One additional aspect of rough-cuts amenable to evaluation is a concern important to AIT and some other producers of instructional materials. Projects developed by a consortium of state, provincial, and local educational agencies in the United States and Canada have a responsibility to represent the great diversity present across these countries. Issues such as diversity of geography and climate, urban and rural environments, the range of socioeconomic settings and racial and sexual demographics are all important. Furthermore, AIT believes it has a responsibility to take each opportunity to illustrate prosocial content and role models. For the ThinkAbout series, an exceptional school television undertaking, evaluators conducted a close examination of the rough-cuts to report roles, values, demographic balance, using a content analytic framework. Used as formative evaluation meant that the content analysis was pro-active, not reactive; groups of five to ten programs were examined and the data reported in a cumulative fashion during the production. ThinkAbout's extensive length (60 programs) made this systematic formative evaluation approach possible. Thus, changes could be made in casting, in the assignment of scripts to various production houses, in the tone and behavior of the characters in the scripts. 306

Figure 6: Attention profile; program Pressure Makes Perfect from the Self Incorporated series (Source: Rockman and Auh, 1976) Attention Percentages 85 90 95

100

I I ι ι ι I I I I ι Fuziy picture. Piano playing Focused picture. Nan playing piano Fantasy of breaking piano and hands Talking with girlfriend after school Playing the recital piece on record Talking. Arguing with mom in kitchen

Practicing on piano Teacher talking in front of classroom Talking with teacher about piano teacher Fantasy of cockroach on piano keys

Practicing piano with re«chcr behind

Mother intruding during piano lesson Nan crying to repeat, unsuccessfully

Nan's dream of blanket recital Heating water. Practicing piano Teacher talking before dismissal After school. Talking to girlfriend "

) Bumming around downtown with friends

Entering house arguing with Mom, Dad

Recital by young boy Nan was introduced. Approaches piano Nan hesitates. Fantasizes. Then plays Nan walks out. Teacher calls on next Closing credit. Production agency, logo

307

The need for content analysis was to obtain a balance that went beyond "not getting into trouble." We saw a potential to increase positive incidental learning from the television images. By seeing the role models act in ways consistent with the best of society's expectations, perhaps viewers would begin to adopt those behavior patterns. By showing a reasonable degree of risk taking, perhaps we might moderate the behavior of some while stimulating risk in others. We hoped to encourage values sought by educators and reduce the presence of inconsistent values in the programs. The outcome of this evaluation is only in the form of a description of the series' content. Assessing long-range impact in this area is a tenuous endeavor. Nevertheless, it is a start in providing prosocial materials for the school.

Implications for production Developers of children's television programs for use in schools need to maintain a perspective on what they do. Most important is recognition that the programs will be used in the context of a classroom. The teacher is present, the environment is somewhat controlled, conditions for learning can easily be put in place. The producer needs to take advantage of this opportunity. The starting point is the instructional design of the program and the refinement of content and production factors over the entire developmental period. The research data collected at AIT over the past 10 or so years, suggest many specific issues for production, but much depends on the specifics of the program being developed. The AIT consortium projects have emphasized the dramtic narrative format, a highly appealing format for children. Within that apprach to programming, concerns for realism, appropriate social models, and the salience of the issues help make lessons that attract and hold the viewer and lead to learning. Methodological issues in evaluation and production are focused on the means to improve programs as they are being created. The emphasis I have described is to conduct much normative evaluation on treatments or scripts rather than waiting until film or tape have been shot and edited. The approach produces savings that permit more evaluation and further revisions. In addition to traditional quantitative approaches to tests, questionnaires, and attention graphs, many insights are derived from interviews and observations — more qualitative methods. And most illuminating have been case studies of how a television series is adopted into the ongoing activities of a classroom. 308

The role of evaluation is integral to creating productions that work in classrooms. No matter how successful a producer has been in predicting the responses of the student viewer, perfect prediction is unlikely. Using data gathered independently to help make decisions has a proven track record and is worth emulating. Funds set aside for evaluation and revision pay enormous dividends in production quality and classroom impact. The experience of AIT and of other producers supports this approach. High quality school television demands it.

References Agency for Instructional Television On the Level final report on formative evaluation (Research Report No. 77). Bloomington, Ind.: Agency for Instructional Television, 1980. Agency for Instructional Television. ThinkAbout content analysis: Final report. Bloomington, Ind.: Agency for Instructional Television, 1982. Cohn, M.R. Teacher use and student response in three classrooms. In J.R. Sanders and S.R. Sonnad (Eds.), Research on the introduction, use, and impact of the ThinkAbout instructional television series (Vol. II). Bloomington, Ind.: Agency for Instructional Television, 1982. Middleton, J. Cooperative school television and educational change. Bloomington, Ind.: Agency for Instructional Television, 1979. Rockman, S., and Auh, T. Formative evaluation of Self Incorporated programs. (Research Report Nr. 30). Bloomington, Ind.: Agency for Instrucational Television, 1976. Sanders, J.R., and Sonnad, S.R. (Eds.) Research on the introduction, use, and impact of the ThinkAbout instructional television series (5 Vols.). Bloomington, Ind.: Agency for Instructional Television, 1982. Sloan, K.R. Thinking through television. Bloomington, Ind.: Agency for Instructional Television, 1980.

309

Konrad J. Burdach

Methodological aspects of formative research

Introductory remarks The concept of "formative research" is closely connected both with the children's television programmes conceived and produced by the Children's Television Workshop at the end of the 60's, which enjoyed world-wide success — Sesame Street in particular and later The Electric Company — and with the names of those who contributed essentially to the educational concept of these series and to the formative preproductional research: names such as Edward L. Palmer, Gerald S. Lesser and Barbara F. Reeves (cf., Palmer, 1974; Lesser, 1975; Reeves, 1970). In subsequent years this new research concept has been applied in a growing number of research projects (e.g., Flaugher and Knapp, 1974; Graves and Nelson-Shapiro, 1975; Laosa, 1974; LaRose, 1979; Rockman, 1976; Rust, 1974; Shapiro, 1975). Today it has therefore become evident that formative research — formative evaluation, preproduction research or whatever this new type of research may be called (cf. Mielke, 1981; Crane, 1980), in any case, research which is typically directed towards optimizing television programmes during or even before the commencement of production — is now considered a promising direction in television research which occupies a position in empirical communication research that should not be underestimated. Moreover, the increasing interest in modes and forms of presentation (e.g. Rice et al., this volume; Salomon, 1979) rather than in the television programme itself reflects this shift of perspective, showing a different view of the phenomena which has much more in common with the focus introduced by formative research than with the "classical" search for television effects. Looking over the recipient's shoulder at the television programme as it is presented and at the subsequent behaviour of the viewer is no longer the only matter which is of interest for social-science research into television; instead, what attracts the attention of the scientists is the producer and his various possibilities for realizing a given content — the presentational modes and formats of television. When Rice et al. (1982) discuss the possibility that one of the main reasons for violent TV programmes causing aggression may be that the violent content is generally linked with the presentational form "action", the shift of scientific emphasis from contents to forms of television is very clearly illustrated. 310

