L "1“?4‘5. my: ,5: ‘ .' 9.12.. v .. h.’ 1 5. - “9451.4.“ 359% m \ ‘ (“25:5, 5 ‘ "gt‘f‘fiwi‘ “' ‘3 t .\‘ ‘r 1 ,1 u ,. : 91-3? .. 5953ch 5... W fl m“? #1:.» ‘ " a w. p?- ‘ M 513‘ r‘ 1 ~- ~ mmvs 5 vii}, " ‘2 J i ,x «var-5,: J. ., :23 ., ., V' ' ‘ l 7155.31, 33'}\. ~>:‘.‘L."_.'\-¢:;-..‘xm . .1 wwfxgfilfi+§h “if“ ‘ ’flkéwmfiu‘t w. 7:: M11 ‘xfifihmm ",, “ox“ ‘L flags. : ”1.2.;st : “41‘. 9:755; .5. n. .1: 7” .‘ ' '1). “113251;? 1 .1‘ . 51:12.1": . «- ~:=:::a- « ‘ W . 13:55:33: 111‘ .m , jxliljiiggh ",3ng 9‘1?" 23939.13)" A {Stark ‘ {.5 41l-‘L5'5--‘;1~£§'~CL;3‘ 3.3:: . 51‘ H -x‘.“ 5 1 13'.“ 'x. . ”gaugi‘f‘i-‘fia. .3 “:14. ‘ ‘ $.35. - ”3393§c54$3:$"3“1§rt¢3$¢i-'#kfi‘wgtzfism. ,. ~ ' V '5' 'N ‘7. We}! ‘v- Exam". 1" "' 5 «3" ‘J’PJ .‘. z . w ,L ..x 3. ‘ , ~ x1. - gn. I, p ‘3. a 1 55(49):» 1N1??? gfiégll «:33... it: $1.1: filfl‘kfgs‘: - v." :c 1 . 1 k : 3‘3? 6:. W. 5‘51? 1_ $211353. ‘3 £953}: '63: .15" 1.: , - ~ . ‘ mm , 11:53:43: .1; {51 .L ":i ‘7 a: 'i.‘ Y? ‘§‘ kiwi 2k. . . ‘ \‘u‘. .. ’1‘ s: ‘ 223%31tkv v4 . 1’. ~ ; J» .1 5‘ ”afi-flc‘n's‘qkfigh 7', . . A.) » . , . . .va,M.(-s.’ 1 ,k ‘- I '.-':‘|'1,v§‘_;i Kiwi?“ u?” 1‘. ,7th 5. ‘5 5' " , I" 13:“? «#353,. 5 ‘_ ggij‘a‘. ég‘r’tcfiiafi-x , . H's” “I“ “’ ' 4' : ""ub “1 2 I W‘ 9"» L. 3 ‘7‘ -. v' 3 """."'- ‘ ,» ‘3: 9 " " ’U~ . p ' ‘1 .~; tawny“: 1. , ._ ‘ T: 1“" Ja- ‘, _. U “My“ 7 ‘- “ 1 33?“. «5:, V , a}. . - ,. 1:3 41M” ”W ' ” .- .: Mm 213:"? “'33:: 1.‘ : 5.14“?” ‘95 W315». 'V‘su .. ' ‘W- -U.- -,,, a,_ 11' . ' x ., .‘ h ‘ ’ . : _ .~ 5 93': - 5.1 n. I: . 4» 03:55:55! 3522;“. “EV $33: tulmza‘lébqfi‘ fut. grab-L: $3 . . ' 41‘ $331: figsi’wtm :N‘ o' L ' '3 “' 34M 5 , v, .4. It '5? 7' :- .5 ,9 x, ‘ . “t }" ‘5‘ v1: 9 .. , b‘: ‘9’: «“254? *1? ‘-“ “flay-3* . 512':- ifs}; . .' .. ' 1mm; i;"":*:5m:‘.::‘~u.r ‘ . ‘ 31:99:33.1 Emit: 23?: 4?: M ‘ 'fi w ’9'7‘313’03'muilgfi4’ ‘ w ‘5. '5:§3.‘I"imm‘w‘3 ‘fl.’ ‘ n5} "Awa‘hvcégtgnu {A i. 5... . t5 ' “p‘jijgii‘fi’k ”W {55? ,L 1' V? '1 . ‘§%1§$‘;3 ' 5", :1 “L ' . g.- t . a. " o ax ‘ ' ~ in 5 f3; 5'“. 1 5:5 . 337? 9(5); 1.55! 1 El} ,‘=Iu:».:..‘ é - ‘ M1: ”Rid". ‘~ - j iigiémiagfifi “9 1a a; ,._ In” : .' If” ' Wm fa H‘ . ~ g ,. .’ :5 t has v' («Hi I“ k ‘1; fifiavqha , » ,.‘ 1 htl ~ 9%}: LR”: - w .» ' - v7 '1 . 1 “1 ‘j: ‘ ”5AA fiMfiu; " t‘fi’xhfic’sx ' ‘ 5;? 10‘ $53); ,“ L ‘. .. .4 \n. '. - , . . Jun g1. f_ “1 - 5 '. :5: ~e ‘ “1 figs—35314???“ E " ~ - suifihad L "3’2? W 1"", ., : - H ~‘ I' ‘ 3;}:‘3. . 32%» £11? _, ‘31.. ~ .. ;_.:, " ..,' .5 . .' , g ”:9- 1W‘ i)" P4 “13 ,. 1' . ‘ A Mm... . :32 $59.12;: . d “ 5“}; 3‘ r 10 V ‘ w: !: n 52%;“. : .3 1&2) - in; .7. v-t‘. ‘ 1: ~‘, . A x. .3 :3“) . . I 1' w c .1. 7-23.? 3'3“:- . 3? m 45mg: 31% .- uw ”x5 wér ., 3: :fiigfififiv‘a}: {51:3352; I} .5 ' ‘3'!- - Wig h. fi~gn~i§ . y." . ~W“{% .Ii' . ,; Vi“ v-l -..'s3‘ %< a: A §§iv~ v 45:? 914:2; i run ‘ ) 3957‘551533 ._ ,. ' , {299%}; “35 93$» 5 $13». .L * . an: )n“ 3-. r. 1 1 «5- ‘r', r E ‘1 avg“ .\ "first? ‘ J dim: v- 9 '_,, .. '. 1' VS} 1 . A}? n- r :1 .4 ‘t . 4‘ ,5 :3. ‘fififiifii 33%? .9 7.: .- .n w ‘ 1: r‘“ JV“ 9‘ s...” “H: ”231;. 2;: . ,' .35.; - . 1. . ‘ 5 5 x. n, 5351!. ‘5.“ ‘1. c 3"“. in}: ' .. '2. 3?- ;g» ‘ ,ra . 3-. 3- “n , . a ""x 351‘ épl‘ 4:53.“: 5;"! if" H. v 1‘1""? ,3! u ”n.3- I~ 5 A .... -- . , «a a: 1:: v , . *4 M5 w;- ”Jarg- . .3. 1. ‘ z? .3" a. 9* “ {‘3 333“}; . M; _ . 3, my, 7‘. '3‘” _ ‘. . < k , A .3. mm , ., H“ ,a. ...4 ~; #319 ~ . .8: 1 ym'Ji‘éh". “I AW“ ”3-633 3:; ~ .1: . .. . ' ‘. a" 4's» - M u "5%”? 9 "‘ . .' "Fun-W ,J‘ ”- ~32: v‘ ‘ . 5 ‘ at L Q- » .v: "I ”M. 9mm - r V . .,~ - . n ~ L‘:h'.§fl" w..." “a 4 pin: uz-VIH—lavdi n-~ ~I‘r‘i. ' “mm“xfm . . ' ' HF .ru—rraua "an” n - u : .. fig“? .... . i Illfllllfllllfllfllmlflll 23 10754 3534 Michigan State University This is to certify that the dissertation entitled Relationships Between Adolescent Media Behaviors and Media Behaviors of Parents and Peers presented by James M. Bernstein has been accepted towards fulfillment of the requirements for Ph.D. degree in Mass Media Ema—38M»..— Qor professorv Date August 9, 1985 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU RETURNING MATERIALS: : Place in book drop to. LIBRARIES remove this checkout from -3—_ your record. FINES will be charged if book is returned after the date , stamped below. “54% 17 K313 -E7”KTZ 100 A339 "3W1 1‘33“ 7 7 8 A lit 1 nailzzfinlm z RELATIONSHIPS BETWEEN ADOLESCENT MEDIA BEHAVIORS AND MEDIA BEHAVIORS OF PARENTS AND PEERS By James M. Bernstein A DISSERTATION submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Communication Arts and Sciences 1985 Copyright JAMES M. BERNSTEIN I985 ABSTRACT ' RELATIONSHIPS BETWEEN ADOLESCENT MEDIA BEHAVIORS AND MEDIA BEHAVIORS OF PARENTS AND PEERS By Jam as M. Bernstein The purpose of the research was (I) to discover within a broad framework the existence of an adolescent culture antagonistic toward adult society, (2) to test in the narrower framework of media behavior whether adolescents modeled their parents to a greater extent than they did their peers, and (3) to reconceptualize modeling and test the relationships posited. Previous research testing a media modeling hypothesis used a similarities conceptualization whereby adolescents behaving similarly to others represented modeling. Previous research, however, failed to consider the perceptions of adolescents. Drawing from social learning and co-orientation research, this study required that adolescent modeling of another's media behavior be based on an accurate perception of another's behavior. Based on research suggesting adolescence is a transition period when teenagers remain influenced by parents, it was hypothesized that adolescents would accurately perceive their parents' media behavior to a greater extent than they would their peers, and that adolescents would behave similarly with their parents' media behaviors to a greater extent than their peers. Media use data were collected from adolescent-parent-peer triads. In addition, adolescents indicated their perceptions of parental and peer media use. To test the hypotheses, a significance test for the difference between Jam es M . Bernstein independent correlations was performed. A significant 2 statistic indicated a difference between parents and peers in terms of agreement with and accuracy of their media behaviors. To test whether perceptions intervened between adolescent behavior and the behavior of others, perceptions of others were statistically removed through partial correlation. Reduction of the zero-order correlation between adolescent behavior and parent or peer behavior would indicate perceptions contributed to the relationship. The data showed few differences between agreement with and accuracy of parents and peers. When differences occurred, they usually were in the hypothesized direction. In every instance where the correlations between adolescent behavior and that of parents and peers were statistically significant, perceptions contributed to the relationship. The findings imply that adolescence is a transition period when parents are still influential. And they indicate accurate adolescent perceptions are important in determining how others' behaviors influence adolescent behaviors. ACKNOWLEDGMENTS This is the last entry into the dissertation, so let's not waste any time getting into the thank yous. If I omit anyone, my apologies; I'm on a deadline to get this to a typist, and I've never been presumptuous enough to compile a running list of benefactors. Two members of my original guidance committee, Ray Cullen and Dean Krugman, have my appreciation for having their doors open when I needed feedback about my ideas. The remaining four committee members, of course, have my eternal gratitude. Although my contact with Bixby Cooper was limited during the last year, his support and encouragement during the entire program have been important. Fred Fico deserves special thanks for getting me over a hurdle when he wasn't even a member of my committee. When I needed a replacement, I never had a second thought. As Keith Adler will attest, I had infinite second thoughts about my research throughout the yearlong dissertation process. But the fact that Keith never seemed to waiver helped me persevere. Finally, I owe committee chair Stan Soffin my gratitude, not only for his assistance on the dissertation, but for many opportunities during the last four years. And a couple or three lunches at the University Club. And how about those family members? Lynn knows how much her support meant to me during the last 3”: years, but let's get it on the record. Seth probably doesn't know, but here it is on the record for him, too. ii TABLE OF CONTENTS List of Tables List of Figures CHAPTER I: INTRODUCTION Not es CHAPTER [1: LITERATURE REVIEW Theoretical Background of the Modeling Hypothesis Application of the Modeling Hypothesis The Modeling Hypothesis in Studies of Adolescent Media Use Conceptualization of Media Modeling Behavior Operationalizations of Media Modeling Behavior Measuring Modeling by Adolescents Placing the Communication Behavior in Context Other Potential Models Alternative Methods of Measuring Modeling Adolescent Media Use: Other Frameworks Summary Notes CHAPTER III: METHOD Data Collection Collecting the Data Definitions Adolescent Co-orientation Concepts: Agreement, Congruency, and Accuracy Media Behavior Modeling Data Analysis Notes iii vii ll 14 15 22 22 23 2‘! 25 3O 37 41 47 47 47 51 51 51 53 55 56 60 CHAPTER IV: RESULTS Description of the Samples The Adolescents The Parents Comparison of Sample with Previous Study Comparison of Media Use among Adolescents, Peers and Parents Tests of the Hypotheses Accuracy/ Movie Information Accuracy/ Music Information Accuracy/ Newspa per Reading Accuracy] Tel evision Viewing Agreem ent/ Movie Information Agreem ent/ Music Information Agreem ent/ Newspaper Reading Agreem ent/ Television Viewing Reconceptualization of Modeling Correlations of Media Use for Movie Information-- Adol escents / Par ents Correlations of Media Use for Music Information Adol escents/ Par ents Correlations of Newspaper Use-Adolescents/Parents Correlations of Television Use--Adolescents/ Parents Correlations of Media Use for Movie Information-- Adol escents/ Peers Correlations of Media Use for Music Information Adolescents/ Peers Correlations of Newspaper Use-~Adolescents/ Peers Correlations of Television Use--Adolescents/ Peers Additional Analysis Summary of Results Tests of the Hypotheses Reconceptualization of Modeling Notes CHAPTER V: DISCUSSION Interpretation of Results Implications of the Study Adolescent Culture Versus Adult Culture? Implications for Public Policy Implications for Research Implications f or Management Notes APPEN DICES A. Procedures Prior to Data Collection B. Questionnaires BIBLIOGRAPHY iv 62 62 62 62 68 72 73 74 75 75 76 77 77 79 79 81 82 82 84 34 86 86 90 91 91 92 94 95 95 101 101 101 103 103 105 106 107 132 lo. 11. 12. 13. 14. LIST OF TABLES Adolescent Demographic Information Demographic Characteristics of Parents Factor Loadings with Varimax Rotation Fremency of Media Use for Movie Information Fremency of Media Use for Music Information Frequency of Newspaper Reading Frequency of Television Viewing Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior, Movie Information Comparing the Correlation Between Adolescent Perceptions and ‘ Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior, Music Information Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior, Newspaper Reading Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior, Television Viewing Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior, Movie Information Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior, Music Information Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior, Newspaper Reading 63 64 66 69 70 71 72 74 74 75 76 77 78 78 15. Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior, Television Viewing l6. Product-Moment Correlations Between Adolescent Media Behavior and Parent Media Behavior for Movie Information 17. Product-Moment Correlations Between Adolescent Media Behavior and Parent Media Behavior for Music Information 18. Product-Moment Correlations Between Adolescent Newspaper Reading and Parent Newspaper Reading 19. Product-Moment Correlations Between Adolescent Television Viewing and Parent Television Viewing 20. Product-Moment Correlations Between Adolescent Media Behavior and Peer Media Behavior for Movie Information 21. Product-Moment Correlations Between Adolescent Media Behavior and Peer Media Behavior for Music Information 22. Product-Mom ent Correlations Between Adolescent Newspaper Reading and Peer Newspaper Reading 23. Product-Moment Correlations Between Adolescent Television Viewing and Peer Television Viewing 24. Product-M oment Correlations Between Adolescent Television Viewing and Parent Television Viewing--Multi-Set Homes 25. Correlations Between Adolescents and Parents for Viewing National News and Comedy Shows, by Age 26. Adolescent and Parent Frequency of Watching National News and Comedy Shows, by Age vi l. 2. LIST OF FIGURES Co—orientation concepts: agreements, accuracy and con gruency Modeling illustrating intervention of adolescent perceptions. vii 51 56 CHAPTER I INTRODUCTION Twenty-five years ago the phrase "the generation gap" became popularized; the impetus for its prominence was the Vietnam War. The widely accepted perception was that opinions about the war could be categorized according to age: older people (those over 40 at the time) supported the war and younger people (usually teen-agers and college-aged people) opposed it. Underlying this perception of a schism over a specific policy issue was the general notion that young people comprised their own separate culture, one that not only disagreed with mainstream society about politics, but also about lifestyles. The size of the group of young people also contributed to its aura. These were the people born after World War II, who collectively became known as the "baby boom generation," the largest segment of the American population.l The sheer numbers of this group prompted interest in researching it 20 years ago, when its members were adolescents, an interest that has remained during the past two decades.2 It is unlikely that another generation will be as influential quantitatively and qualitatively and, as a result, be as highly researched. Research of adolescents has remained essential, justified by the importance of knowing about future generations. For example, present-day adolescents, though lacking the numbers of the "baby boomers," certainly yield influence to a degree similar to their predecessors. The idea of an adolescent culture is as strong as it was 25 years ago. Perhaps what makes the study of adolescents in the 19805 important are the intriguing political differences between them and their predecessors. First-time voters in the 1984 presidential election among those 18 to 24 years old supported President Reagan by nearly 3- to-l ratios.3 Experts said the young people of the 19805 expressed traditional values to a greater extent than the adolescents of the 19605.4 And a nationwide survey of high-school students indicated respondents favored restrictions on individual legal rights and religious freedom .5 Given these findings, the youth of the 19805 could certainly be characterized as more conservative than generations in the 19605 and 19705. It is the existence of these distinctions that justify continued research of adolescents and raise many of the questions that were asked 25 years ago: Does an adolescent culture exist that is independent of and, perhaps, antagonistic toward adult society? Or is adolescence merely a transition when peer influence is temporary and parental (adult) influence is still great? What kinds of values are young people bringing to adulthood as they pass through adolescence? These questions provide the broad framework for the present study. The research presented here narrows these questions and considers a specific type of influence--modeling—-in a specific type of context--the use of media. Previous research of adolescent modeling of others' media behaviors has not been developed in the framework of a parental-peer difference, having been limited to modeling of parents. Furthermore, the findings of these studies are inconclusive, in part because the research contains conceptual and methodological problems that will be addressed in this research. Addressing the general questions posed here, the present study compares the influence (in the form of modeling) of parents and peers on adolescent media behavior. In addition to the general research implications dealing with the existence of a division between parental and peer influence, the study has narrow research implications. For public policy, the importance of adolescent media use is twofold. In education, structured media use is already common.6 But increased knowledge of media use at home and elsewhere outside the classroom can help educators more effectively design programs that involve the family and community. In regulatory policy making, the implications of this type of research are well established. Modeling research developed from a concern by parents and consumers rights advocates that excessive television viewing would adversely affect children by fostering aggressive behavior learned from TV violence. Today the concern is over the effects of advertisements on teen-agers,7 but the importance of research of adolescent media use remains. Policy makers need to be aware of the potential influence of parents and peers in the assessment of effects of media on children. For researchers, this study is important in that it reconceptualizes the term modeling and derives new operationalizations from these concepts. The addition of the peer modeling variable, previously overlooked in media use modeling research, creates the potential for developing later research on interpersonal influences on media use. Both of these developments are unique to this study, thereby producing the ability to extend present knowledge about the use of media by adolescents. For management, knowledge of how adolescents use the media and the reasons they use them can be important for two reasons. In the short term, media management can learn about a market with substantial buying power potential.8 Because many teen-agers hold jobs, possess disposable income, and assume responsibility for certain family purchase decisions, they comprise an attractive market for media and their advertisers. The importance of adolescent media use for the long term has to do with the