Despite the fact that formative research is apparently of practical as well as of conceptional relevance, there has been — with one exception (LaRose, 1980) — very little critical assessment of a comprehensive nature concerning this new approach within communication research and above all concerning the methodological aspects. Most probably, part of the reason for this is the fact that the findings of the relevant projects have not been sufficiently published and are therefore not readily accessible. In this respect, formative research is quite different from other types of research whose authors generally endeavour to have their findings published, circulated and discussed. It seems in fact that formative research is so radically oriented towards practical application, i.e. towards the planning and production of a special TV programme, that the tasks of the participating researchers are terminated once the programme has been produced. They step back behind the programme. The fact, on the one hand, that the findings of formative research often fail to gain the attention of the experts, because they are not sufficiently available in published form, and the exclusive orientation towards the production of a specific media programme, might lead to the assumption that formative research is possibly an ad-hoc type of research, hastily conceived, working with insufficient methodological effort, and that its main purpose may be to provide certain especially well-funded educational programmes with a kind of "quality stamp", in order to improve their starting position in any discussions concerning quality and educational effectiveness. On the other hand, formative research can be regarded as the very model of succesful socialscience research because it is a truly remarkable combination of research and practical application, melting together programme production and empirical evaluation, and is therefore not faced by many of the usual problems of empirical research, such as that of transferring scientific findings into practice, or of applying laboratory results to real-life situations. It is probably impossible at this stage to pass a final verdict on the question of whether formative research should be considered to be a transitory concomitant of the educational euphoria of the early 70's or rather a promising innovation of empirical research. However, a discussion of methodological problems and. difficulties often confronting formative research may be a first step towards a comprehensive evaluation of this approach. The following questions are of particular interest in this context: Should formative research be classified as social-science research in the conventional sense, or is it something different? Can we apply the "classical" criteria (e.g. reliability, validity), which are normally used for the evaluation of empirical 311

research, in the same way to formative research, or do we have to use different criteria? What claims does formative research make in regard to the reliability, validity and generalizability of its findings, and does the methodological quality of its finding satisfy these claims? In this paper, the first topic will be the discussion of the typical characteristics of formative research. We try then to give a brief description and methodological evaluation of three formative research projects which were carried out by or for the Children's Television Workshop, reflecting various stages of formative research during the last 15 years. Finally, we will try to provide answers to the questions mentioned above and will draw attention to some problems which should be taken into account when future studies are being planned.

Formative, summative and basic research What are the differences between formative studies and other empirical studies in the field of television research? Let us first make clear the way in which formative research proceeds, taking a fictitious example, namely of a formative study about the best way of teaching preschool viewers the numbers 1 to 20, using an instructional TV programme. As there are a number of different possibilities for realizing this instructional goal in the programme — e.g. embedding the teaching material in dramatic scenes, using song-games or presenting it via a conventional teaching situation — the producer must first of all decide in what form the teaching material should be presented. Now, this decision is not based on the judgment and opinion of the producing team but on empirical studies whose purpose is to ascertain what forms of presentation are best suited for teaching the subject matter in question. Certain criteria have to be laid down for this purpose, which should be closely related to the suitability of the programme, on the one hand, and be accessible to empirical measurement, on the other. In our example, criteria such as the entertainment value of a presentational format, the attention gained as well as the recipients' learning progress in naming numbers, are employed. The entertainment value is operationalized for the empirical evaluation by counting the number of "laughs" per time unit, attention by the relative amount of viewing time related to programme duration and, finally the learning effect by the difference between a numerical achievement test before and after the presentation of the programme. Test programmes are then produced for all three presentational forms, they are shown to several groups of children, and the above-mentioned variables are measured. One possible result might be that the song-game form is the most suitable for teaching 312

numerical ability. The production decision would thus be made in favour of the song-game. If we compare this fictitious formative research study with any study from the field of basic TV research, e.g. one of the well-known experiments about the correlation between violent TV programmes and subsequent aggressive behaviour, the differences between formative and basic research become very clear: whereas basic research is mainly concerned with a problem which has to be operationalized and made verifiable by the formulation of hypotheses, formative research is concerned with a product, or rather the best way of presenting such a product. The next question is what kind of criteria are applied for this optimization. In our example, the improved level of numerical ability would be taken as the main criterion, accompanied by others which are possibly related to media-learning processes, too, such as attention to the medium or the entertainment value of the programme being shown. Generally speaking, it can be said that formative research is usually concerned with questions connected either directly or indirectly with the — mostly instructional — objectives of a television programme. Or, to put it more precisely, formative research is intended to provide the producer as early as possible with information as to whether specific elements of the programme are effective or not within the framework of the given objectives, so that even at this stage any programme segments which prove ineffective can be replaced by more effective ones. Formative research, therefore, is mainly concerned with the degree of efficiency of a TV programme, which is, however, also the focus of summative research. What is the difference between the two approaches? Whereas summative research examines the intended programme effects after completion of the programme, it is a characteristic of formative research that it starts working before and during the development of the programme, so that its findings can influence production decisions. From the perspective of the producer, summative research, as portrayed by Mielke in a very vivid manner (Mielke, this volume), can indeed be compared with an autopsy, in its diagnostic function — the patient is beyond help, the programme can no longer be modified — whereas formative research is "therapy"-oriented, aiming to make changes and improvements to the programme. Can basic, formative and summative television research complement one another, or are they all completely autonomous approaches, with only one of them at a time being best suited to solve the given problem, depending on the type of question involved? The fictitious example quoted above of formative research indicates that, as far as the conception, production and evaluation of preschool programmes 313

are concerned, there is certainly a methodological context within which all three methods can play a role: the conceptional stage of such a project is influenced by the application of existing findings of basic research, for example from educational psychology. In our example the type of problems which basic research findings could help to solve would be those involving the postulating of the learning objective (improved numeracy for instance), or establishing the degree of correspondence between instructional goals and recipient characteristics (tailoring the material to be taught to the abilities of the recipients). At this stage of the project existing research findings generally have to be relied upon, because it is generally not possible to engage in special empirical research in order to clarify the many questions of this production phase. Once the conceptional framework has been set up in this manner, certain problems regarding the adaptation of the audiovisual material to the given objectives, on the one hand, and tó the recipients, on the other, are handled during the production phase using social-science methods: this is the domain of formative research. Finally, during and after the presentation of the programme, the question of whether the intended effects — improved numerical ability in our example — have been achieved is investigated by means of summative research methods, e.g. by comparing the test results (naming of numerals, solving of simple sums) of the "viewers" and "nonviewers" respectively. Basic, formative and summative research are thus methods which are not mutually exclusive but can be employed in conjunction with one another in certain problem areas, each of them nevertheless retaining its full autonomy and functional capacity. Despite the differences between these approaches, they therefore have one characteristic in common: they are all concerned with the same problem context, within which each of them assumes a specific function. Does the fact that they belong to the same problem context also extend to the quality criteria of the three approaches which this discussion is attempting to differentiate? Or, to put it in another way: are the same evaluative criteria to be used for evaluating basic, formative and summative research? Let us begin with basic research: for the assessment of the methodological quality of studies belonging to this field, the "classical" criteria of reliability, validity, generalizability and relevance are usually applied. Reliability means that the results of an empirical study are so stable that they can be reproduced by other researchers under different research conditions (e.g. using different test persons or different stimuli). Validity requires that the findings must indeed measure what they claim to measure, that they are of conceptual value 314

in the sense that they confirm or falsify prior results, that they are linked with previous theories and help to form new ones. The validity of much basic research work on the correlation between violence, as presented via the media, and aggressive behaviour, for instance, has frequently been questioned because it involves the measuring of an individual's aggression under laboratory conditions, which leads to doubts as to whether the behaviour recorded really has anything to do with "real" aggression, or with aggression as understood by social psychologists (cf., e.g., Bandura, Ross, and Ross, 1961).

Generalizability, the third of the above-mentioned criteria, refers to the transferability from the special conditions of a study to the general conditions of a real-life context. It concerns the question of whether the results also extend to areas not tested in the study itself (e.g. different age groups, different fields of stimuli). The relevance criterion, finally, is mainly to be applied to studies that are actually oriented towards practical application (that means, not necessarily for all basic studies), indicating the probability that the research findings will lead to a change or improvement in actual practice. Now, are these criteria, which are normally used for the assessment of basic research work, valid in the same way for the evaluation of formative research? The reliability, or reproducibility, of research findings might appear to be of slight interest when formative issues are at stake, as the main objective here is to achieve programme effectiveness and appeal, whilst the reliability of the results seems to be of more theoretical relevance. As we will see later on, formative researchers are indeed often willing to tolerate results of minor reliability, as long as there is enough likelihood of "programme success" in the above-mentioned sense. This may be regarded as a kind of contradiction: on the one hand, production decisions are based on the findings of empirical studies, while, on the other hand, a low degree of reliability is often tolerated, which can only mean that there is a willingness to make production decisions of far-reaching significance on the basis of, in the worst case, dubious and uncertain information. Reliability is thus an essential, though frequently underestimated, quality criterion of formative research. The validity criterion normally does not involve any problems for studies concerned with formative issues, because the intended validity area, i.e. the programme to be produced, is limited and largely coincides with the field of investigation, which is the assessment of the effectiveness of programmes or programme segments. In the example we gave at the beginning of this discussion, for instance, it was not intended to make general statements concerning, for example, the connection between intelligence and numerical ability on the basis of a single experiment: this would be the task of basic 315

research. The intention was to test the effectiveness of various programme segments (field of investigation), in order to build up a programme (area of application). Validity problems can, of course, arise with respect to the recording of "indirect" objectives (LaRose, 1980, p. 280 ff.), if such an anticipated control of programme effectiveness is not possible. This may be the case, for instance, if the given objectives relate to complex social behaviour characteristics that are difficult to measure empirically, so that the formative evaluation has to be content with measuring behavioural characteristics which are only indirectly connected with the actual learning objectives, such as attention and comprehension. Formative research is at any rate nothing but an anticipatory assessment of the effectiveness of the programme, the success or failure of which is revealed in the summative check on the effects. This means that the findings of the summative research represent an important criterion of the validity of the formative measurements. The criterion of relevance is apparently also relatively unproblematic, as the area of investigation and the area of application overlap to a great extent, thus providing direct practical relevance. Similarly there would appear to be few problems regarding the generalizability of the findings of formative research, as generalization only takes place from programme segments to the whole programme and not — as in many questions of basic research — from one stimulus to the universe of all possible stimuli relevant to the question concerned. On closer scrutiny, however, problems do crop up as regards generalization from the recipient group investigated to the entire audience of the finished programme. This type of generalization from the sample to the population is usually also aimed at in basic research investigations, it is true, but — unlike the formative research situation — without the population actually being confronted with the stimulus after completion of the investigation, and the results or predictions of the investigation — e.g. via viewing scores or the results of the summative research — thus being subjected to a "generalization test", as it were. Other aspects of generalization — such as generalization from the test situation to the real reception situation, from individual prototype programme segments to the whole programme or, in the case of a series, to all the shows in the series — are not without their problems, either. The generalizability would there-fore appear to be a quality criterion of particular importance for questions dealt with by formative research: it is only possible to develop a successful programme on the basis of formative research findings if the conditions under which the formative research is carried out and the reception reality largely correspond, which means that conclusions can be drawn about actual reception behaviour from the research findings. 316

To sum up: the main difference between formative research and basic research is that the former is not primarily concerned with a given problem, but with adapting a product to given objectives. These objectives usually relate to the learning effect which the programme is intended to achieve amongst viewers. The criteria of success are therefore: the appeal of the programme, i.e. its chance of actually reaching a comparatively large number of viewers, and the effectiveness of the programme, i.e. the probability of its achieving learning success in the intended direction amongst the viewers who are reached and addressed by the programme. Formative research aims at improving the programme at the developmental stage; summative research at compiling the effects of the programme during and after transmission of the completed programme. Basic research and the formative and summative research approaches certainly relate to different aspects of a TV programme, but they complement one another. The "classical" quality criteria for evaluating empirical research seem to be less important for formative research, as apparently it is the "success" of the programme under investigation which alone determines the quality of the formative research. On closer examination, however, it becomes clear that it is above all the reliability and generalizability of the research findings which determine the predictive quality of the formative research in the anticipatory assessment of the programme's success and which must therefore be regarded as important quality criteria.

Formative research in practice: Three examples Having discussed some theoretical aspects, we will now take a closer look at formative research in practice. It is, of course, the methodological questions which interest us here, for example how the appeal and effectiveness of a programme can be operationalized empirically, whether the research methods used (test procedure, samples, experimental conditions etc.) are suitable for examining the specific hypotheses in question, and to what extent the conclusions drawn from the research findings regarding the development of the programme are justified. From existing formative research studies we have picked out three investigations which are particularly suitable for showing some of the typical problems which face formative research as regards methods (Reeves, 1970; Shapiro, 1975; Mielke, 1981) and the development of the formative approach during the last 15 years. Space will not permit us here to describe these large studies in detail, however, so we will have to confine ourselves to discussing the essential methodological aspects.