assumption that certain lifelong media-use habits develop during the adolescent period.9 The "fledgling adults" make greater use of the media to make decisions, exposing themselves to a wider variety of media than during pre-adolescent years. Understanding the explanations for media use by teen-agers can help address potential audience problems media companies may experience. Research on adolescent use of the mass media has had two primary focuses during the last two decades, one concerned with theoretical and public policy ramifications, the other with management and marketing. The first focus, usually associated with adolescent use of television, addresses the effect of a medium on the individual.10 The research has typically posed the qJestion of whether television has a negative impact on teen-agers. The second approach poses these questions: What media are teen-agers using? For what reasons are they using them? How and why are they using them? This was typically the focus of researd‘n conducted by the newspaper industry,11 which was concerned about declining circulation. Both types of research have explored the possibilities that adolescent media use can be explained by the modeling of parental media use. In studies dealing with the effects of television on adolescent socialization, the modeling hypothesis implied that certain adolescents could be influenced to watch certain programs because their parents did.12 Similarly, the hypothesis also implied that parental avoidance of certain television programming would result in adolescent avoidance as well.13 Support for the modeling hypothesis meant parents could alter their television viewing patterns in order to induce appropriate viewing behavior in their children.” Research from the newspaper industry indicated that "parental models play powerful roles in shaping children's newspaper reading."15 The implications from the modeling hypothesis in this research is that media are likely to become more attractive to children as parents pay more attention to those media.16 Support for the modeling hypothesis has varied in research of adolescent media use. Researchers have found evidence that adolescents model their parents' behaviors where certain types of family communication patterns exist.l7 Other research, however, suggests alternative explanations for similarities between adolescent and parental media use.18 The lack of stronger support for the modeling hypothesis is partially a conceptual and methodological problem. Conceptually, previous research deals with modeling as a similarity in media behavior. The methodological problems, which come from the conceptualization in the previous research, relate to a failure to consider the issue of whether the adolescent's behavior results from the parents' behavior or from an awareness of the parents' behavior. The present study, through a critical analysis of these concepts and operationalizations, will suggest both conceptual and methodological alternatives in the research of adolescent modeling of parental media behavior. Additionally, this study, broadly framed within the context of parental vs. adolescent influences, will extend the range of models for adolescents by including adolescent peers. Although research in other disciplines deals with the competing influences on adolescent behavior,l9 research on adolescent media use, and specifically modeling, has not. The present study, then, will test modeling explanations for adolescent media consumption. The study will extend previous research in that it will reconceptualize and re-operationalize variables that earlier modeling studies have used. It will also serve as an extension of the earlier studies by testing the influence of both parental and peer models, the rationale for which comes from studies that suggest an adolescent society independent from the adult world. NOTES 1In fact, in his book on the baby boom generation, Landon Jones contended that the generation gap was not so much between the old and the young, but between the many and the few. Landon Jones, Great Expectations: American and the Bab Boom Generation (New York: Coward, McCann and Goeghegan, 1980), pp. 88-91. 2For example, the number of people brought to the 18-24 age group by the baby boom went from 16 million in 1960 to 24 million in 1970, 50-percent increase. U. S. Bureau of the Census, Current Population Reports (Washington: Government Printing Office, 1977). 3"Reagan's Youthful Boomlet," Time, October 8,1984, p. 25. 4"American's Youth in Search of a Cause," U.S. News and World Report, April 16, 1984, pp. 31-32. 5$tanley M. Elam, "Anti-Democratic Attitudes of High School Seniors in the Orwell Year," Phi Delta Kapgn, January 1984, pp. 327-332. 6The best example is the Newspaper-in-Education program sponsored by the American Newspaper Publishers Association. 7Charles Atkin, John Hocking and Martin Block, "Teenage Drinking: Does Advertising Make a Difference?" Journal of Communication 34 (Spring 1984): 157-167. 8Jerald G. Bachman, "Premature Affluence: Do High School Students Earn Too Much? Economic Outlook USA (Summer 1983): 64—67. 9C_hildren and Newspapers (New York: Newspaper Advertising Bureau, 1980), pp. 34-39. 10John P. Murray, Television and Youth: 25 Years of Resggch and Controversy (Boys Town, Neb.: The Boys Center for the Study of Youth Development, 1980), pp. 17-101. The volume contains an extensive bibliography. 11Two examples are Children and NewspaErs (New York: Newspaper Advertising Bureau, 1980) and Mass Media in the Family Settng (New York: Newspaper Advertising Bureau, 1980). 12Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parent- Adolescent Similarities in Television Use," paper presented to the Association for Education in Journalism, Washington, 1970. l3Ibid. 14Ibid., p. 2. l5Chilc1ren and newsEErs, p. xiv. 161bid. 17Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parental Influences on Adolescent Media Use," American Behavioral Scientist l4 (January/February 1971): 323—340. 18Chaffee, McLeod and Atkin, "Parent-Adolescent Similarities," pp. 2-3. 19For example, Bruce J. Biddle, Barbara J. Bank and Marjorie M. Marlin, "Parental and Peer Influence on Adolescents," Social Forces 58 (June 1980): 1057-1079. The article also contains a thorough review of this type of research in various disciplines. CHAPTER II LITERATURE REVIEW This chapter contains a critical analysis of various theoretical frameworks used to study media behaviors of adolescents. Of primary concern is the discussion of previous research that has tested a modeling hypothesis of adolescent media use. That is, the studies under discussion have used the assumptions from Bandura's social learning analysis of observational learning that individuals imitate behavior they observe. This chapter also includes reviews of representative studies in other disciplines that used modeling as an explanation for adolescent behavior and studies that apply frameworks different from modeling to the research of adolescent media use. These frameworks include uses and gratifications, cognitive development, and cumulative acquisition. This chapter also analyzes the concepts and methods used in the previous research of modeling and adolescent media behavior. Suggestions for reconce ptualizati on follow. Theoretical Background of the Modeling Hypothesis The modeling hypothesis, as researchers have named it, states simply that individuals imitate behavior they observe. The hypothesis is derived from Bandura's social learning analysis of observational learning, which assumes: 1. modeling influences produce learning through their informative functions, and 2. observers acquire symbolic representations of modeled activities instead of stimulus-response relationships. This latter assumption contradicted a previously accepted emphasis among social learning theorists that the observer had to receive a reward for learning to take place.1 Instead, Bandura, critical of the one-way influence process implied by the earlier formulations, conceived a "continuous reciprocal interaction between behavior and its controlling conditions," which were personal and environmental factors.2 Within this framework, modeling or observational learning is governed by four interrelated subprocesses: attentional processes, retention processes, motoric reproduction processes, and motivation processes. For individuals to learn and imitate what they have observed, they have to attend to and accurately perceive the essential characteristics of the model's behaviors. Among the many factors that will determine whether one will attend to a model are the characteristics of the model, the characteristics of the observer, and the effectiveness of the behavior being observed. Adolescents regularly associate with a variety of models-parents, peers, other adults with whom they come into contact, and prestige models, such as celebrities seen on television or in films. The behavior evoked from these models would vary depending on the status of the models in the eyes of the observers. Observer characteristics and previous learning would also affect attention to modeled behavior. For example, those who lack confidence or self-esteem would be susceptible to model behavior, as would these who had been rewarded for imitative behavior. In a like manner, the functional value of a model will also be influential in determining attention to models. That is, the effectiveness of a modeled behavior will affect how much one pays attention to it. Another process governing modeling is retention. To reproduce a model's behavior without the model as guide, one must have stored response patterns in long-term memory through the use of symbolic coding. Two systems of representation allow symbolic coding and facilitate observational 1earning--an 10 imaginal representation system and a verbal one. Imaginal representation is important during early childhood when verbal skills are few or non-existent and learning does not lend itself to verbal coding. But adults, too, can retrieve images of people, places, and things to reproduce behaviors. Most of the cognitive processes that regulate behavior, however, are verbal rather than visual. And verbal coding results in longer retention and more accurate reproduction than imagery. Rehearsal also aids the retention process after symbols have been encoded. Motoric reproduction processes involve the observer's ability to translate symbolic representations into overt actions. The ability of the observer to accurately perform the modeled behavior depends on the availability of the component skills and physical limitations of the observers. If these skills are not available, further modeling and practice will be required to develop them. A third impediment to the motoric reproduction processes results because responses to observer actions are not always observed enough to make corrections needed to closely match the model. This is particularly true for skills rewiring physical dexterity, such as swimming and golf. The final factor determining the performance of modeled behavior is motivation. One can acquire, retain, and possess the necessary skills to execute the behavior, but if it is not sanctioned, it will not be performed. Motivation, or the incentive to perform, affects not only the performance of matching behavior, but also the level of observational learning through the control of what people attend to and how actively people code and rehearse observations. Motivation comes in the form of external reinforcement (feedback from others), vicarious reinforcement (consequences received by models), and self- reinforcement (self-1m posed standards). 11 Applications of the Modeling Hypothesis The most frequent applications of the modeling hypothesis can be found in studies of aggressive behavior by children. In a study by Bandura, Ross, and Ross, children who observed aggression by a model subsequently displayed more aggressive behavior than children who were non-aggressive.3 Another study by the same researchers found an increase in aggressive behavior regardless of whether the modelrwas a real adult, an adult on film, or an adult in a cartoon.4 The studies by Bandura and his colleagues have consistently found support for modeling. Their explanation of modeling as attributable to two processes, however, raises some issues. The first process deals with the child's exhibiting novel behavior by imitating a model's aggressiveness. The second process is a disinhibiting one. That is, the child's observation of a model's aggressiveness lessens the child's inhibitions and increases the likelihood that the child will perform an aggressive act already in his repertoire. As Zigler and Child point out, the explanation provided by the second process raises the questions of where, how, and to what extent the child learned aggressive behavior.5 Modeling has also been tested in studies other than those involving aggressive behavior, many of them recent studies assessing the influence of "parents and peers as models for adolescents. Many of these studies also assess the influence of normative standards established by parents and peers. The studies focus primarily on illicit adolescent behavior, such as drug use, alcohol use, and smoking, although the influence of parents and peers on academic achievement and aspirations has also been a research topic. While results of these studies vary, it is apparent that the relative influence of parents and peers depends on the topic.6 Concerned with the fact that earlier studies had been inconclusive about the origins of influence, Biddle, Banks, and Martin established as one of their research objectives a desire to discover whether others' influence depended on the content of behavior considered.7 Their results indicated that the influence on adolescent behavior was content related. For example, their data showed greater parental influence on school achievement and greater peer influence, particularly modeling, on drinking.8 In another study of the relative influence of peers and parents on adolescent consumption of alcoholic beverages, Forslund and Gustafson contrasted peers and parents to determine which reference group was more influential.9 Although Biddle et a1. categorize this study as a modeling study,10 in fact it deals with imitative behavior by adolescents only in terms of parents. The researchers did not test modeling of peer behavior, but rather tested peer influence by measuring the amount of pressure adolescents perceive getting from their peers. Although the researchers use two different ways to conceptualize influence, they were more consistent in their measurement of the concepts. In both instances, they measured the adolescents' perceptions of peer and parental influence. Herriott also used adolescent perceptions to determine parental and peer influences of adolescent educational aspirations.ll In this case, teens were asked their perceptions of educational aspirations held by 11 types of persons. The researchers then used these perceptions as sources of data on parental and peer influence. Other researchers of peer and parental influence on adolescent behavior find the use of perceptions inadequate and potentially misleading. Associations based on perceptions may be inflated and the adolescent's own patterns of drug use may determine his1 Zperception of drug use by others around him, whether peer or parent. In fact, Kandel's data showed parental influence on marijuana-smoking exaggerated when based on the child's perception of parental use of drugs. No data were presented dealing with the possible inflation of peer influence because adolescents were not asked their perceptions of peer drug use. The study did indicate, however, that peer drug use (measured by a self-report from the peers) was of greater importance than parental drug use (also a self-report). Kandel's findings on perceptual data, though well taken, should not prevent the consideration of adolescent perceptions of other behaviors. First, as will be discussed later, using associations of self-reports of adolescents and parents leaves room for several alternative explanations of certain behaviors, particularly media behaviors. Furthermore, the issue of the content of the modeled behavior must be raised. It is possible that the adolescents had inaccurate perceptions because of the illicit nature of drug use, even those legally prescribed psychotropic drugs used for legitimate psychiatric and medical purposes. In other words, parental use of drugs may be a more difficult behavior to observe than other behaviors, such as media use. This would, therefore, make difficult the adolescent observer's ability to make accurate perceptions of his parent's drug use. Unlike the aforementioned studies, recent research in consumer socialization has been more explicitly guided by social learning theory. A review of violence and unsafe acts depicted in children-directed television commercials found that television may induce previously learned aggressive behavior, but only if the child encountered a situation in which aggression was an appropriate response. Otherwise, the study concluded, little evidence existed that children directly imitated television behavior.13 Atkin's study of children's observational learning from television commercials found that most children-directed advertising facilitated response rather than producing new behaviors. For example, a weak relationship was found between exposure to candy advertising and the number of candy bars l4 eaten. But a somewhat stronger relationship resulted between exposure to candy advertising and reduction of inhibitions about excessive candy eating.” Research by Moschis also uses the social learning model to study adolescent consumer socialization by the mass media and by peers, parents, and school. But none of the potential influences is considered as a model. For example, parental influence is the amount an adolescent talks with his parents about consumption; school influence, the number of consumer education courses taken. Likewise, the effect of the mass media was not conceptualized as a modeling effect, but rather the amount of exposure, the type of medium, and reasons for using the medium.” The Modeling Hypothesis in Studies of Adolescent Media Use As early as the mid-19505, researchers began testing the possibilities that young people followed the example of television-watching set by their parents. On the basis of positive correlations between patterns of media use among parents and children, Himmelweit, Oppenheim, and Vince concluded that parental example was an important factor in determining how much children watch television.16 Three years later, Schramm, Lyle, and Parker reached a similar conclusion on the quantity of television children watch17 and found additionally that parental example influenced the child's progam selection. If a parent views educational television, then a child is almost certain to do so; and if neither parent does view educational television, the child is almost sure n_o_t to do so. This is a very potent kind of influence . . . . Example is the best persuader . . . . The strong conclusion that modeling of parental television viewing exists raised several issues about the studies. Subsequent research of adolescent media behavior has questioned the two early studies in both conceptual and methodological terms. In the following sections, these issues are discussed, as are problems with the later research. l5 Conceptualizations of Media Modeling Behavior The notion that parent-child similarities in media behavior implies "parental example," or imitative behavior or modeling, is clearly expressed in the research of both Himm elweit et a1. and Schramm et al. The criticisms leveled at their research from Chaffee, McLeod, and Atkin,l9 however, pointed out other plausible explanations for the findings, including other types of modeling. 1. Negative modeling. The child may emulate the parent in terms of avoidance of television or other media. If a parent avoids television to participate in another activity that the child imitates, the child would be modeling the parent's non-use of television. 2. Reverse modeling. The parent may be following an example set by the child. That is, a child's viewing may result in parents' viewing, too. 3. Opportunity. Because someone has the television set on, another person may be exposed to the set whether he chooses to or not. Chaffee et a1. claim that this situation limits the chances that the second person could avoid a program he dislikes. Furthermore, they point out that personal influence would have little to do with television viewing; the person who controlled the dial would be the "influencer." 4. Third factors. It is possible that a child and parent would be independently attracted to the same medium or content by other attributes members of the family share. For example, parents and children are likely to be similar in terms of socioeconomic background, mental and physical capabilities, and availability of other media. These factors may contribute to similarities in media use, irrespective of effects from modeling or opportunity. Conceptually, then, the Chaffee, McLeod, and Atkin study in 1970 treated modeling in a manner similar to previous research; that is, parental influence over child behavior. And they also conceptualize modeling in terms of child influences over the parent (reverse modeling) and parental influence over child avoidance of behavior (negative modeling). Although negative modeling is a form of influence of the parent over the child, it is not necessarily imitative behavior. Rather it is-in the case of the studies of Chaffee et al.--the child not performing a behavior a parent is not performing. For example, Chaffee et al. inferred from negative correlations between adolescent's TV time and parents‘ news reading that "if the parent reads news materials a great deal, the adolescent is somewhat more likely to either read or watch news presentations but it is definitely less likely to devote much time to television."20 Negative modeling may be a form of influence, but it is not necessarily imitative behavior or modeling. The concept of reverse modeling is a plausible explanation for similarities in parent-child television use, although Chaffee et a1. did not find support for it in and of itself. Only in combination with opportunity, they said, could the "reverse" explanation be possible. That is, the adolescent TV "expert" may draw the parent to a program the adolescent is watching, most frequently an entertainment program.21 An additional test of reverse modeling came from research by Surlin, Wurtzel, and Whitener in which they accurately conceptualize reverse modeling as ". . . the child's media use influences the parent's, rather than vice versa."22 But methodologically, they were not dealing with child influence in the manner of an "expert" whose behavior would be imitated by parents. Rather, they dealt with the influence of "mature theme" warnings on parents' own viewing decisions. That is, if parents heeded the warnings and exercised discretion for their children's television viewing, they and their children would watch alternative programs. This is an indirect influence on the parents' TV watching, of course, but hardly modeling or reverse modeling as it has been previously conceptualized. Though the reverse modeling explanation is conceivable for parent-child similarities in television use, it seems unlikely as an explanation for similarities in newspaper or magazine reading. And, in fact, studies that have found similarities between parent and child reading have not offered reverse modeling as an explanation.23 Newspaper reading is an entirely different behavior from television viewing because the child develops the requisite skills and topic interests later in life. The child would be unlikely to acqlire the "expert" role he could acquire as a television viewer. Significantly, neither of the early studies had as its primary purpose a test of whether parental television-viewing behavior led to or explained similar children's television-viewing behavior. Rather, both studies evaluated the effects of television on children. Subsequently, it seems, both research teams discovered a relationship between parental use of television and child use, thereby suggesting that parental example was a primary way to mitigate adverse effects of television and increase positive effects. The later studies by Chaffee et a1. picked up on this theme, suggesting the potential for adverse effects as a rationale for testing the modeling hypothesis. By testing the hypothesis in isolation from other possible explanations of adolescent media use, they were suggesting modeling as a causal variable. But the existence of other possible explanations (even, as these researchers put l8 forth, other forms of modeling) makes this conceptualization of modeling questionable . This is a dilemma reminiscent of the theoretical examination of attitude- behavior inconsistency. The inconsistency has resulted when expressed attitudes of individuals have not corresponded with their subsequent behaviors, thus eliminating the attitudes as an explanation for the behaviors. Two conceptual solutions have been developed to deal with the inconsistency. One treats the attitude as one of many variables causing a behavior, the other variables affecting behavior in a way that results in inconsistency with the attitude. In these terms, the research task is to identify other variables. The second conceptual solution is to treat an attitude as a correlate with a behavior where the size of the correlation and the identifying conditions that affect the size are not seen as significant research problems?“ In the research of adolescent modeling of parental media behavior, a related problem has developed. Because of low correlations between parental use and adolescent use, Chaffee, McLeod, and Atkin all but dismissed parental use as a cause of adolescent use of media. Other conceivable explanations of adolescent media use suggest similarities with the first conceptual solution to the attitude-behavior inconsistency. The research task, then, would be to find other variables explaining adolescent media use, which the researchers have done, but only in a limited manner. A more parsimonious solution, and one used in the present study, is treating the parental behavior as a correlate, not a cause of, adolescent behavior. Another problem with previous research of adolescent modeling of parental media behavior is the notion that similarities between adolescents and parents constitute the equivalent of imitative behavior. This study suggests that the concept modeling should incorporate a more active role on the adolescent's part, 19 an awareness of the model's behavior, and, at the least, an implied desire to behave similarly. Up to now, the discussion of modeling in studies of adolescent media behavior has been confined to those that see the concept as a form of direct influence. Some later studies saw modeling as a form of indirect influence of parents over children in the form of family communication. This idea of indirect influence through interpersonal communication processes was first conceptualized by Maccoby in a study of the television- viewing habits of five- and six-~year-olds.26 She found that the upper-middle class children who are highly frustrated in their home life (subject to restrictions and not treated permissively or warmly) spend more time watching television than similar children who are not so frustrated. In the upper-lower class, no relationship existed between television-viewing and frustration. Maccoby interpreted the differences between the classes as meaning that in the upper- lower class, where parents watch a lot of television, the child has more positive motivation for watching. That is, the child will be drawn to it even in the absence of frustration because it is a dominant family activity. In the upper- middle class, where adult television-viewing is not so great, the absence of frustration also results in the child's participating in a dominant family activity, something other than watching television. The modeling influence here is an indirect one in that children who are not frustrated (and, therefore, who have warm relationships and good communications with their parents) will want to be like their parents. In the upper-middle class, children who are not frustrated will avoid television as their parents do; in the upper-lower class, children do as their parents do and watch television. 20 A later study by Chaffee, McLeod, and Atkin27 further investigated modeling as a function of the indirect influence of parental-child communication. Their research used the family communication pattern typology to determine the relationship between communication and modeling of media behavior. The typology consisted of two dimensions of family communication. The socio-oriented dimension stressed relations between the child and parents and has been comparable to unidimensional, social-power concepts, such as "autocratic-democratic" and "controlling-permissive." Parental emphasis is on maintaining interpersonal harmony and avoiding controversy. The concept- orientation encouraged the child to express ideas at the risk of exposure to controversy.28 Although the two dimensions have been treated as continuous variables, analysis has typically been done by dichotomizing both dimensions at median scores to create a four-fold typology of family communication.29 Laissez-faire families emphasize neither socio- nor concept-oriented relations (usually little parent-child communication takes place). Protective families emphasize socio- oriented communication at the expense of ideas. That is, obedience and social harmony within the family are valued at the expense of conceptual matters. Pluralistic patterns are the opposite of protective in that communication and discussion of ideas are encouraged and little emphasis is put on social constraint. The consensual family is one in which both relations exist. The child is encou'aged to seek new ideas, as long as he or she does not disturb the family's power and harmony structure. Studies using this typology assume that parent-child communication is such a pervasive influence for the child that the child uses the relations of family communications as a prototype for each newly encountered situation. In other _—~...-..—r~__ 21 words, the patterns of parent-child communication result in ways of coping that the child applies in a variety of circumstances. Implied in the family communication model is that transmission of knowledge and values of a generic nature characterizes parent-child communication across specific knowledge and values. Research employing the typology and testing the modeling hypothesis found greater parent-child media use similarities among socio—oriented families than among the concept-oriented. The inference, the researchers said, is a greater degree of modeling among socio-oriented families. Because this orientation stresses avoidance of controversy and repression of feelings on extrapersonal topics, they said, the similarities could be attributed to parental example.” In other words, the child has media preferences similar to his parents to maintain harmonious relations. This is, of course, an inference in direct contradiction to the explanation first expressed by Maccoby. Equating the socio-orientation with Maccoby's frustration (restrictive with lack of warmth), one would expect children in socio- oriented families to behave in a manner unlike their parents, as frustrated children in Maccoby's study did. Similarly, one would expect concept-oriented children (Maccoby's non-restricted, warmly treated children) to desire to be like their parents. There is also evidence, both from the Maccoby study and that of Chaffee et al., that parent-chfld communication, regardless of form, is a function of socioeconomic status. For example, statistical control of socioeconomic factors have resulted in partial overlap between status differences and family communication patterns .31 22 Operationalizations of Media Modeling Behavior This section centers on the ways in which the studies of parental influence on adolescent media behavior have measured modeling. Specifically, it deals with the following issues: (1) previous measurements of modeling, (2) an alternative method of measuring modeling, (3) the inclusion of another group besides parents as potential models for adolescents, and (4) placing the communication behavior in the context of another behavior. Measuring Modeling by Adolescents In both early studies that inferred parental example, the researchers used an "amount of time" measure for assessing television viewing. Himm elweit et a1. relied on children's reports of parental behavior, while Schramm et al. used self reports from both groups. The issue here is not perceptions versus self-reports (that issue will be addressed later). At issue is a "general" media use measurement of the amount of time spent watching television. The strong relationship, as they see it, between the amount of time parents watch television and the amount of time children watch television is an indication of parental example. A few issues can be raised here. A strong relationship between the amount of time a child watches television and the amount his parent model watches could indicate nothing more than coincidence. That is, the child could be watching after school or after dinner while the parent watches later in the evening, unobserved by the child. Furthermore, if the concern of those who have tested the modeling hypothesis or those who have used a modeling explanation is the possibility that observing children will watch similar programming as their models, then the variables being measured should be content related rather than time related. 23 The limitation of using amount of viewing time was overcome by Schramm et al., who showed significant parent-cth correlations for viewing specific types of programs, such as crime shows and westerns. Similarly, the two subsequent studies by Chaffee et al. used an "amount of time" measure and a more specific measure dealing with program categories. Only the latter two studies use indices to measure the amount of time spent watching television, although both also asked respondents to estimate "time spent viewing on an average day." Two indices-mews viewing and entertainment viewing--consisted of three items each, items measuring the frequency (sometimes, often, never) with which respondents watched certain types of programs (comedies, national news, Westerns, news specials, adventure or spy shows, and interview shows). Both the research reported by Chaffee et al. and Schramm et a1. use self-reports from their young respondents and from the parents of those respondents to measure the viewing behaviors of both groups. the relationships between the behaviors reported by the two groups measured the extent to which modeling existed. An analysis of that technique comes later in the chapter. Placing the Communication Behavior in Context Since the research of Chaffee et al. in the late 19605, several studies have emerged placing communication behavior in relationship with another behavior. Studies of communication behavior as part of a broader investigation of political behavior initiated this kind of research .32 But recent research has linked the use of media by adolescents to the broader study of consumer behavior and socialization into the economic system.33 Heeding the words of McLeod and O'Keefe,“ these researchers seem to believe that communication behavior is best understood in relationship with other kinds of behaviors. 24 Other Potential Models McLeod and O'Keefe have also suggested that adolescents may be more influenced by their peers than their parents, accurately pointing out that most research of adolescents and media use modeling includes parental influence but overlooks a peer component.” A vast majority of studies dealing with influences on adolescent smoking, drinking, and educational and occupational aspirations include comparisons of the relative influence of parents and peers.36 Underlying most of this research are findings that adolescents spend more time with peers their own age than they do with their parents and, consequently, are influenced by their peers. ‘ The result, these researchers argue, is the existence of a youth culture that dominates the behavior of young people, causing them to be independent of the adult world and less closely related to their families.” Although not derived from this research, studies of how adolescents learn consumer behavior also include parental and peer influences.38 The research suggesting potential influence of both parents and peers notwithstanding, other researchers reject a generation gap and its effects. For example, Campbell called adolescence a transitional period during which the individual comes to rely less on parents and more on peers. But, he said, parents remain the most important influence on adolescents.39 Likewise, media industry research implies parental influence when it assumes that certain media-use habits of adolescents are formed in the homefto The apparent disagreement over the existence and effects of a generation gap provides a key component of the present study. Previous research of modeling influences on adolescent media behavior has dealt only with parental modeling. Research in other disciplines on relative influences of parents and peers and generation gap research suggest the possibility of peer influence on adolescent media behavior. 