317

The first year of Sesame Street: The formative

research (Reeves, 1970)

The research reported on here began in 1968, eighteen months before the regular broadcasting of the first series of Sesame Street, and continued until well into 1970, till the end of the first broadcasting period. It is composed of very heterogeneous individual investigations, which diverge widely, for example, as regards the questions, the quality of the execution of the tests and their validity, and which give some idea of the creativeness and pioneering spirit — but maybe also the hectic activity — which apparently characterized the pre-production phase of this mammoth undertaking. The questions examined in the investigations reflect the information needed in each phase of production and can therefore be dealt with in corresponding order. Testing for competence After producers and scientists had agreed upon an extensive catalogue of learning objectives (.Reeves, 1970; pp. 44-51), the first questions put to the Research Department by the producers related to the abilities and knowledge which the target group would already possess in the individual learning target areas. The idea behind this was that in order to develop programmes intended to promote children in a particular age group, in this case between 4 and 6 years, in various areas, it was first necessary to ascertain the level of knowledge and achievement of the age group in question in these areas (Reeves, 1970; p. 10 f.) It was, however, almost impossible to find this type of basic information, for example about the children's literacy and numeracy, in the existing literature. In order to obtain the relevant information, special studies had to be designed and carried out. The first study examined 68 four-year-olds in a New York day nursery to establish their knowledge of the alphabet. This study investigated among other things whether the children were in a position to recite the alphabet or parts of it, to name letters and to write their own names. The second study was concerned with the children's numeracy. The children in the test group were asked to count as far as they could, to name numbers, to ascribe numbers to given objects etc. Finally, the third study was concerned with the formation of concepts and the ability to name parts of the body. This study investigated the extent to which the children were able to recognize common attributes of given objects or differences between them, and to identify various parts of the body correctly. The determinants of appeal The formative research was, secondly, concerned with examining the appeal of existing children's TV programmes and the presentation factors connected 318

with the appeal of the programmes and/or the attention they elicited amongst the children. The primary aim at this stage was to gain acceptance for the series about to be produced in the face of numerous rival shows: "No matter how effective the show would be in teaching the child, it had to reach him first" (Reeves, 1970; p. 12). At the time in question — the end of the Sixties — programme appeal and the attention elicited in child viewers were criteria which had not as yet been satisfactorily operationalized, i.e. were scarcely accessible to empirical research. One of the great achievements of the formative research for Sesame Street is that it devised a method for measuring attention and programme appeal in a quasi-experimental test procedure. This method has gone down in the literature as the distractor method. Here the viewer is confronted with two different visual or audio-visual presentations competing for his attention: one is the test programme whose appeal is to be determined, shown on a TV monitor, the other is a series of slides projected by a slide projector in a certain time sequence (e.g. every 8 seconds) onto a screen next to the TV set. Using a technical device, an observer registers how long the viewer's attention is distracted from the TV programme, i.e. how long he allows his eyes to wander from the TV screen to the projected slides. If this distraction time is related to the total length of the test programme, one obtains a standardized index of a programme's appeal. The distractor method was used to assess the appeal of existing children's programmes already broadcast. The research team wanted in particular to establish a link between the main characteristics regarding forms and formats of the programme — e.g. the use of animated cartoons or puppets"- and the programme appeal and attention level amongst the recipients. Forms of presentation which proved to be attractive were to be adopted in the Sesame Street production. A total of 29 different children's programmes were shown to small groups of children and the attention level for each programme determined in the manner described above. Special conclusions were drawn from this registration of attention values of the children's programmes (Reeves, 1970; pp. 16-20), some of which were later taken into account in the production of Sesame Street, e.g. that animated cartoons achieved particularly high attention values, that dialogue between adults holds little attraction, or that children are chiefly fascinated by child performers and animals. A further series of investigations intended to provide information about the connection between programme characteristics and the attention level of the child recipients were the "small group observations". Here again various programmes were shown to small groups of four-year-olds, but this time a note was made not only of the attention elicited in the children but of all behavioural characteristics connected with the programme, e.g. verbal 319

utterances, gestures and facial expression, motor activity and imitative behaviour. Barbara Reeves' report does not indicate, however, whether there was any collection of data at these sessions. She describes only the conclusions (Reeves, 1970; pp. 22-28) which were derived from the observations and conveyed to the production team. These conclusions point out, for example, the importance of the soundtrack or the effectiveness of slapstick elements in gaining children's attention, and establish that children are inclined to imitate certain modes of behaviour seen in the TV programme and that the repetition of programme elements can under certain circumstances heighten the appeal of a programme. Achievement research on individual film

segments

The next phase of the research was concerned with the development of programme elements in the series. The research team wanted to find out what characteristics are required of various programme segments to achieve maximum learning effectiveness. For these tests the film material was varied as the independent experimental variable, while the achievement under the various experimental conditions was investigated in a pretest-posttest design. The criterion of achievement in all these experiments was without exception the children's ability to handle letters. The investigations examined, for example, whether frequent showing of a letter sequence was more effective than less frequent presentation or even one-off sequences, whether massed presentation achieved greater learning progress than intermittent presentation, whether the length of a sequence affected the test achievement, and whether lessons about various letters obstructed or promoted learning progress with regard to an individual letter. Again, no details are given in the report about data-collection, the nature of the measuring instruments or the statistical treatment. Research on five test-shows The last phase of research before the regular broadcasting of the programme was concerned with the learning effect achieved by the entire programme. Five complete one-hour Sesame Street test-shows were presented to various groups of viewers in a field experiment and tested as to their learning effectiveness. This last phase of the research already includes, in addition to features of formative research, components which are more characteristic of summative research. The programme was largely finalized, the fundamental production decisions were already made. The task was not to evaluate programme segments but to check the effectiveness of whole test-shows. On the other hand, it was still possible at this stage to incorporate last-minute changes, in 320

other words there was a formative side to the research in that the research findings could potentially still affect the development of the final programme. The five test-shows were shown to three different groups of four-year-olds in New York and Philadelphia within the space of a fortnight, while two control groups watched different TV programmes. The test and control programmes were broadcast in both cities on different channels. Children were assigned to the experimental or the control group by drawing the parents' attention to the channel and the broadcast time which corresponded to the experimental conditions in question. The samples ranged from 20 to 40 subjects. Before the start and after completion of the fortnight test period, the children's achievement scores were collected in five test areas roughly corresponding to those mentioned in the first section (Testing for Competence). The Sesame Street programme was watched partly in the children's own homes and partly in the nursery school. Barbara Reeves' report does not contain any further precise information about the findings obtained, but in a summary simply states the general conclusions relating to the learning progress said to have been achieved. Formative research during the broadcast period The second part of the report contains a detailed description of investigations carried out during the first period of regular broadcasting. These studies — although referred to as formative research in the report — quite clearly belong to the field of summative research and will therefore not be discussed here. Comments In evaluating the research findings reported on by Barbara Reeves we want to take up our initial reflections and so evaluate the methodological quality of these findings, applying the quality criteria described at the beginning. Contrary to our expectations, we find that a criterion which seemed to us to be unproblematic for the field of formative research turns out to be surprisingly significant for the evaluation of the research findings shown here, viz. the validity criterion. In fact the validity of the research findings shown here proves to be extremely doubtful. A comparison of the achievement areas examined with the explicitly formulated objectives reveals in fact that only a small proportion of the extensive catalogue of objectives is covered by corresponding measurements in the formative research, while a far larger proportion of the objectives is left completely unconsidered. This means that the relevance of these formative research findings clearly has to be assessed as very slight, simply because only a very small proportion of the initially 321

formulated objectives, in particular literacy and numeracy, has been included in the check on the programme's effects. As far as reliability is concerned one has to state that the results do not satisfy even the minimum requirements which should be made of empirical socioscientific investigations, and this is in fact mentioned in the report itself several times (e.g. on p. 29). Frequently the samples used are only very small, with the consequent risk that the claimed "effects", such as the "average attention" elicited by the children's programmes investigated, are influenced most of all by the differences between the individuals taking part in the investigation rather than reflecting the varied appeal of the programmes. The reader of this report learns next to nothing about the data on which the conclusions are supposed to be based, the way they are processed, or about possible statistical treatment. It seems that for a number of investigations no data in the true sense of the word were really collected and that only ad hoc observations were made. Many of the so-called conclusions do not represent the findings of empirical research but are to be considered as the result of subjective interpretation or even speculation. As regards the generalizability of these findings any further discussion is unnecessary: if the research findings in themselves are so unreliable, there can be no question of generalizability, which presupposes reliability. Now it could be objected that Reeves makes no claim at all to be carrying out investigations and experiments in accordance with the rigid demands of socialscience methodology. On the contrary, she clearly concedes certain deficiencies in her methods: "In order to provide information on the many program decisions that had to be made, we were often forced to short-cut traditional research methods. The objective of the research was not rejection or acceptance of a production decision at the .05 level of significance. We wanted to provide information that would increase the probability of a successful decision" (Reeves, 1970; p. 29). Any claim regarding the generalizability of the results beyond the framework of the Sesame Street programme being produced is also expressly rejected: "Most of the research on film segments has little applicability beyond this specific project" (Reeves, 1970; p.30). Palmer also makes a similar statement, pointing out that the interpretation of formative research results often has to go beyond the data: "The formative researcher's work frequently results in statements of principles and guidelines that go well beyond anything which could be decided strictly from the data" (Palmer, 1981; p.237). Is criticism of the formative research portrayed above therefore unjustified, because evaluation criteria are applied that the author herself admits are not fulfilled, and the fulfilment of which she herself does not even regard as 322

necessary? In this connection we should perhaps make clear to ourselves once again what the actual function of formative research is: its purpose is to obtain information that is as "objective" as possible concerning the appeal, comprehension and effectiveness of programmes from empirical investigations, to be used as the basis of the corresponding production decisions. Objectivity in this case is to be achieved above all by not relying on observations and conclusions based on individual intuition, but by employing the methods of empirical research, such as systematic observation, random sampling, application of statistical significance tests. It may be asked, therefore, whether it makes sense to abandon the claim of reliability, thus taking the risk of abandoning the objectivity of the results, too. Despite these methodological inadequacies, the success of the final product "Sesame Street", which was produced to a considerable extent according to the results of the formative research described above, is indisputable. In this case there seems to be a contradiction between, on the one hand, the methodological quality and the validity of the formative research that was carried out, and on the other, the advantages obtained for programme production from these results, despite their unreliability. To formulate it in a rather extreme manner, it looks as if wrong calculations can lead to the right result, and, moreover, even to results that in practice have quite obviously proved their value. Sesame Street's success proves that this is, in fact, possible. What was decisive in this case was surely the fact that researchers and producers had been sensitized to each other's problems, that sufficient time and money were available to discuss in full detail the questions requiring decisions, and, above all, that the participating team was made to observe children watching TV programmes. The conclusions drawn and applied for the programme are therefore based probably more on the personal intuition, sensitiveness and intelligence of those concerned rather than on the interpretation of reliable social-science results. This, and the unusually high amount of creativity and know-how of those who participated in the production of children's programmes may explain why the formative research still led to success in the production of Sesame Street, despite all its methodological inadequacies. Formative research on Sesame Street: Programming affective goals (Shapiro, 1975)

with social and

While in the first formative research work for Sesame Street above all very basic and general questions necessarily received the most attention, such as, for example, the choice of performers (puppets, children), the establishing of the programme structure (magazine format) or the speed of scene sequences, 323

the present study, which was made about five years later, concentrates mainly on special questions regarding the production and complementation of the programme. The research described by Shapiro deals with extending the Sesame Street programme to include elements designed to promote pre-school children's social and affective behaviour. Aims of this kind were, it is true, also originally planned, but then apparently postponed in favour of more cognitive aims. Five years later CTW had developed 11 programme segments for four of these aims, namely, "entering social groups", "coping with failure", "coping with basic emotions", and "sex role stereotyping". These had to be examined to see how far their message would be understood by preschool children and transferred to the conditions of their own lives. Divided up into small groups 73 children between 3 1/2 and 6 1/2 years of age were shown the test programmes as well as control material from Sesame Street series already broadcast. As they watched, the attention of the children was assessed on a five point scale. Then interviews and in part also questionnaires were used to check whether the various segments of the test programme had been understood in the way indended by the producers, and whether the child viewers were able to connect the message of each programme segment with their own experience of life. In addition, each child was asked to establish a preference list of children's programmes. The results contained details for each of the 11 programme segments on the attention obtained, detailed information about the cognitive processing of the segment, especially regarding the consistency or inconsistency of the scene elements, as well as suggestions for the production team with regard to changes and/or additions. Comments Shapiro's research differs from the formative research studies described above mainly because it limits itself to a single group of questions, that is, the effect on the audience of the given 11 test segments. Probably as a result of the methodological difficulties appearing in earlier formative studies, right from the beginning no attempt was made to record the actually intended changes in behaviour, e.g. the overcoming of difficulties when entering a new playgroup. Only certain responses, which are to be regarded as intervening variables, such as attention and comprehension were measured. In this way, better methodological effectiveness, that is, greater reliability, was achieved. On the other hand, no exact information was obtained about the actual behaviour it was intended to achieve, since this would have in fact caused considerable methodological problems in the intended behaviour changes. A further problem of this study that indeed often causes difficulties in formative research is the "non-obtrusiveness" of the test procedure: the 324