25 Alternative Method of Measuring Modeling ‘ An assumption of the research done by Chaffee et a1. is that people model only what they can observe. And the basis for their operationalization (the relationship between the adolescent and parent self—reports) is the assumption that adolescents observe their parents using the media. One would assume that they were testing the modeling hypothesis for adolescents and their peers, they would use a similar operationalization. The fact is that they operationalized the variables of the modeling hypothesis in a way that prevents any inference about influence, much less the direction of the influence. A high correlation between parent use and adolescent use merely indicates a similarity in their use of media, their conceptualization of modeling. A modeling hypothesis could find greater support in this situation if the adolescent were observing a parent's novel behavior or observing a parental behavior for the first time. Such is not the case in all probability and so the findings of similarities would require alternative explanations. In fact, in most media use situations, the inference could never be made that the adolescent child models his parents' behavior based on the relationship between parent and adolescent self-report alone. For example, if the adolescent and the parent are similar in their use of media, but no measurement exists that the adolescent is aware of or has made a judgment about the parental media use, the modeling hypothesis would lack support. In this case, alternative explanations for the similarity would prevail (the Chaffee, McLeod, Atkin paradigm). It is fair to assume, however, that adolescents come to many situations, including those involving media, with a certain amount of experience and the ability to make judgments about the value of another's behavior. The failure of previous research to take those factors into account provides the rationale for proposing an alternative method of testing the 26 modeling hypothesis. The next section provides further details of the method by establishing conditions under which modeling exists, conditions that derive from the co-orientation research approach. The co—orientation research approach finds its origins in Newcomb's ABX paradigm,“1 but uses the terminology and measurement techniques of person- perception research.“2 The term co-orientation refers to a situation in which a person is aware of the same object as another person. Co-orientation could also pertain to the first person's awareness of the second person's interest in the object and attitude toward it.” In the simplest situation, co-orientation exists when one person is focused on the same object as the second person. In this two-person situation, four cognitions are possible: person A knows what s/he thinks about an object and has a perception of what person B thinks; the reverse is also true. Implied in this situation is a social-systemic“ (in this instance, dyadic) model rather than one whose unit of analysis is at an individual level. This approach and the individual agreement approach comprise the two main traditions in research of consensus.” In studies of consensus using the individual agreement approach, researchers would measure the extent of agreement on an object between individuals in a group. The research of parental influence on adolescent media behavior has measured modeling in much the same manner. The conceptualization of modeling in these studies-similarities between parent and child-~compels the researchers to use this operationalization. In the study of consensus, however, Scheff has argued that the individual agreement approach "makes no provision for perceptions of agreement, which may be independent of actual agreement."46 A parallel argument is made here pertaining to the study of modeling. Because of its reliance on an individual agreement conceptualization, the 27 approach used by Chaffee et. al. makes no provisions for perceptions, which the co—orientation approach does provide. And like the co-orientation approach in consensus research, perceptions in modeling research may be independent of actual agreement. The independence of perceptions from actual agreement, however, may be problematic in modeling research. For example, if one's perceptions of another are similar to one's own behavior, would that person be modeling if the perceptions were inaccurate? Perhaps not, although the person's perception could result in another type of influence. One could argue that if modeling required imitative beahvior, then an action based on an inaccurate perception, though influential on one's behavior, precluded modeling. This relates to one of the four processes discussed earlier in this chapter, the attentional process. Implied in Bandura's analysis is the necessity for an individual to accurately perceive the actions of a model, not just have exposure to the model. A person cannot learn much by observation if he does not attend to, or recognize, the essential features of the model's behavior... Simply exposing persons to models does not in itself ensure that they...will even perceive accurately the aspects they happen to notice. Testing the modeling hypothesis by relating the behaviors of teen-agers and those of their parents implies that modeling can occur with simple exposure to the parental behavior. This research, however, fails to consider whether the adolescent recognizes the parental behavior. The present study attempts to consider that necessity through the measurement of adolescent perceptions of models and the accuracy of those perceptions. With these issues raised, it is now necessary to address them and relate them to the co-orientation approach and its constructs. As mentioned previously, co-orientation exists when two individuals are focused on a similar object, the result from this two-person situation being four cognitions: what the 28 first person knows and what that person thinks the second person knows and vice versa. From these cognitions come three structural relations. First is agreement, the relationship or comparison between what the first person knows and what the second person knows.48 Second is congguency, whereby the comparison is between what the first knows and what he perceives the second person to know.“9 Finally, accuracy forms from the comparison of the first person's perception of the second person's cognitions and the second person's actual cognitions. It should be pointed out that congruency is not a true co- orientational construct in that it is strictly intrapersonal. It does, however, exist as a third element in the context of co-orientation, agreement and accuracy being the other two. Furthermore, a lack of uniformity in labeling and conceptualizations exists in co-orientation research and related approaches. Different labeling may result in part from the various research streams that have contributed to co- orientation or have formed simultaneously with the co-orientation approach. For example, Newcomb uses the term "consensus" or "similarity" to represent agreement. He called congruency "perceived symmetry" or "perceived consensus."5O Person-perception researchers have called agreement "mutuality,”l while elsewhere it has been referred to as "concurrence" or "concordance."52 Discrepancies in the way congruency is conceptualized also appear in the literature, although these may be readily explained by the ways in which the concept is used in various types of research. Because co-orientation and its concepts originate with Newcomb's ABX paradigm, a look at his "perceived symmetry" or "perceived consensus" concept will be instructive in this discussion. 29 Newcomb's research concluded that congruency (or perceived consensus, perceived symmetry, perceived agreement) is a variable. As a criterion variable, congruency "varies as a function of intensity of attitude. . . and of attraction."53 So, if person A and person B are co-oriented to X, person A is more likely to perceive agreement with B in regard to X if person A considers person B attractive. As an independent variable Newcomb's perceived symmetry is the "determinant of instigation to symmetry, directed communication."54 Here, if person A perceives that person B differs in regard to X, person A will communicate his point of view to B, will reassure himself that disagreement with B does not exist, or will provide more information about X to B. Whether the independent or criterion variable, perceived symmetry could be operationalized by the question: "To what extent do you and person B agree on X?" Person A in this instance would be required to decide what person B believes about X and assess his own agreement with person B. In the present research the term congruency refers to a construct of co- orientation whereby person A would only be required to judge person B in regard to X. The purpose of this research is not to determine the effects of perceived symmetry or agreement or how perceived symmetry or agreement is affected, but rather to determine the e_x;tgg_t_ of perceived symmetry or perceived agreement. Because those terms from Newcomb's research imply completely subjective responses, however, the term congruency from co-orientation is used here. Congruency is the relationship between what person A believes and what he perceives person B to believe, a relationship that is judged objectively, even though person A's perception is subjective. Accuracy is the one co-orientation concept for which uniformity does seem to exist, both in terminology and definition. It is the extent to which a person correctly perceives another's behaviors or simply "correctness of perception."55 30 In summary, research on adolescent modeling of parental media behavior has taken an objective approach to testing the modeling hypothesis. That is, self-reports of media use by adolescents and parents have been compared. This measurement approach to modeling is reminiscent of the individual agreement approach to measuring consensus, an approach that disregarded perceptions of an individual and used the individual as the unit of analysis. The co-orientation approach, on the other hand, takes into consideration perceptions and allows a dyadic unit analysis. Two co-orientation concepts are important to the study: congruency (the relationship between one's perceptions and another's behavior and one's own behavior) and accuracy (the relationship between one's perceptions of another's behavior and the other's actual behavior). The extent of congruency will indicate influence on one's behavior; that influence cannot be considered modeling unless accuracy exists. Adolescent Media Use: Other Frameworks Modeling and co-orientation are among several research approaches to the study of adolescent media use. This section discusses and analyzes other frameworks that have been used in research of media behavior by adolescents, including the uses and gratifications approach and the cumulative acquisition model. Regardless of the theoretical underpinnings of the research, most studies of adolescent media use deal with adolescent use of television. Furthermore, a great deal of this research attempts to determine preference and patterns of television viewing on the basis of demographic differences, rather than differences derived from more formal theoretical frameworks. Socioeconomic variables, for example, have played an important role in research of television 31 use by adolescents, often with contradictory findings. A comprehensive review of the topic points out that among teen-agers, those from lower-income households and from households headed by college-educated parents watch less television than other types of teen-agers.56 Two other studies found that lower- income and working-class adolescents watch more television than other types.57 And another found no significant relationship between a teen-ager's level of participation in social activities and television watching.58 Comstock et al. also portray this composite picture of television viewing by young people in terms of age: Television viewing increases in early adolescence (12-14 years), but declines during high school years when competition from other activities ensues.59 Finally, several studies report a negative correlation between intelligence and the level of teen-age television viewing,60 but another suggests this relationship is disappearing.61 Aside from the studies that use demographic variables to explain television viewing by adolescents, the most prevalent research framework has been the uses and gratifications approach. This research perspective emphasizes the individual as an active seeker, selector and user of the media.62 Proponents of the uses and gratifications approach suggest that members of the audience use the media to satisfy extrinsic individual needs. Researchers have attempted to explain the motivations for specific media behavior and have stressed functions and consequences of media to better understand the role of media in the lives of the audience. The rationale for this approach is expressed well by Johnstone: ...the analysis assumes that the media can have little or no impact on persons who have no use for them, that media fare is selected rather than imposed, and that particular media offerirkgs are chosen because they are meaningful to those who choose them. Thus, the uses and gratifications perspective undermines the media effects approach in that it recognizes that "even the most potent of mass media cannot 32 ordinarily influence an individual who has 'no use' for it in the social and psychological context in which he lives."64 An elaborate critique of the uses and gratifications approach suggests that studies using the framework have failed to clearly and precisely analyze concepts, have not elaborated how uses and gratifications explanations differ from other explanations of media behavior, and have not taken into account the basic tenet of the approach, that people actively use the media, rather than passively.“ Swanson argued the conceptual ambiguity results when the uses and gratifications researchers identify three theoretical positions from which the approach can be used: functionalism, a structural/cultural position, and an action/motivation position. The problem, as Swanson saw it, is that the positions "are fundamentally contradictory" in a way that makes combining them confusing in explaining data. Each position, he said, looks at behavior in a distinct way. For example, in the functionalist position, individuals are conceived as active pursuers of information to maintain roles in everyday life. The structural/cultural approach focuses on society as the unit of analysis, rather than the individual. This approach conceives of the individual as part of a collective system whose actions emanate from the culture. Finally, the action/motivation approach considers the individual as its unit of analysis. The individual, however, is not conceived as behaving in relation to the system, but for more self-actualizing reasons. Swanson argued that acceptance of any one of the positions precluded acceptance of the others. The acceptance of more than one, he said, resulted in an eclecticism that defies a satisfactory conceptual framework. Furthermore, the lack of specificity for major concepts employed in uses and gratifications 33 research (e.g. "use," "gratification," "motive," "need") "may...make research employing such vacuous and amorphous concepts a self-validating exercise."66 Another Swanson criticism--a confused explanatory apparatus-«also seems to result from conceptual problems, in this case with the concept "media use." Is "media use" a motive as in: "in order to _____"? Or is "media use" a function fulfilled by media consumption? Is it a cause of behavior or a consequence? Failure to answer these questions results in an unclear explanatory framework. Finally, Swanson pointed out that uses and gratifications researchers fail to consider one of their basic premises: that persons are active in their media behaviors rather than passive. And, as he indicated, one basic way they are active is through the process of perception. It is ironic that uses and gratifications research, so grounded in the idea of active persons, does not consider persons' perceptions of media content. Failure to do so, Swanson said, means leaving unconnected persons' interpretations of media content and the other stages of the uses and gratifications mass communications process. In general, then, criticism of uses and gratifications has dealt with the inability of its proponents to develop a grand theoretical framework that links the approach to social and psychological origins. The basis for this criticism is the development of a number of audience function typologies from the various uses and gratifications studies. The variety of typologies, critics say, has prevented synthesis and integration of research. For example, Brown employed a uses and gratifications approach to determine the frequency of television use by children (including adolescents) for certain purposes, purposes defined by 13 function statements.67 Citing changes in gratifications across age groups, Brown found support for a functional reorganization hypothesis. The hypothesis states that uses of media by children change as they undergo an ongoing process of reorganization.68 34 Dembo and McCron list 11 uses/gratifications of media to discover the relationships between their gratifications statements, media exposure, and self- conceptions of adolescent boys. Although the researchers find a relationship exists between the boys' self-concepts and media involvement, their data do not indicate differences across self-concepts (street-culture orientation vs. education orientation) for media functions.69 Greenberg factor analyzed 31 "reasons for watching television" and