behaviour being recorded should be influenced as little as possible by the measurement process. In the case of preschool children this is certainly an especially difficult requirement to fulfil, since measurement procedures, such as questionnaires, tests or interviews, usually demand an intellectual and personal maturity that children of this age rarely possess. The requirement of "non-obtrusiveness" should, however, be taken into account as far as possible in preparing any research design and also in interpreting the results. The attention shown by the children taking part in this study, for example, probably does not correspond to the attention shown by children in a real television situation, since, on the one hand, they were told that the object of the study was to improve the Sesame Street programme, and, on the other, because the children were placed in an artificial, unusual television situation and in addition were observed by an adult. Under these circumstances the results of the preference ratings of the children's programmes — more than 50% of the children said Sesame Street was their favourite programme — are almost worthless: they are an expression of the children's readiness to adapt and to cooperate, but they reveal hardly anything about the programmes really preferred. The same can be said of the way the interviews were carried out: under such conditions, and especially when details relating to the children's own behaviour are concerned, the child must always be expected to say what the adults probably want to hear, rather than express what it really thinks and feels itself. Difficulties of this kind could have possibly been avoided or at least reduced, if there had been no direct reference to Sesame Street in the instructions, if the test programme had been presented in surroundings that were as natural as possible and less recognizable as a test situation, and if the interviews had been carried out in a manner less obvious to the children, for example by their nursery-school teachers, with whom they were already familiar. Making contact: Formative and Chen, 1980)

research in touch with children

(Mielke

The latest formative research project of Children's Television Workshop was concerned with the programme development for the series 3-2-1 Contact. Compared with Sesame Street there are important differences with regard to objectives, content and target audience, of the programme. The series is intended to provide schoolchildren with knowledge of the principles, function and importance of science. The programme aims less at promoting definite intellectual abilities or the development of constructive social behaviour, as was attempted in the first and second phase of the Sesame Street programmes, than at teaching general and specific information relating to science and, possibly, also at changing attitudes towards science in a more positive 325

direction. The target group does not consist of preschool children, but of 8to 12-year-old schoolchildren. The programme is intended both to form part of teaching in the school and also to be viewed "voluntarily" at home, that is to say, to be successful in competition with other TV entertainment programmes. The formative research for this series can be subdivided into four differing phases, like the research for the Sesame Street programme carried out ten years before. The first phase is described as "needs assessment". In contrast to earlier studies, the focus of interest is not only the competence already existing in the target group, but also — and above all — the needs and motivation relating to what it is intended to teach. Phase 2, "format development", is concerned with the appeal of rival programmes already broadcasV and with the development of presentation forms, appropriate to the target group as well as to what is being taught. The test programme produced on the basis of the results obtained here is subject to comprehensive empirical examination in the third phase, in which greatest emphasis is placed on appeal, understandability, ability to compete successfully against rival TV programmes, and the possibility of using the programme in school lessons. During the production of the series, in phase 4, "formative researchers functioned as representatives of the target audience in daily consultations with production and content staff" (Mielke and Chen, 1980; page 10). Comments A detailed evaluation of the very extensive research carried out at great expense on 3-2-1 Contact would go far beyond the scope of this paper. Only some points will therefore be taken up to make clear that the methods of formative research were refined and further developed in this study, that the typical mistakes of earlier formative research studies have in part been avoided, and in general a considerably increased sensivity to methodological problems is to be found. In the first place, the more advanced age of the audience allows the use of structurized paper-and-pencil procedures in assessing the appeal of children's programmes, whereby the children have to decide between three given alternative answers ("yes", " n o " , and "don't know"). This avoids the scarcely assessable influence of the interview situation on the results of programme preference testing. The questionnaires were given to the children by their teachers, so that biases, which occur when the interviewer is a person previously unknown to the child, could be excluded. No mention is made either of Children's Television Workshop as the promoter of the research, to avoid provoking any positive response biases. Further progress compared with the preceding formative research carried out for CTW can be seen in: 326

— the improvement of continuous programme evaluation by allocating positive and negative assessments to two different random samples (a "positive" and a "negative" assessment group); — the discontinuation of the use of descriptions such as "likes" and "dislikes", in which it was always unclear whether the children were referring to the programme content, e.g. to a "bad man" on the screen at the moment, or, as intended, to the appeal of the programme; — the employment of experimental set-ups and electronic equipment that make possible an immediate evaluation of the experiment, and in this way make the results immediately available for production decisions; — the considerable increase in the number of random samples, which usually consist of several thousand children.

Concluding remarks What is now the result of the considerations, partly theoretical, partly practical, regarding the methodological problems of formative research? At first it has become clear that the formative approach differs from the other approaches of television research only with regard to the application of the results obtained, that is to say, to optimize the effect of a television programme, but not with regard to research-specific features. An empirical investigation of the educational effectiveness of a television programme intended to help 4- to 6-year-olds in the difficulty they experience in dealing with numbers, can, for example, be regarded as basic research (research into the connection between age and learning processes), formative research (assessment of the effectiveness of the programme during production) and also summative research (assessment of the effectiveness of the programme after the production has been finished), depending on the time at which the research is carried out and ori what those commissioning the research are interested in. Formative evaluation or research is accordingly less to be regarded as a separate type of research than as a variant in the application of empirical research. Nevertheless, by reason of the special problems and the special circumstances under which formative research is carried out, there are certain methodological peculiarities and risks that must be kept in mind, in particular in planning such studies, but also in interpreting their results. Formative research almost always takes place under considerable pressure of time: since the results have to serve as the basis of production decisions, they must be available when such decisions are made, and are therefore part of the production planning, which is usually extremely rigid. This often leads to the fact that, in order to be able to have the results at the right time, the 327

methodological effort is kept to a minimum, or may even lead to the decision to abandon data collection altogether, as we have seen in the example of the early Sesame Street research. Behind this can be felt, although usually not put into words, the attitude of the practitioner that if he has to convert the empirical information into hard production decisions as quickly as possible, the "whole fuss made about methodology" can be shortened or completely dispensed with, more or less on the principle: "What interests me are not the methods but exclusively the results." This overlooks the fact that results obtained using inadequate methods cannot be regarded as results at all but only as random data, which could turn out in one way, but could also be completely different. Expressed in more scientific terms: If research findings are unreliable, then there is no need to think further about their validity and relevance. Unreliable, that is to say, possibly "wrong", results can have more fatal effects, particularly for the questions asked in formative research with its direct practical relevance for programme production, than perhaps in basic research, in which research producing important findings is usually critically evaluated, if not replicated. If time and circumstances do not permit systematic data collection to be carried out using sufficiently large random samples and adequate statistical significance tests, the paradigm of empirical social-science research should rather be abandoned, and pre-scientific methods of acquiring knowledge should be resorted to, such as individual observation, informal questioning, personal understanding, insight and intuition, which apparently proved their value in the pre-production phase of Sesame Street research. Another danger resulting from the interweaving of research and programme production, typical of the formative approach, could be that the instructional objectives of the programme are adapted to the needs of formative research — not vice-versa — for example, in that particular aims are pursued that are relatively easy to measure empirically. If the Sesame Street programmes were strongly characterized in the first years in which they were broadcast by programme segments promoting reading skills and arithmetical ability, then this may be accounted for probably not least by the fact that the abilities mentioned could be relatively easily measured by corresponding tests and the progress achieved in these fields thus relatively easily proved. All in all, no more should be expected from the results of formative research than this approach is capable of giving. Only in exceptional cases, with welldefined, limited and easily measurable behavioural objectives, will it be possible to measure directly the educational effectiveness of a television programme and to actually maximize it through the employment of forms of presentation that in the course of formative research have proved to be 328