found six "clusters of reasons" that assume "a unique role in the orientation" of children to television: habit, arousal, companionship, relaxing, forgetting and learning. Each of these "motivations" was felt more strongly by younger children]0 Replication of the Greenberg study with American children produce similar results.71 In a defense of the uses and gratifications approach, McGuire accurately pointed out that uses and gratifications researchers are usually using the gratifications offered by media as a starting point, using the potential rewards media offer as explanations for use of the media]2 Critics of the approach, however, claim that media consumption is not a deliberate process, but a haphazard one, determined by chance and external circumstances, such as opportunity to read a newspaper or the available television schedule. McGuire admits the truth of these arguments, but says personal motivations or gratifications could also account for media consumption, particularly "when we turn to the question of maintaining continued exposure once one has found appropriate mass communications material."73 And Blumler deflects the criticism of lack of a grand theory by refocusing on the fact that the purpose of uses and gratifications research is "to get to grips with the audience experience itself ."7"‘ 35 These last two statements relate to Swanson's criticism that the uses and gratifications paradigm fails to recognize persons as active in their perceptions of media content. If, in fact, the purpose of the approach is to understand the "audience experience" with or "personal motivations" for media consumption, a greater awareness of audience perceptions of media would be important. Applying the conceptual glue to the uses and gratifications framework, however, is beyond the scope of the present study. No doubt that adolescent exposure to media is an ongoing process, not an original one. Appropriately analyzed, uses and gratifications concepts may provide an adequate explanatory framework for adolescent media behavior. In a like manner, the study of adolescent media behavior from a modeling perspective assumes continuous rather than original media consumption, especially when the comparison of parental and peer influences is considered. This similarity notwithstanding, the broad range of conceptual and methodological issues and the necessity to bring so many additiOnal variables into the study recommend that a uses and gratifications approach not be incorporated here. Another theoretical framework for adolescent behavior is cognitive development theory. Cognitive development comes primarily from the work of Jean Piaget and views the emergence of the individual as the result of interaction between the individual and his environment over time.75 Socialization of the child takes place in qualitative stages (the theory is sometimes called stage theory), although in application of the theory, age serves as a proxy variable. This type of application of cognitive development has been used recently in studies of adolescent consumer behavior,76 even though critics argue the theory requires true development stages with psychological properties that do not arbitrarily correspond to chronology]7 36 The most recent model developed to explain adolescent media behavior is the cumulative acquisition model.78 Drawing from cognitive development theory, the model posits a progression of media use by the adolescent from electronic media (radio and television) to print media (newspapers and magazines). Chaffee and Tims provide the following rationale for the ordering: There is an overriding expectation of a developmental broadcast- print sequence based on the assumption that print requires a higher level of cognitive skill and involvement, and a fairly extensive knowledge base, for the person to establish a pattern of regular use. Secondly, within both the broadcast media and the print media there are empirical probabilities that encourage us to expect sequential dominance of specific media. The model "is cumulative in that the attainment of each succeeding level presupposes mastery of those that precede it."80 Empirical tests of the model have had mixed results. The initial test, based on a single cross-sectional sample, supported the basic premise that a cumulative progression existed from broadcast news to print. Guttman scale analysis identified that 86 percent were "scale" types: people who watch TV news two days a week, but did not read the newspaper three days a week or read three magazine articles (a "1" score); people who watched TV news two days a week and read the newspaper three days a week, but who did not read three magazine articles (a "2" score); and people who did all three (a "3"). Further analysis attempted to show a relationship between scale scores and news media reliance. For example, reliance on newspapers as one's principle source of news increases as scale scores increase. The progression, however, does not fully support the model in that heavy reliance on television exists at level "2" and on television and newspapers at level "3".81 Subsequent longitudinal analysis using the cumulative acquisition model found even less support for the premise that young people's news media uses progress in stages related to the Complexity of the news media content. The 37 only evidence of "cumulative growth" was among those in the fourth through sixth grades at the outset of the study. According to the authors, this suggests that adolescents have achieved a consistent level of news media use by early adolescence . 32 The cumulative acquisition model of adolescent news media use is intuitively appealing, but it evokes previous criticisms of cognitive development research. Development theories have been conceptualized so that qualitative development and quantitative development proceed in stages together in discontinuous fashion. There is some question, however, about whether cognitive growth "jumps" from one step to another rather than in a more discrete manner. Questions also exist over equating age or grade with a particular stage of development. For example, an assumption of the cumulative acquisition model is that news media use "levels" will increase with age when, in fact, the empirical data showed it leveled off in early adolescence. Finally, one cannot assume that a child at a certain stage in one area will be at a similar stage in another area. Again, the cumulative acquisition research serves as an example for it incorrectly assumes that one's reading level parallels one's newspaper-reading frequency. Summary This review has shown the problems with previous studies of adolescent modeling of parental media behavior. Conceptually, these studies treat modeling as the imitation by an observer of a behavior performed by a model. Similarities between the behaviors of an adolescent and those of a parent, for instance, would constitute modeling on the part of the adolescent. But the similarities conceptualization allows several alternative explanations, including 38 parental modeling of adolescent behavior. Were the modeled media behaviors novel behaviors, this conceptualization might suffice. In all likelihood, however, adolescents are not observing novel media behaviors by potential models. The adolescent comes to a media behavior situation with previous experience and, for the purposes of modeling, the ability to judge the media behaviors of others. For this reason and because the similarities conceptualization allows a number of alternative explanations, the present study posits a more rigorous conceptualization of modeling as imitative behavior based on a correct perception by the observer of the model. That is, for modeling to occur, the adolescent must perceive the media behavior of another, be accurate about the perception, and behave in a similar fashion to the perception. This suggests the following model, where X = one's perception of another's behavior; Y = one's own behavior; and Z = another's behavior: Z >X )Y In other words, one's perception of another's behavior intervenes with another's behavior and one's own behavior. Another's behavior would also relate to one's own behavior directly in the modeling outcome. The model suggests that when one's perceptions are held constant, the relationship between one's own behavior and another's behavior will decrease or disappear. The relationship change as posited means one's perceptions of another would serve as an intervening variable between the behaviors of another and one's own behavior. This means the perceptions would mediate the relationship between another's behaviors and one's own behaviors. For purposes of example only, assume that another's behaviors and one's own behaviors are not related directly, but only indirectly through the perception of another's behaviors. This assumption can be tested by controlling or removing the effects of the 39 perceptions. A zero or close to zero relationship between another's behaviors and one's own behaviors would support the assumption. In addition to reconceptualizing modeling, the present study also deviates from previous research on adolescent modeling of media behavior in that it considers peers as models in addition to parents. Previous research has indicated that adolescence is a transition period, one during which adolescents spend more time with same-age peers that they had previously. Other researchers claim that despite this transition, parents maintain the most influence over their adolescent children to parents. It is with these criticisms in mind that the following hypotheses are stated: H1: Adolescents will correctly perceive the media behaviors of their parents to a greater extent than they correctly perceive the media behaviors of their peers. H2: Adolescents will behave similarly with their parents' media behaviors to a greater extent than they will with their peers'. H3: Adolescents will model their parents' media behaviors more than they will those of their peers'. Several issues raised in the previous discussion are relevant to the hypotheses. In terms of the predicted direction, these hypotheses are based on the theoretical approach that regardless of any peer influence, parents remain the "major shaping influences in the lives of adolescents...Chi1dren tend to be like their parents."82 Empirical research also provides some basis for the hypotheses. Atwood found that young people accurately perceived the program preferences of their parents.83 Abel, meanwhile, found that among certain children, the perception of what mothers preferred their children to watch was related to what the mothers actually preferred.“ Industry research, in which tests were not performed, showed a tendency for the development of media behavior habits in the home.” Based on these studies, one would expect adolescents to correctly perceive their parents more than their peers. 40 The hypotheses are not only positing expected direction vis-a-vis parents and peers, but they also provide the operational definitions of the key co- orientation concepts discussed earlier. For example, the first hypothesis deals with the equivalent of accuracy, and predicts that adolescent accuracy of parental media behavior will be greater than adolescent accuracy of peer media behavior. The second hypothesis asserts that agreement between adolescent media behavior and parent media behavior will be greater than agreement between adolescents and peers. Previous research of adolescent modeling of others media behavior would have tested only the second hypothesis. But because of the more rigorous modeling conceptualization in this study, a third hypothesis has been stated based on the expectations of the first two. If the expectations of the first two hypotheses are fulfilled, the third will be supported. That is, if greater accuracy and agreement exist for parents than for peers, greater modeling of parents will exist than of peers. The reconceptualization requires agreement and accuracy, therefore if either of the first two hypotheses is not supported, the third will not be. For example, if adolescents correctly perceive parents' media behaviors (accuracy) more than they do peers' media behaviors, but behave more similarly to peers (agreement) than to parents, the first hypothesis would be supported, but not the second or third. Similarly, adolescents could behave similarly (agreement) to parents to a greater extent than to peers, but correctly perceive peers to a greater extent than parents (accuracy). Both the first and third hypotheses would not be supported in this case. #1 NOTES lMichael Ray, "Psychological Theories and Interpretations of Learning," in Consumer Behavior: Theoretical Sources, ed. Scott Ward and Thomas S. Robertson (Englewood Cliffs, N.J.: Prentice-Hall, 1973): 76-77. 2Albert Bandura, Social Learning Theory (Morristown, N..:J General Learning Press, l97l),p. 2. 3Albert Bandura, Dorothea Ross and Sheila Ross, "Transmission of Aggression Through Imitation and Aggressive Models," in Readings in Social Ps holo , ed. Jonathan L. Freeman, J. Merrill Carlsmith and David O. Sears (Englewood Cliffs, N.J.: Prentice-Hall, 1971): 162-175. “Albert Bandura, Dorothea Ross and Sheila Ross, "Imitation of Film- Mediated Aggressive Models," Journal of Abnormal and Social Psychology 66 (1963): 3-11. 5Edward Zigler and Irvin Child, "Socialization," in The Handbook of Social Psychology, ed. Gardner Lindzey (Reading, Mass.: Addison-Wesley, 1969), p. 530. 6Bruce Biddle, Barbara J. Bank and Marjorie M. Marlin, "Parental and Peer Influence on Adolescents," Social Forces 58 (June 1980): 1057-1079. 71bid. 31bicl. 9Morris A. Forslund and Thomas J. Gustafson, "Influence of Peers and Parents and Sex Differences in Drinking by High School Students," Quarterly Journal of Studies on Alcohol 31 (1970): 868-875. loBiddle, Banks and Marlin, "Parental and Peer Influence." 11Robert E. Herriott, "Some Social Determinants of Educational Aspiration," Harvard Education Review 33 (1963): 157-177. 12Denise Kandel, "Inter- and Intragenerational Influences on Adolescent Marijuana Use," Journal of Social Issues 30 (Spring 1974): 112. 13Richard P. Adler et al., Research on the Effects of Television Advertising on Children, National Science Foundation Report (Washington: Goverrment Printing Office, 1978 8): Chapter 5. 1"Charles K. Atkin, "Children's Social Learning from Television Advertising: Research Evidence on Observational Modeling of Product Consumption," Advances in Consumer Research 3 (1976): 513-519. 42 15George Moschis and Gilbert Churchill, "Consumer Socialization: A Theoretical and Empirical Analysis," Journal of Marketing Research 15 (November 1978): 599-609. l6Hilde Himmelweit, A.N. Oppenheim and Pamela Vince, Television and the Child (London: Oxford University Press, 1958). l7Wi1bur Schramm, Jack Lyle and E.B. Parker, Television in the Lives of Our Children (Stanford, Cal.: Stanford University Press, 1961). 18lbid., p. 132. 19Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parental- Adolescent Similarities in Television Use," paper presented to the Association for Education in Journalism, Washington, 1970. 20Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parental Influence of Adolescent Media Use," American Behavioral Scientist l4 (January/February 1971): 329. 2“Chaffee et al., "Parental-Adolescent Similarities," p. 11. 22Stuart Surlin, Alan Wurtzel and Linda Whitener, "Parental Control of Children's Viewing Behavior: Support for the Reverse Modeling Principle," paper presented to the International Communication Association, Chicago, 1978. 23Chaffee, et a1 ., "Parental Influence." 2“Allen Liska, The Consistency Controversy (New York: John Wiley and Sons, 1975), pp. 8-15. 25Chaffee et al., "Parental Influence." 26Eleanor Maccoby, "Why Do Children Watch Television?" Public Opinion Quarterly 18 (Fall 1951!): 239-2114. 27Chaffee et al., "Parental Influence," pp. 330-338. 281bid. ”Jack M. McLeod and Steven H. Chaffee, "The Construction of Social Reality," in The Social Influence Processes, ed. James T. Tedeschi (Chicago: Aldine-Atherton, Inc., 1972), pp. 50-99. 30Chaffee et al., "Parental Influences," p. 337. 31Jack M. McLeod, Steven H. Chaffee and Daniel B. Wackman, "Family Communication: An Update Report," paper presented to the Association for Education in Journalism, Boulder, Colo., 1967. 43 32Steven H. Chaffee, L. Scott Ward and Leonard P. Tipton, "Mass Communication and Political Socialization," Journalism Quarterly 47 (Winter 1970): 647-659; Steven H. Chaffee, Jack M. McLeod and Daniel B. Wackman, "Family Communication Patterns and Adolescent Political Participation," in Socialization to Politics, ed. Jack Dennis (New York: John Wiley and Sons, 1973), pp. 349-364; and M. Margaret Conway, A. Jay Stevens and Robert G. Smith, "The Relation Between Media Use and Children's Civic Awareness," Journalism Quarterly 52 (Autumn 1975): 531-538. 33Moschis and Churchill, "Consumer Socialization"; Roy L. Moore and George P. Moschis, "The Role of Family Communication in Consumer Learning," Journal of Communication 31 (Autumn 1981): 42-51; George P. Moschis and Roy L. Moore, "Family Communication and Consumer Socialization," in Advances in Consumer Research vol. 6, ed. William L. Wilkie (Ann Arbor: Association for Consumer Research, 1979), pp. 359-363; Gilbert A. Churchill and George P. Moschis, "Television and Interpersonal Influences on Adolescent Consumer Learning," laurnal of Consumer Research 6 (June 1979): 23-35; and Scott Ward and Daniel Wackman, "Family and Media Influences on Adolescent Consumer Learning," American Behavioral Scientist 14 (January/February 1971): 415- 427. 3"’Jack M. McLeod and Garrett J. O'Keefe Jr., "The Socialization Perspective and Communication Behavior," in Current_Perspectives in Mass Communication Research, ed. Gerald Kline and Phillip Tichenor (Beverly Hills: Sage Publications, 1972), pp. 121-168. ”mm. 36Forslund and Gustafson, "Influence of Peers and Parents"; Kandel, "Inter- and Intragenerational Influences"; and Biddle et al., "Parental and Peer Influence." 37James L. Coleman, The Adolescent Society: The Social Life of a Teen- Ager and Its Impact on Education (New York: Free Press, 1961.) 38Moschis and Churchill, "Consumer Socialization." 39Ernest Q. Campbell, "Adolescent Socialization," in Handbook of Socialization Theory and Research, ed. David Goslin (Chicago: Rand McNally, 1969'). pp. 821-859. 40Mass Media in the Family Setting (New York: Newspaper Advertising Bureau, 1980). “Theodore Newcomb, "An Approach to the Study of Communication Acts," Ifiychologigal Review (1953): 393-404. “zRenator Tagiuri, Jerome S. Bruner and R.R. Blake, "On the Relationship Between Feelings and Perceptions of Feelings Among Members of Small Groups," 44 in Readings in Social Psychology 3rd edition, ed. Eleanor Maccoby, Theodore Newcomb and E.L. Hartley (New York: Holt, Rinehart and Winston, 1958). ”McLeod and Chaffee, "The Construction of Social Reality." “Thomas J. Scheff, "Toward a Sociological Model of Consensus," American Sociological Review 32 (1967): 32-45. ”lbw. 451bid., p. 33. Q7Bandura, Social Learning Theory, p. 6. 48McLeod and Chaffee, "The Construction of Social Reality." ”Steven H. Chaffee, Jack M. McLeod and Jose L. Guerrero, "Origins and Implications of the Co-orientation Approach in Communication Research," paper presented to the Association for Education in Journalism, Berkeley, Cal., 1969. 50Newcomb, "An Approach." 51Taguiri, Bruner and Blake, "On the Relation." 52Reuben Hill, J.M. Stycos and K.W. Back, The Family and Population Control: A PuertoLRican Experiment ifin Social Change (Chapel Hill: University of North Carolina Press, 1959f Denise B. Kandel and Gerald S. Lesser, "Parental and Peer Influences on Educational Plans of Adolescents," American Sociological Review 34 (1969): 213-223. ”Newcomb, "An Approach," p. 397. 54lbid. 55Taguiri, Bruner and Blake, "On the Relation ." 56George Comstock, Steven Chaffee, Natan Katzrnan, Maxwell McCombs and Donald Roberts, Television and Human Behavior (New York: Columbia University Press, 1978). 57Schramm, Lyle and Parker, Television in theLLives of Oujrf Children; and Bradley S. Greenberg, "Gratifications of Television Viewing and Their Correlates for British Children," in The Uses of Mass Communicatigy ed. Jay Blumler and Elihu Katz (Beverly Hills: Sage Publications, 1974), pp. 71-92. ”Jack Lyle and H.R. Hoffman, "Children's Use of Television and Other Media," in Television and SoLcial Behavior, Vol. 4: Television in Day-to—Day Life: Patterns of Use, eds., Eli Rubinstein, George Comstock and John Murray (washington: Government Printing Office, 1972). 45 ”Comstock et al., Television and Human Behavior. 6OSchramm et al., Television in the Lives of Our Children; and Jack M. McLeod, Charles K. Atkin and Steven H. Chaffee, "Adolescents, Parents and Television Use: Adolescent Self-Report Measures from Maryland and Wisconsin Samples," in Television and Social Behavior, Vol. 3: Television and Adolescent Aggressiveness, ed. George Comstock and Eli Rubinstein (Washington: Government Printing Office, 1972). 61Lyle and Hoffman, "Children Use." 62Alan M. Rubin, "T elevision Use by Children and Adolescents," Human Communication Research 5 (Winter 1979): 109-120. 63John W. C. Johnstone, "Social Integration and Mass Media Use Among Adolescents: A Case Study," in Uses of Mass Communication, ed. Jay G. Blumler and Elihu Katz (Beverly Hills: Sage Publications, 1974), p. 35. 6“Elihu Katz, "Mass Communication Research and the Study of Popular Culture," Studies in Public Communication 2 (1959): 1-6. 65David L. Swanson, "The Uses and Misuses of Uses and Gratifications," Human Communication Research 3 (Spring 1977): 214-221. 66lbid., p. 217. 671R. Brown, "Children's Use of Television," in Children and Television, ed. Ray Brown (Beverly Hills: Sage Biblications, 1976), pp. 116-136. 681bid. 69Richard Dembo and Robin McCron, "Social Factors in Media Use," in Children and Television, ed. Ray Brown (Beverly Hills: Sage Publications, 1976), pp. 137-166. 70Greenberg, "Gratifications of Television Viewing." 71 Rubin, "Television Use by Children." 72William J. McGuire, "Psychological Motives and Communication Gratifications," in The Usgsi of Mass Communication, ed. Jay G.B1umer and Elihu Katz (Beverly Hills: Sage Ptlblications, 1974), pp. 167-196. 73Ibid. ”Jay Blumler, "The Role of Theory in Uses and Gratifications," Political Communication Review 3 (1978): 3. 46 75Lawrence Kohlberg, "The Cognitive Development Approach to Socialization," in Handbook of Socialization Theory and Research, ed. David Goslin (Chicago: Rand McNally, 1969), pp. 347-480. 76Moschis and Churchill, "Consumer Socialization." 77Zigler and Child, "Socialization," pp. 450-589. 78Steven H. Chaffee and Albert R. Tims, "The Development of News Media Use in Adolescence: Implications for Political Cognitions, " paper presented to the International Communication Association, Boston, 1982. 791bid., p. 4. ”mm. 311bid., pp. 7-10. 82Albert R. Tims and Steven H. Chaffee, "A Test of the Cumulative Acquisition Model of Adolescent News Media Use," paper presented to the Association for Education in Journalism, Corvalis, Oregon, 1983. 33Campbell, "Adolescent Socialization." 8"L. Erwin Atwood, "Perception of Television Programming Preferences Among Teenagers and Their Parents," Journal of Broadcasting 12 (Fall 1968): 377-3880 85John D. Abel, "The Family and Child Television Viewing," Journal of Marriage and the Family 38 (May 1976): 331-335. )85Children and Newspapers (New York: Newspaper Advertising Bureau, 1980. CHAPTER III METHOD This chapter comes in three sections and provides details about how the study was conducted. The first part deals with data collection (see Appendix A for information on procedures prior to data collection). The second section explains concepts being used in the study, including modeling, media behavior, and the co-orientation concepts--agreement, congruency, and accuracy. In some instances the discussion will be repetitious of previous parts of the study, so a detailed analysis will not be necessary. The final section of this chapter deals with data analysis. Information will be provided about the statistical tools used in the study--correlation and partial correlation--as well as about the methods testing the study's hypotheses. Data Collection Collecting the data. The present study required the following data: (1) media use of adolescents, (2) perceptions of parental media use by adolescents, (3) perceptions of peer media use by adolescents, (4) media use of parents, and (5) media use of peers. Relationships among the first three types of data were used to determine congruency, the relationships between (2) and (4) and the relationships between (3) and (5) indicated accuracy, and the relationships between (I) and (4) and the relationships between (1) and (5) indicated agreement. 47 48 Because of the relationships under study in this research, data were collected from three groups: adolescents, their parents, and their peers. Additionally, the breadth of information needed to study the relationships recommended the administration of a survey rather than another type of design, such as experimental or a field study. The lack of an adequate sampling frame containing suitable groups of adolescents, parents, and peers made sampling on a probability basis difficult. Trying to locate an appropriate population of adolescents resulted in a similar problem. Because of these problems and economic limitations, a convenience sample of adolescents was used. The decision to use a convenience sample, of course, has implications. It becomes more difficult to generalize the findings of a study of adolescents when a sample may not be representative of the teen-aged population. A later section of this Chapter, however, will explain attempts to validate the sample by replicating parts of a study in which the sample was representative of the teen- aged population. The data collection requirements of this study necessitated this alternative method of sampling and sample validation. The first step in selecting respondents was to find a group of adolescents that could participate in all phases of the data collection. That is, the adolescents had to respond to a questionnaire and then collect data from members of the other two groups. The adolescents participating in the study were members of eight chapters of the Lansing area Explorers, the senior division of the Boy Scouts of America. The group demonstrated characteristics that were attractive to the goals of data collection for this study. First, Explorer members were all within the age category required for the research. Second, the Explorer posts or Chapters met on a regular basis, providing the researcher with the opportunity to appear 49 before captive audiences when he requested the posts to participate in the project. Third, previous contact with Explorer members indicated that they were motivated individuals, interested in participating in the project.1 The posts also conducted several fund-raising activities, which would include the present study.2 Because of the need to provide respondents with more anonymity than they would receive in a telephone or personal interview, data collection was self- administered. The technique was also effective because the questionnaire did not require probing or complicated explanations3 and because the Explorer members were to serve as both respondents and data collectors. The Explorers were given envelopes containing three questionnaires and a set of instructions (see Appendix B). The members were instructed in writing to fill out by themselves the questionnaire labeled "Media Use Questionnaire-- Adolescent." In addition, they were instructed to have their best friends complete the "Media Use Questionnaire--Best Friend" and have one of their parents complete the "Media Use Questionnaire--Parent." Both the "Best Friend" and "Parent" questionnaires consisted of the items dealing with media use for entertainment information and general newspaper and television use scales. In addition, the parent was asked several questions dealing with personal characteristies, such as income and education, that could have been used as control variables in subsequent analyses. The Explorer members were also instructed to provide their names and phone numbers and the names and phone numbers of their best friends. This information was used for spot-checking questionnaires and response verification purposes only.“ Data collection took place in January and February 1985 with 87 completed sets of questionnaires out of 122 distributed or 71 percent. To validate the sample, items used in other studies of adolescent media use were included in the 5O questionnaire.5 Convergent results in this study with those in the previous research would provide validity. That is, if evidence yielded by administering the instrument to different groups in different places yielded dissimilar results, one could account for the differences either through the non-validity of the instrument or the sample. Similar results, then, would support the notion of validity for the instrument and the different sam ples.6 The results are reported in the next chapter. Even with added validity from convergent results with other studies, the sample still has limitations, both in terms of Size and representativeness. A larger sample of adolescent-parent-peer triads would have alleviated several of the methodological issues that can be raised with the study. Small sample size will often reduce variance which, in turn, can reduce the number of significant relationships. In addition, a larger sample could have contributed to greater reliability in media use scales. Because of the small size in the study, the number of points on the media frequency of use scales was increased to increase variance. Increasing the scale points, however, also resulted in a sacrifice of scale reliability.7 Increasing the sample size could have resulted in the necessary variance with fewer scale points (perhaps five or seven instead of eleven) and, ultimately, greater reliability in the measure. The issue of representativeness in the sample is raised because of the income and educational characteristics of the parents of adolescent respondents. The mean income for the sample of parents was higher than that of families in Lansing, and a greater percentage of parents in the study had high school degrees than Lansing adults. Income and education are related to the media behaviors of adults,8 and evidence exists that lower parental socioeconomic status is related to adolescents' imitating the media behaviors of their parents.9 For this study, heavy media users, particularly newspaper readers, were overrepresented among 51 parents. Whether these factors affected the specific relationships being researched here remains unknown. In future research, however, adolescent respondents should be Chosen from a population where control over representativeness is more attainable, such as a high school student body. At that point data analysis could include a test of differences between different socioeconomic groups. Definitions Adolescent. A person 13 to 18 years old. To determine whether respondents or their same—age peers were in this age range, they were asked to indicate their ages on a questionnaire. Co-orientation concepts: agreement, congruency and accuracy. The previous chapter provided detailed background on the co-orientation approach, its three major constructs, and their conceptualizations. Here the conceptual definitions will be briefly repeated and operational definitions will be offered (also see Figure 1). ADOLESCENT < AGREEMENT ———> PARENT OR PEER MEDIA BEHAVIOR [MEDIA BEHAVIOR] CONGRUENCY ACCURACY PERCEPTION OF MEDIA BEHAVIOR I Figure 1. Co—orientation concepts: agreement, accuracy and congruency. 52 Aggeement is the relationship between one person's cognitions and another person's cognitions, or, in the case of the present study, between one person's media behavior and another person's media behavior. The relationship between cognitions and behaviors in this study comes from research that indicates a high relation between cognitions and behaviors that are prototypic or most representative of their particular category. Conversely, behaviors that are not prototypic will not be related to the cognitions.10 To put it another way, an idea in one's mind that one behave a certain way is more likely to result in that behavior than a notion that one will not or than no notion at all.11 In this study, the relationship between two persons' media behaviors was measured by com paring the behavior of the adolescent respondent with the behavior of either the parent or best friend who completed the questionnaire. For example, each adolescent respondent indicated on a 0-to-10 scale how often he/she used newspapers to get information about movies, as did the parent and peer respondent for that adolescent. If the responses between the adolescent and the parent or peer were comparable, high agreement would exist. If the responses were not comparable, little or no agreement would exist. The second co-orientation concept, congruency, is the relationship between a person's media behavior and what that person perceives to be the media behavior of another person. The relationship was measured in this study by asking adolescent respondents their media behaviors and their perceptions of the media behaviors of their parents and peers. For example, each adolescent respondent was asked to indicate how much he/she uses newspapers to get information about movies. They also were asked to indicate how much they believe their parents and best friencb use newspapers to get information about movies. Comparable responses for the adolescent and the adolescent's 53 perception of another person (parent or peer) would indicate high congruency. The converse would indicate little or no congruency. The final co-orientation concept is accuracy, the comparison of a person's perception of another person's behavior and the other person's actual behavior. In this case, measurement was made by comparing the adolescent respondent's perception of his/her parent's or peer's media behavior with the actual behavior of the parent or peer. If an adolescent respondent said his/her parent ‘uses newspapers a great deal to get information about movies and the parent actually does use newspapers a great deal to get information about movies, the inference is that accuracy exists. If the parent does not use newspapers a great deal to get information about movies, but the adolescent said the parent does, accuracy would not exist. Media behavior. Media behavior is operationalized in two ways, although in both cases it is conceptualized as the frequency of use of media. The first operational definition pertains to the frequency of use of media for specific types of information. For this study, "specific types of information" refers to information about movies and musical performers or, generally speaking, entertainment information. Use of the entertainment context relates to the section in the last chapter dealing with relating media behaviors with other types of behavior. Entertainment was used in this study because adolescents share entertainment experiences with both their parents and their peers. Furthermore, studies have shown teen-agers are attracted to media for this type of information. For example, teen-age newspaper readership is high for music columns, movie advertisements, and movie reviews.12 Respondents answered questions about how frequently they used four media (newspapers, television, radio, and magazines) and interpersonal sources to acquire entertainment information, using a 0-to-10 scale (0 = never, 10 = always). 54 The second method to identify media behavior also employs the "frequency of use" conceptualization, but pertains to the general use of two media-— newspapers and television. That is, indices used in previous research” were used to measure newspaper and television habits of respondents. Again, 0-to-10 scales were usedfifor both the seven-item newspaper-reading index and the nine- item television-viewing index. For analysis, the items in the indices were measured individually and as part of an overall scale measuring the use of the two media. Respondents indicated on a 0-to-10 scale (0 = never, 10 = always) how often they read the following in a newspaper: . 1. news about government and politics, 2. news about the economy, 3. sports news, 4. adverti sem ents, 5. comics, 6. entertainment news, and 7. crime news. Likewise, respondents indicated how often they watched the following on television: 1. national news, 2. local news, 3. sports events, 4. movies, 5. variety shows, 6. cartoons, 7. police and adventure shows, 8. comedy shows, and 9. soap operas. 