especially effective. In most of the cases, the work of formative research will be mainly to determine intervening variables, such as the appeal and comprehension of the programme, which are to be regarded as important indicators of future programme effectiveness. At any rate, the latest studies carried out by Children's Television Workshop (Mielke and Chen, 1980, 1981) have shown that formative researchers are now very well aware of the specific methodological difficulties of their research approach and have come quite a way along the road towards solving the methodological problems.

References Bandura, Α., Ross, D., and Ross, S. Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 1961, 63, 575-582. Crane, V. Content development for children's television programs. In Edward L. Palmer and Aimee Dorr (Eds.), Children and the faces of television. New York: Academic Press, 1980 Flaugher, R. I., and Knapp, J. Report on evaluation activities of the Bread and Butterflies project Princeton, New Jersey: Educational Testing Services, 1974. Graves, S., and Nelson-Shapiro, B. Formative research on Vegetable Soup. Amherst, Massachusetts, University of Massachusetts.: 1975. Laosa, L. B. Carrascolendas: A formative evaluation. Los Angeles, Cal.: University of California, 1974. (ERIC ED 090 968) LaRose, R. Final report on formative evaluation of Freestyle. Los Angeles, Cal.: University of Southern California, 1979. LaRose, R. Formative evaluation of children's television as mass communication research. In B. Dervin and M.J. Voigt (Eds.), Progress in communication sciences. Vol.2. Norwood, N.J.: Ablex, 1980. Lesser, G. S. Children and television: Lessons from Sesame Street. New York: Random House, 1974. Mielke, K.W., and Chen, M. Formative research in touch with children. CTW International Research Notes, 1980 (3), 9-14. Mielke, K.W. and Chen, M. Children, television and science: An overview of the formative research for 3-2-1 Contact. New York: Children's Television Workshop, 1981. Palmer, E.L. Formative research in the production of television for children. In D. Olson (Ed.), Media and symbols: The forms of expression, communication and education. Chicago: Univ. of Chicago Press, 1974. Palmer, E.L. Shaping persuasive messages with formative research. In R.E. Rice and W.J. Paisley (Eds.), Public communication campaigns. Beverly Hills, Cal.: Sage, 1981. Reeves, B.F. The first year of Sesame Street: The formative research. New York: Children's Television Workshop, 1970. (ERIC ED 047 822) 329

Rice, M.L., Huston, A.C. and Wright, J.C. The forms of television: Effects on children's attention, comprehension, and social behavior. In D.Pearl, L. Bouthilet, and J.Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties. Rockville, Md.: National Institute for Mental Health, 1982. Rockman, S., and Auh, T. Formative evaluation of Self Incorporated programs. Bloomington, Ind.: Agency for Instructional Television, 1976. (ERIC ED 119 712) Rust, L.W., Visual attention to materials in 'The Electric Company'. New York: Children's Television Workshop, 1974. (ERIC ED 122 813) Salomon, G. Interaction of media, cognition, and learning. San Francisco: JosseyBass, 1979. Shapiro, B.N. Comprehension of television programming designed to encourage socially valued behavior in children: formative research on Sesame Street programming with social and affective goals. New York: Children's Television Workshop, 1975. (ERIC ED 122 863)

330

About the contributors DANIEL R. ANDERSON received a Ph. D. from Brown University, Rhode Island, Pa., in 1971. He has conducted numerous studies on the effects of television on children's attention, most of them in connection with Sesame Street programs. He is now Professor of Psychology at the University of Massachusetts, Amherst. Together with J. Bryant he is editor of Children's Understanding of Television: Research on Attention and Comprehension (New York: Academic Press, 1983). KONRAD J. BURDACH holds a Ph. D. and a diploma in psychology from the University of Hamburg. He was research collaborator at the Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen, Munich, and is now lecturing at the Institute of Communication Science at Munich University. His academic interests include media psychology and music psychology. PIERRE CORSET is a collaborator of the research team Recherche prospective at the Institut National de l'Audiovisuel, Paris. He has published various articles on the subject of children and television and is co-author of the book L'Enfant devant la Télévision (with M. Chalvon and M. Souchon; Paris: Casterman, 1979). AIMÉE DORR is Professor of Education at the University of California at Los Angeles. She received a Ph.D. in developmental psychology from Stanford University and then held academic positions at Stanford University, Harvard University and the Annenberg School of Communications at the University of Southern California. She is co-editor, with Edward L. Palmer, of Children and the Faces of Television: Teaching, Violence, Selling (New York: Academic Press, 1980). Her interests are in the ways in which children use, understand and are affected by media. CATHERINE DOUBLEDAY, co-author with Aimée Dorr, is a doctoral candidate in education at the University of California at Los Angeles. Her doctoral studies focus on television's influence on the socialization of norms for affective expression and display. DIANE E. FIELD, co-author with Daniel Anderson, is a graduate student in the Ph.D. program in Developmental Psychology at the University of Massachusetts. She has a B.A. degree from the University of Pennsylvania.

331

LINDA S. HUNTER is a doctoral candidate at the Institute of Communications Research, University of Illinois at Urbana-Champaign. She is coauthor with Ellen Wartella. ALETHA C. HUSTON is Professor of Human Development and Psychology at the University of Kansas. She is Co-Director of the Center for Research on the Influence of Television on Children. She received her Ph.D. from the University of Minnesota and has served on the faculties of Cornell, Temple, and Pennsylvania State University. PETER KOVARIC, co-author with Aimée Dorr, is a doctoral candidate at the Annenberg School of Communications at the University of Southern California. His research interests focus on the ways in which children and their families use and are affected by television and the newer computer technologies. KEITH W. MIELKE is Associate Vice President for Research at the Children's Television Workshop. Previously, he had served on the faculty for 13 years at Indiana University's Institute for Communication Research and the Department of Telecommunications. He has developed a strong interest in the design and evaluation of goal-directed uses of television. Since joining the CTW staff in 1977, Dr. Mielke has been Executive Director for Seasons I and II of 3-2-1 CONTACTa series about science and technology, designed for children 8-12 years of age viewing both in homes and schools. EDWARD L. PALMER has been the director of the Children's Television Workshop's research activities since CTW began in 1968. As Vice President for Research for CTW, he has designed and carried out numerous tests of television's effect on learning, and advised other countries on adaptation of CTW's programs. He has been an associate at the Harvard University Center for Research in Children's Television since 1974, and is currently authoring a study of television for children in four cultures. He received his doctorate in the field of educational measurement and research design from Michigan State University in 1964. In 1974 he received the American Psychological Association's annual award for Distinguished Contribution for Applications in Psychology. Together with Aimée Dorr he published Children and the Faces of Television: Teaching, Violence, Selling (New York: Academic Press, 1980). MABEL RICE joined the Center for Research on the Influence of Television on Children as a research associate in 1978, following the completion of her Ph. D. at the University of Kansas. Her interest in children's television is part of a concern with children's acquisition of communicative skills. Her previous 332