55 An ll-point scale was used for all media-use measurements in this study in an attempt to increase variance.” Although the question of whether to use an odd number of steps was not at issue (the choice of neutrality, the midpoint of odd-stepped scale, was not relevant), anchoring the scale with 0 and 10 was. It was believed respondents relate 0 to "never" more than they would 1 and that they would relate 10 to "always" more so than another number. Modeling. Research on modeling of media behaviors has relied on a similarities conceptualization that required only that the behaviors of one person be similar to another's. The existence of a relationship between one person's behavior and another person's behavior allowed an inference that imitative behavior-and, therefore, modeling--existed. This conceptualization and the operational definitions derived from it, however, allow too many alternative explanations, including coincidence and reverse modeling (a parent modeling a Child's behavior). This study takes into account the adolescent child's perception of his/her parent and also requires that imitative behavior based on an accurate perception exist for modeling to occur. Modeling, then, is conceptualized as behavior imitative of another and based on accurate perceptions of another. The operationalization of modeling is the existence of agreement and accuracy. If the two conditions exist, modeling would also exist. Furthermore, if agreement and accuracy exist, congruency would also exist. In this study, this situation would occur as follows: adolescent media behavior and parental media behavior would be related, as would adolescent perception of parental media behavior and parental media behavior. Because parental media behavior would be related to both adolescent behavior and adolescent perceptions, these two variables would be related to each other. The co-orientation constructs and their structural relationships could be interchanged, and the conditional nature of the typology 56 would still exist. That is, if agreement and congruency exist, so does accuracy. Or if accuracy and congruency exist, so does agreement. Data Analgis Like other studies of adolescent media use and modeling, the present study will use correlation as a major statistical tool. For example, the study will determine the degree of congruency by correlating adolescent perceptions with adolescent behavior, agreement will be the correlation between adolescent behavior and the behavior of another (either parent or peer), and accuracy will be measured by the correlation between the behavior of another and the adolescent perception of that behavior. The analysis will differ from other modeling studies, however, by using partial correlation as well. Because of the similarities conceptualization in previous studies, correlation was an adequate statistical technique. The reconceptualization of modeling as an imitative behavior based on an accurate perception of another makes partial correlation more appropriate than correlation. As proposed earlier, operationalizing this reconceptualization means adolescent perception of another's behavior (Y) serves as an intervening variable between another's actual behavior (X) and the adolescent's behavior (2) (see Figure 2) . X = another's (parent's or peer's) media behavior Y = adolescent's perception of another's behavior Z = adolescent's media behavior Figure 2. Model illustrating intervention of adolescent perceptions. 57 The diagram indicates how partial correlation can be used. First, if one's perceptions are intervening between another's behavior and one's own behavior and enhancing the relationship between the latter two variables, partial correlation will help reveal this intervention. The technique will statistically remove the effect of the intervening variable. If the intervention exists, the relationship between another's behavior and adolescent behavior will decrease in this case. Referring to the diagram, then, X is related to Z through Y. Or another's media behavior is related to the adolescent's media behavior through the perception of another's behavior by the adolescent. In causal terms, the media behavior of another would cause perceptions of that behavior which, in turn, would cause adolescent behavior. By statistically removing the effects of the perception with partial correlation, one would expect the effects of another's behavior on adolescent behavior (X > > > Y) to be reduced. As mentioned in Chapter 11, however, the relationships being tested are not being considered causal on an a priori basis but, rather, associational. Correlational data cannot show causality, but can allow one to make "inferences concerning the adequacy of causal models."15 Such is the case here: if the data are inconsistent with the diagrammed models, those models can be eliminated. These methodological techniques, however, only indirectly involve testing the hypotheses stated in the previous chapter. As stated, the first two hypotheses, if supported on an individual basis, would not indicate the possibility of modeling. For example, the first two hypotheses deal with two conditions for modeling, accuracy (H1) and agreement (H2). That is, similarities in behavior (agreement) based on accurate perceptions (accuracy) provide the situation in which modeling could exist. Agreement without accuracy would not indicate modeling, but merely a similar behavior. The similar behavior, then, would perhaps be the result of other variables, such as family socioeconomic status and 58 opportunity. Accuracy without agreement, on the other hand, would mean no influence at all. Partial correlation, as described above, will allow analyses of these interrelationships to test whether these conditions for modeling exist. The scope of the study, however, does not go beyond testing for the existence of modeling. That is, the study does not test the necessity of modeling for the use of media by adolescents, although it assumes modeling as a sufficient reason for adolescent media use. Oncejthe existence of modeling has been established, however, additional analyses must be performed to test the hypotheses. Tests of the first two hypotheses, however, can be performed irrespective of the existence of modeling. They state merely that the agreement and accuracy relationships will be stronger between adolescents and parents than between adolescents and peers. To test these hypotheses, a significance test for the difference between independent correlations was done.16 The test allows the computation of a Z statistic that can determine whether two correlation values from different samples are likely to be truly different. A test for independent samples was used because the units of analysis for the test--the adolescent/parent dyad and the adolescent/peer dyad--are different. A test for differences between dependent correlations would require the variables to come from the same unit of analysis. For example, the test could be used to see if differences exist between the correlation of adolescent media use and grade point average and the correlation of adolescent media use and birth order. In each case, the unit of analysis is the adolescent; therefore, the correlations come from the same sample. Although the first two hypotheses can be tested regardless of whether modeling exists, the third hypothesis, which states that adolescents will model their parents more than their peers, can be supported only when support exists for the first two. The various tests--correlation, tests of significant differences ’59 between correlations and partial correlation--will be performed for each of the aforementioned media behaviors. 6O NOTES 1In fact, this relates to the concerns expressed above on threats to inference resulting from a convenience sample. Personal observation and discussion with Explorer officials (conversation with Reid Smalley, executive director of the Lansing Explorers, June 17, 1985) indicate Explorer members are probably more achievement oriented and goal directed than normal teen-agers. The primary threat from this situation is, again, to the generalizability of the study. 2Advisers of the Explorer posts were contacted to obtain permission to attend the post meetings and to solicit participation in the project. In all instances, the advisers granted permission and the researcher attended meetings. of the eight Explorer posts, which all had career orientations (for examples, members who were interested in medical careers met an an area hospital; those with law enforcement career interests met at a police department). 3Fred N. Kerlinger, Foundations of Behavioral Research, 2nd ed. (New York: Holt, Rinehart and Winston, Inc., 1973), p. 487. “Explorer members were asked to return the three questionnaires in their envelopes at the next post meeting, two or three weeks later. Each individual Explorer who successfully completed the three-part task received $5, although in some instances the money went to the Explorer post for fund-raising. Several members of various posts received packets of questionnaires at the first meeting but missed the second meeting. In these cases the researcher followed up in a third meeting to collect completed sets of questionnaires. In addition, members who missed the researcher's first appearance but attended the second meeting received questionnaires that were returned at the researcher's third visit. 5Brad1ey S. Greenberg, "Gratifications of Television Viewing and Their Correlates for British Children," in The Uses fiof Mass Communication, ed. Jay Blumler and Elihu Katz (Beverly Hills: Sage Publications, 1974); and Alan M. Rubin, "Television Use by Children and Adolescents," Human Communication Research 5 (Winter 1979): 109-120. ‘ 6K erlinger, Foundations, p. 462. 7The alpha reliability estimate for the seven-item newspaper reading index was .61, less than the minimum criterion for reliability. The alpha estimate for the nine-item television viewing index, .70, met the criterion for reliability. See Jum C. Nunnally, stchometric Theory, 2nd ed. (New York: McGraw-Hill, 1978), p. 245. 8Leo Bogart, Press and Public (Hillsdale, NJ: Lawrence Erlbaum, 1981). 9For example, studies have found relationships--either direct or indirect- between low family incomes and media behavior imitative of parents. But the relationships had nothing to do with how adolescents perceived their parents or whether they found parental behavior desirable. See Eleanor Maccoby, "Why Do Children Watch Television?" Public Opinion Ouartefl 18 (Fall 1954): 239-244; and Steven H. Chaffee, Jack M. McLeod, and Charles K. Atkin, "Parental 61 Influence of Adolescent Media Use," American Behavioral Scientist l4 (January/February 1971): 323-340. loSusan T. Fiske and Shelley E. Taylor, Social Cognitions (Reading, MA: Addison-Wesley, 1984), pp. 370-3 71. llJoseph Woelfel and Edward L. Fink, The Measurement of Communication Processes (New York: Academic Press, 1980), pp. 162-165. 12Children, Movie-Going and the Newspaper (New York: Newspaper Advertising Bureau, 1979); and Children and Newspapers (New York: Newspaper Advertising Bureau, 1980). l3George Moschis and Gilbert Churchill, "Consumer Socialization: A Theoretical and Empirical Analysis," Journal of Marketing Research 15 (November 1978): 599-609. l“Jum C. Nunnally, Psychometric Theory, 2nd ed. (New York: McGraw- Hill, 1978), PP. 245. l5Hubert M. Blalock, Jr., Causal Inferences in Nonexperimental Research (New York: W. W. Norton 6: Company, 196‘0, pp. 62. 16Jacob Cohen and Patricia Cohen, Applied Multiple Regression_/Corre1ation Analysis for the Behavioral Sciences (Hillsdale, NJ: Lawrence Erlbaum Associates, 1975), PP. 50-53. CHAPTER IV RESULTS This chapter presents the findings of data analysis described in the previous chapter, including tests of the hypotheses and analyses resulting from the reconceptualization of modeling. In addition, data reported here will describe the sample of adolescents and their parents; compare media use of adolescents, their peers and their parents; and attempt to validate the sample of adolescents by comparing it with other samples of adolescents in connection with uses and gratification measures. Description of the Samples The adolescents. Nearly two-thirds of the adolescents surveyed were young women, seven of 10 were at least 16 years old, and almost all were in senior high school (grades 10 to 12). The mean age of the sample was 16.6 years (see Table l). The parents. Table 2 provides demographic information about the parents of adolescent respondents. In general, the parents tended to be more "upscale" than adults in the Lansing metropolitan area; that is, they had more education than adults in Lansing 25 years and older, and they reported household income greater than the mean income of households in Lansing. For example, 76.2 percent of Lansing adults 25 years old and over are high school graduates; 96.6 percent of the parents in the study had high school degrees. Mean income for families in Lansing is $24,685}; for respondents' families, the mean income was $3 6, 766. 62 63 TABLE 1 Adolescent Demographic Inform ation A. Gender of Respondents n PCT. Male 33 37.9 Female 54 62.1 TOTAL 87 100.0 B. Age of Respondents n PCT. 13 years old 1 1.2 14 years old 2 2.3 15 years old 22 25.3 16 years old 31 35.6 17 years old 21 24.1 18 years old 10 11.5 TOTAL 87 100.0 C. Grade of Respondents n PCT. Seventh grade 1 1.2 Eighth grade 1 1.2 Ninth grade 2 2.3 Tenth grade 34 39.1 Eleventh grade 25 28.7 Twelfth grade 23 26.4 Missing data 1 1.2 TOTAL 87 100. 0 64 TAB LE 2 Demographic Characteristics of Parents A. Gender of Parents n PCT. Male 24 27.6 Female 63 62.4 TOTAL 87 100.0 B. Age Level of Parents n PCT. 34 years or less 4 4.6 35 to 49 years old 72 82.8 50 years and over 10 11.5 Missing data 1 1.2 TOTAL 87 100.0 C. Education Levels of Parents n PCT. Less than a high school degree 3 3.4 High school degree 23 26.4 Some college 31 35. 6 College degree 13 14.9 Some graduate work 5 5.7 Graduate degree 10 11.5 Missing data 2 2.3 TOTAL 87 100.0 (continued) 65 D. Income Level of Parents n PCT. 0 to $15,000 6 9.2 $15,001 to $35,000 28 43.1 $35,001 and more 31 47.7 TOTAL 65 100.0 Comparison of Sample with Previous Study Because a convenience sample was used in this study, an attempt was made to validate the sample by comparing it with an adolescent sample used in another study.l This was done by administering and factor analyzing 23 uses and gratifications items that started with the words "I watch TV . . . . " Each item completing the sentence represented a reason for watching. In a replication of Rubin's study, respondents indicated "how much each reason is like you" by using one of four response options: a lot, a little, not much, and not at all. Principle factor analysis with iterations was then performed on the data, using a varimax rotation. Also in replication of Rubin, a factor had to explain 18 percent of the variance in a particular item, while another factor could explain no more than seven percent of the same variable. In other words, an item with .42 loading on one factor could load no higher than .27 on another. Additionally, a minimum criterion of two principle loadings per factor was used. Using these criteria, six factors were revealed (see Table 3), with the basic structure generally consistent with Rubin.2 Some exceptions were noted, however. The reason "when there's no one to talk to or play with" had significant factor loadings on "pass tim e/habit" (.57) and "companionship" (.47), but loaded significantly only on "com panionship" in Rubin's 66 we. a we. 1- No. 1 cm. 5. 3 mm. 833 ocov rco>mc H mmcfp ow 8. Bo: :32 :8 _ 0m 8. 1 2. i N“. me. E. no. .. 3.03 of E wcmcoaam; mchu Scam Emo— cmu _ am we. 1 D. 3. mm. g. me. a: 2: 9:? t 33802 n _. mo. 3. a 0.. Nm. 3. 9: motuxo t 0383 :. no. no. 6. mm. 2. wczzcfi 9: 8:83 mN. mm. 3. mo. 1 a. mm. ow 3 chHoEOm 0E 93% z omamoon mo. m L. n _. no. 3. Nu. xmkm 083 on» 883 : 3233 mm. s. 3. .. No. 1 mm. on . not; 3 ox: «mi _ 0.0.383 no. N_. s we. 5. .. No. i on. ou 8. L033 wcfuoc o>mz _ con? wm. no. .. me. No. .. mm. mm. «Em: m M: 3:88 .. . i 683 poms". yam—om QEchEEEoU wEEmoq 13:92 «Em—LEE: «6mm cofimuom metm> :23 $5284 ..oUmL m mam»; 67 R. t. on. 8. 1. on. fl. 9:8 E; :23 E0: >956 How cmo _ 0m 2. no. No. B. ., 3. mm. €2,062 95 328 Scam ”romeo“ cmo _ 0m 3. .. ma. 2. E. t. 2. u .35 Em _ :23 :38 «E mac—mu x 33.63 mm. 8. 2. 8. .. so. 3. £63 2: no so. 2: E9; imam wow :8 _ am so. . NN. an. 2. z. B. £2838. $3 oE 8x2: t 0383 am. 9. R. no. 5. 2. 253 on. Eco? _ cm i. 1 am. mo. E. mm. B. o>mn _ 9:0an oEmm of :33 Eve 0303 .850 26; £65 3 3:6qu mo. . no. a No. 1 mm. No. 1 mm. .. Bozo... 5 Eng race _ mchu. 0.: 8663 t 0383 5. om. _ we. no. no. a me. i 2: 3 comm—m; 2:8 was? «none 562 emu _ am No. 3. mo. mm. mu. mo. £9...an Scam wwfif Emo— 08 33; z 3:83 ooo>p—| 50H“;— yomLom cam—om EcmcomcmanU wEEmoA 339:1 “Ear? EC. mama voucflcoo .m 2an 68 study. It did not, therefore, meet the criteria for inclusion in a factor in this study, but did in the Rubin analysis. The item "so I can get away from the rest of my family" loaded with the "relax" factor in the present study, but was one of three items loading on Rubin's "forget" factor. And "because it's a pleasant rest" loaded on Rubin's "relaxation" factor, but failed to meet the criteria for the analysis in this study. Despite these deviations from the Rubin study within separate factors, the data reported here and those reported previously were sufficiently convergent to lend validity to this sample of adolescents. Like the Rubin study, six factors were revealed in the present research. Three of the six consisted of the same items as corresponding factors in Rubin's study. Two of the other three factors deviated on one of the three items, while the last factor deviated on two of three. Comparison of Media Use among Adolescents, Peers and Parents This section provides an overall look at the frequency of getting entertainment information from a variety of sources. It compares adolescents with their peers and their parents over frequency of use, which was measured with a O-to-lO scale (0 = never, 10 = always). Table 4 shows no differences existed among adolescents, peers, and parents for obtaining movie information from newspapers or magazines. And all three groups used newspapers for movie information to a comparable degree. Adolescents and their peers said they used other people for movie information more frequently than did the parents of adolescents. The same situation applied to the use of television to get movie information: parents of adolescents used the medium less frequently than their children and their children's peers. Adolescents and their peers also used radio to a greater extent for movie information than did the parents of adolescent respondents. 