experience includes clinical work with language disordered youngsters. Her other publications are in the area of language and cognitive development. SAUL ROCKMAN is Director of Research for the Agency for Instructional Television. Over the past 15 years he has been designing, supervising and conducting AIT evaluation activities, and he has developed several techniques now in wide use for the evaluation of school television materials. His academic background is in mass communications, social psychology and research methodology. He has published on evaluation, school television, public policy and computer-based instruction. INGEGERD RYDIN holds a degree in psychology from Uppsala University. Since 1969 she has been research associate at the Audience and Programme Research Department of the Swedish Broadcasting Corporation. She has published various articles and research reports on the effects of radio and television on children. Her field of interest is in the effects of special programme features on learning and comprehension. GAVRIEL SALOMON is an Associate Professor of Educational Psychology and Communication at the Hebrew University of Jerusalem. He was written books on Interaction of Media, Cognition and Learning (San Francisco: Jossey-Bass, 1979) and on Communication and Education: Social and Psychological Interactions (Beverly Hills: Sage, 1981). He received his Ph. D. in educational psychology and communication from Stanford University. He ist presently interested in research on children's mental involvement with television and other media. ELLEN WARTELLA, Ph. D., is Research Assistant Professor at the Institute of Communications Research, University of Illinois at UrbanaChampaign. Before this, she held academic positions at Ohio State University and the University of Minnesota. Her research focuses on the application of cognitive development theory to children's perception and comprehension of television programming and advertising. She is co-author of the book How Children Learn To Buy (Beverly Hills: Sage, 1977) and editor of Children Communicating: Media and the Development of Thought, Speech, Understanding (Beverly Hills: Sage, 1979). JOHN C. WRIGHT is Professor of Human Development and Psychology at the University of Kansas. He is Co-Director of the Center for Research on the Influence of Television on Children. He received his Ph. D. from Stanford University and served on the faculty of the University of Minnesota before going to the University of Kansas.

333

COMMUNICATION RESEARCH AND BROADCASTING A publication series Editor: Internationales Zentralinstitut für das Jugend- und Bildungsfernsehen (IZI)

No. 3 Women, Communication,, and Careers Ed. by Marianne Grewe-Partsch and Gertrude J. Robinson 1980. 138 pages. DM 2 8 ISBN 3-598-20202-4

No. 1 School Radio in Europe A documentation with contributions given at the European School Radio Conference Munich 1977 1979. 198 pages. DM 2 8 ISBN 3-598-20200-8

No. 4 Hertha Sturm, Sabine Jörg Information Processing by Young Children Piaget'sTheory of Intellectual Development applied to Radio and Television 1981. 95 pages. DM 2 4 , ISBN 3-598-20204-0

No. 2 Effects and Functions of Television: Children and Adolescents A bibliography of selected research literature 1970—1978 Compiled by Manfred Meyer and Ursula Nissen 1979. 172 pages. DM 2 8 . ISBN 3-598-20201-6

No. 5 Health Education by Television and Radio Contributions to an International Conference with a Selected Bibliography Edited by Manfred Meyer 1981. 476 pages. DM 4 2 ISBN 3-598-20203-2

K*G*Saur Ν Unchen · NewYork »London · Paris K-G-SaurVerlag KG • Postfach 711009-8000 München 71 -Tel.(089) 798901 K-G -Saur Publishing, Inc. -1995 Broadway · NewYork, N.Y. 10023 -Tel. 212 873-2100 K-G-Saur Ltd. • Shropshire House · 2-20 Capper Street · London WC1E6JA-Tel. 01-637-1571 K-G-Saur, Editeur SARL.· 6, rue de la Sorbonne · 75005 Paris - Téléphone 3288601

saur SCHRIFTENREIHE Hrsg.: Internationales Zentralinstitut für das Jugendund Bildungsfernsehen (IZI)

Nr. 11 Schulfunk in Europa Eine Dokumentation mit Beiträgen zur Europäischen Schulfunkkonferenz 1977 Herausgeber: Paul Lohr 1978. 198 S. Brosch. DM 2 4 ISBN 3-598-03561-4

Nr. 12 Hertha Sturm / Sabine Jörg Informationsverarbeitung durch Kinder Piagets Entwicklungstheorie auf Hörfunk und Fernsehen angewandt 1980. 95 S. Brosch. DM 2 4 ISBN 3-598-20752-2

Nr. 13 Michael Schmidbauer Kabelfernsehen in der Bundesrepublik Deutschland: Die Interessen von Wirtschaft, Politik und Publikum 1982. 118 S. Brosch. DM 2 4 , ISBN 3-598-20755-7

Nr. 14 Frauenrollen, Kommunikation und Beruf Beiträge von: Lois Wladis Hoffmann, .Aimee Dorr, Gerald S. Lesser, Ellen Wartella, Gertrude J. Robinson, Marian Jelinek, Joyce Hocker Frost 1983. 178 S. Brosch. DM 3 2 . ISBN 3-598-20754-9

Nr. 15 Gesundheitserziehung in Fernsehen und Hörfunk Beiträge zu einer internationalen Konferenz mit einer annotierten Auswahlbibliographie Herausgeber: Manfred Meyer 1982. 437 S. Brosch. DM 4 8 , ISBN 3-598-20753-0 Nr. 16 Michael Schmidbauer / Paul Lohr Die Kabelpilotprojekte in der Bundesrepublik Deutschland Ein Handbuch 1983. 175 S. Brosch. DM 2 8 , ISBN 3-598-20756-5

K*G*Saur München · New York· London · Paris K-G-SaurVerlag KG · Postfach 711009· 8000 München 71 -Tel.(089) 7989 01 K-G · Saur Publishing, Inc. · 1995 Broadway · New York, N.Y. 10023 -Tel. 212 873-2100 K-G-Saur Ltd. · Shropshire House · 2-20 Capper Street · London WC1E6JA -Tel. 01-637-1571 K-G · Saur, Editeur SARL. · 6,rue de la Sorbonne · 75005 Paris -Téléphone 32886 01