69 TABLE 4 Frequency of Media Use for Movie Information Means* (Standard Deviations) m Adolescents Egg Eggn—ts Newspapers 7. 743 (2.46) 7. 573 (2.92) 7. 77a (2.87) Television 5.6la** (3.06) 5. 87a*** (2.89) 4.2213 (3.21) Radio 3.34a** (3.09) 3.47a** (3.06) 2.01b (3.07) Magazines 1.813 (2.45) 2.34a (2.71) 2.21a (2.86) Other people 7.19a*** (2.46) 6. 99a** (2.48) 5.63t> (2.85) *Measured on a 0-to-10 scale NOTE: Media use measured on a 0-to-10 scale. NOTE: Within each row, means with different subscripts differ from each other: *p < .05 **p < .01 ***p < .001 The differences were more pronounced among adolescents, peers and parents for getting information about music (see Table 5). No differences existed between adolescents and their peers for any source of music information, while adolescents differed from their parents for all five sources. They used television, radio, magazines, and other people more than did parents, but used newspapers less than their parents did. Similar differences were discovered between peers and parents. 70 TABLE 5 Frequency of Media Use for Music Information Means (Standard Deviation) Sgur_c_e_ Adolescents P_e_e;s_ 3% Newspapers 3.26a (3.33) 3.403 (3.11) 5.34b*** (3.36) Television 5,433“ (3.56) 5.473“ (3.11) 4.10,, (3.34) Rad“ 8'02““ (2.84) 7. 935,4H (3.02) 4.681, (3.56) Magazines 4103* (3.62) 3.91,} (3.49) 2.99,) (3.03) Other people 7-35a (2.81) 6.983*** (2.79) 4.46,, (3.01) NOTE: Within each row, means with different subscripts differ from each other: *p < .05 **p< .01 ***p< .001 In reading newspapers, the parents of adolescent respondents read with greater frequency than did the adolescents and their peers (see Table 6). The data indicate also that adolescents and their peers tend to read comics and entertainment news more frequently than adolescents' parents, while the parents of the adolescent respondents used the newspaper for government and political news and economic news to a much greater extent than did the adolescents. The differences between the adolescent groups and the parent group were also consistent for the data on television use (see Table 7). Adolescent respondents and their peers indicated greater use of television for information about movies and music than did the parents of adolescents. And both adolescent groups indicated greater overall frequency of television viewing than 71 TABLE 6 Frequency of Newspaper Reading Means (Standard Deviations) Item Adolescents Peers Parents Government and politics 3.12a (3.00) 2.54,1 (2.48) 6. 66b*** (2.87) Economy 2.713 (2.73) 2.213 (2.17) 6. 52b*** (2.54) Sports 4.893 (3.59) 5-20a (3.69) 4.438 (3.79) Advertising 4.47;:1 (3.05) 4.333 (3.10) 6. 07b*** (3.23) Comics 7. 54am” (2.89) 7.47311” (3.09) 4.321, (4.01) Entertainment 6. 60a** (3. 06) 6. 903** (2.74) 5.701, (3.02) Crime 6. 04a (3.12) 5. 463 (2.97) 7.13,,M (2.72) Summated scale 35.17a** (11.84) 34.95a** (12.95) 40.781; (12.45) NOTE: Within each row, means with different subscripts differ from each other: *p < .05 **p < .01 ***p< .001 the parent group. Once again, differences between the parent group and the adolescent groups were associated with content. Parents were more frequent viewers of national and local news, but less frequent viewers of entertainment programming than the adolescents. 72 TABLE 7 Freqiency of Television Viewing Means (Standard Deviations) Item National news Adolescents 4.67a (3.33) Peers 3. 93a (2.94) Parents 7.52b*** (2.96) Local news 5.333 (3.12) 5.143 (2.92) 7. 75b*** (2.94) Sports 4.83a (3.31) 5. 7613* (3.34) 4.5421 (3.43) Movies 7.72a*** (2.47) 7.93a*** (2.25) 6.27a (2.74) Variety shows 6.823*** (2.85) 6. 703“ (2.82) 5. 20b (3.16) Cartoons 3.723*** (3.25) 3.64a*** (3.16) 1.59a (2.44) Police and adventure 6. 833*)” (3.12) 6.35a** (3.26) 4.89b (3.40) Comedy 7.27a*** (2.75) 7.29a*** (2.50) 5.581) (3.05) Soap operas 5.51a*** (4.21) 5.23a*** (4.41) . 2°70b (3.64) Summated scale 52.703,H (15.65) 51.96,,” (13.98) 45.90,, (16.96) NOTE: Within each row, means with different subscripts differ from each other: *p<.05 **p<.01 ***p < .001 Tests of the Hypotheses The first hypothesis dealt with adolescents' correct perceptions of the media behaviors of their parents and peers. H1: Adolescents will correctly perceive the media behaviors of their parents to a greater extent than-they correctly perceive the media behaviors of their peers. 73 The second hypothesis involved adolescents' behaving similarly to their parents and peers in terms of media behaviors. H2: Adolescents will behave in a similar fashion with their parents to a greater extent than they will with their peers. The hypotheses were based on research that indicated that parents remained influential in shaping the lives of their adolescent children, even though adolescents spent more time with their same-age peers than they did in earlier stages of life.3 Other research also provided some basis for the hypotheses in terms of media behaviors.“ Correct perception, or accuracy, was measured with a Pearson product- moment correlation between adolescents' perceptions of parental (or peer) media use and actual parental (or peer) media use. The test of the hypothesis, then, required that accuracy as measured by the correlation coefficient be higher for the parent than for the peer. A significance test for the difference between independent correlations was then performed to determine whether the differences in correlations were statistically significant at the p< .05 level. The test yielded a 2.: statistic, so a critical value of 1.65 was necessary to be significant at the .05 level using a one-tailed test (the direction of the difference is predicted). Similar tests were used for the second hypothesis, but the correlation coefficients being tested were for the relationship between adolescent media use and parental (peer) media use. The two least important sources of information for adolescents were omitted from the analysis: radio and magazines for obtaining movie information and newspapers and magazines for getting music information. Accuracy/ movie information. The first hypothesis was supported in terms of the direction of the prediction for three of the five information sources for movies (see Table 8). The differences, however, were not significant at the .05 74 level; therefore, it cannot be said that adolescents were accurate in their perceptions of parents to a greater extent than they were of peers in obtaining movie information from newspapers, television and other people. TABLE 8 Com paring the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior Movie Information Source Peers Parents 2 Significance Newspapers .35 .49 1.07 ms. Television .18 .21 .19 ns. Other people .30 .46 1.18 ns. TABLE 9 Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions of Parent Media Behavior Music Information m Keg} Parents 2 Significance Television .19 . 49 2.1 4 p . 02 Radio .39 .24 1.04 ns. Other people .17 .47 2.11 p . 02 Accuracy/music information. Two significant relationships emerged from the data on getting music information, both in the hypothesized direction (see 75 Table 9). Adolescents were more accurate in perceiving parents than peers for getting information about music from television and from other people. In both cases the differences between the correlations were significant at the .02 level. Accuracy/newspaper reading. Adolescent accuracy of parents was greater than that of peers with respect to reading comics in newspapers (see Table 10). None of the other differences were statistically significant. TABLE 10 Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior Newspaper Reading It_erfl 133351 Parents g Significance Government and politics .57 .44 1.09 ns. Economy .49 .29 1.47 ns. Sports . 62 . 54 . 75 ms . Advertising .2 9 . 43 l. 00 n .s . Comics .28 .57 2.23 p .01 Entertainment .34 .43 .66 ns. Crime .51 .44 .56 ns. Summated scale .52 .53 .00 ns. Accuracy/television viewing. Four of the differences in adolescent accuracy of television viewing were statistically significant, all in the hypothesized direction (see Table 11). Adolescents were more accurate in their perceptions of parents than of peers for national news, local news, movies, and 76 comedy shows. And for the nine-item summated index of television viewing, adolescents more accurately perceived their parents than their peers. This difference was also statistically significant. TABLE 11 Comparing the Correlation Between Adolescent Perceptions and Peer Media Behavior with the Correlation Between Adolescent Perceptions and Parent Media Behavior Television Viewing Item E_e_e_r_s_ Parents _Z_ Significance National news .38 .66 2.45 p .01 Local news .29 .61 2.55 p .01 Sports . 49 . 58 . 78 ns . Movies .23 .54 2.29 p .01 Variety shows .41 .42 .07 ns. Cartoons .44 .34 .73 ns. Police and advent ure . 61 . 51 . 91 n .s . Comedy .34 .62 2.29 p .01 Soap operas .74 .74 .00 ns. Summated scale .35 .64 2.45 p .01 Agreem ent/ movie information. None of the correlations between adolescent behavior and parental behavior was greater than those between adolescent behavior and peer behavior in getting movie information. Although parent-adolescent correlations were higher than peer-adolescent correlations for 77 getting movie information from newspapers and television (see Table 12), the differences were not statistically significant. TABLE 12 Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior Movie Inform ation M Pieg Parents 2 Significance Newspapers .00 .25 1.60 n.s. Television .10 .18 . 51 ms . Other people .09 .09 .00 n.s. Agreement/ music information. Adolescents are more similar to their peers than to their parents in getting music information from radio (see Table 13). The difference in the correlations for getting music information from radio is significant at the .05 level. No other difference in correlations was statistically significant. Agreement/newspaper reading. In general, adolescents showed no greater similarity with their peers than with their parents for newspaper reading (see Table 14). Adolescent-peer correlations were higher than adolescent-parent correlations for reading government and political news, economic news, sports, entertainment news, and crime news, but not in a manner that was statistically significant. Greater correlations with parents than with peers existed for reading advertising and comics, but again the differences were not significant. 78 TABLE 13 Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior Music Information Source Peers Parents ; Significance Television .07 .28 1.36 n.s. Radio .33 - .04 1.94 p .05 Other people .14 .22 .52 n.s. TABLE 14 Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior Newspaper Reading Item .Iie_e_rs_ Parents g Government and politics .18 . .12 .38 Econom y . 32 . 07 1. 62 Sports .27 .22 .33 Advertising .14 .32 1.19 Comics - .05 .19 .88 Entertainment .28 - .05 1.49 Crime .33 .23 .68 Summated scale .17 .05 .76 Significance n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 79 Agreement/television viewing. For watching television, adolescents showed greater similarities with their parents than with their peers for watching national news and comedy shows (see Table 15). Adolescents were more similar to peers than to parents in watching soap operas. The difference between parent and peer correlations for watching soap operas is significant at the .01 level. TABLE 15 Comparing the Correlation Between Adolescent Media Behavior and Peer Media Behavior with the Correlation Between Adolescent Media Behavior and Parent Media Behavior Television Viewing m m Parents g Significance National news .00 .30 1.94 p .05 Local news - .04 .12 .51 n.s. Sports .28 .34 .41 n.s. Movies .12 .16 .25 n.s. Variety shows .28 .04 1.55 n.s. Cartoons .08 .28 1.30 n.s. Police and adventure .29 .30 .64 n.s. Comedy .04 .34 1.95 p .05 Soap operas .55 .25 2.32 p .01 Summated scale .26 .31 .34 n.s. Reconceptualization of Modeling One of the major purposes of the study was to reconceptualize modeling, particularly for media behaviors, as similar behavior based on accurate 80 perceptions. This conceptualization was developed because of the inadequacy of what can be called the similarities conceptualization used in previous studies of adolescent modeling of parental media behavior. This conceptualization asserted the existence of modeling when similarities existed between the behaviors of parents and adolescents. The new conceptualization required not only the adolescents behave similarly to others, but also that the behavior be based on accurate perceptions that adolescents have of others. As discussed previously, this conceptualization gave way to a model whereby perceptions of the adolescents intervened between the behavior of another and the behavior of the adolescent. The effect of perceptions, therefore, was to enhance the effects of another's behavior on the adolescent's behavior. If, in fact, this were the case, removing the effects of the perceptions would lessen the effects of another's behavior on the adolescent's behavior. In the present study the behavior in question-~media use-was measured at the interval level, allowing the use of partial correlation to control the perceptions variable. For example, the zero-order correlation between parental media use and adolescent media use would show similarities, but would not take into account perceptions of parental media use by the adolescent. By taking the zero-order correlation and statistically controlling for or removing the effects of the perceptions of parental media use, the degree of intervention by the perceptions variable was discovered. If perceptions of parents were intervening between parental media use and adolescent media use, then the relationship between parental media use and adolescent use should have been reduced when the effects of perceptions as an intervening variable were removed. To test these relationships, Pearson product-moment correlation analysis was conducted on the media use data. Relationships between parental media use and adolescent media use and between peer media use and adolescent media use 81 were tested. Relationships for all media use variables were analyzed, including separate analyses for every item in the newspaper-reading and television-viewing indices, as well as for the summated scales. The data for parent-adolescent relationships will be presented first. Correlations of media use for movie information--adolescents/iarents. The proposed relationships were borne out in the data for adolescent/parental media use for obtaining movie information from newspapers (see Table 16). That is, by controlling for adolescent perceptions of parental media use for obtaining movie information, reduced product-moment correlations occur between adolescent media use and parental media use. For example, the zero-order correlation between adolescent use of newspapers for getting movie information and the same parental use is .25 (p < .05); removing the effects of adolescent perception of parental use of newspapers for getting movie information reduces the correlation to .09. TABLE 16 Product-Mom ent Correlations Between Adolescent Media Behavior and Parent Media Behavior for Movie Information S_o_u5$ Zero-order Correlation Partial Correlation* Newspapers .253 .09 Television .1 8 . 09 Radio .11 - .02 M agazi nes .17 . 05 Other people .09 - .03 *Controlling for adolescent perceptions of parent behavior. ap < .05 82 Correlations of media use for music information--adolescents/parents. Again, removing the effects of adolescent perceptions of parental media use reduces the relationship between adolescent media use and parental media use for music information from television and other people (see Table 17). The zero- order correlations between adolescents and parents for obtaining music information from television (r = .28) and from other people (r = .22) are significant at the .05 level; the correlations are no longer statistically significant when perceptions are controlled. TABLE 17 Product-Mom ent Correlations Between Adolescent Media Behavior and Parent Media Behavior for Music Information m Zero-order Correlation Partial Correlation* Newspapers .12 - .01 Television .28b .12 Radio - .04 - .09 Magazine .00 - .01 Other people .22a .13 *Controlling for adolescent perceptions of media behavior. ap < .05 bp (.01 Correlations of newspaper use--adolescents/ parents. The zero-order correlations for sports, advertising, comics, and crime items of the newspaper- reading index were reduced when adolescent perceptions of parental newspaper 83 reading were controlled (see Table 18). The adolescent perceptions of parental reading of sports and advertising in the newspaper did not contribute as greatly to the adolescent-parent relationship as did the adolescent perceptions of reading comics and crime. In other words, the effects of the perceptions on the adolescent-parent relationships for crime news and comics were greater than the effects on the adolescent-parent relationships for sports and advertising. TABLE 18 Product-Mom ent Correlations Between Adolescent Newspaper Reading and Parent Newspaper Reading Item Zero-order Correlation Partial Correlation* Government and politics .12 .06 Economy .07 .00 Sports .22a .15 Advertising .32C .20a Comics .193 .05 Entertainment - .05 - .20 Crime .23a .07 Summated scale .05 - .11 *Controlling for adolescent perceptions of parent reading. ap < .05 bp < .01 Cp < .001 84 Correlations of television use-adolescents/parents. Controlling for the adolescent perceptions of parental television use reduced the relationships between adolescent television use and parental television use for the following items: national news, sports, cartoons, police and adventure shows, comedy shows, soap operas, and the overall summated scale (see Table 19). This indicated, again, that perceptions of parents enhanced the similarities adolescents have with parental television viewing. Variations existed, however, in the degree of the effects of the perceptions. For example, for watching national news, removing the effects of the adolescent perceptions reduced the correlation between parental use and adolescent use from .30 (p < .01) to .08. But removing the effects of the perception of watching sports reduced the correlation between adolescent sports viewing and parental sports viewing from .34 (p < .001) to .19 (p < .05). And for soap operas, the relationship dropped from .25 (p < .01) to .22 (p < .05). Correlations of media use for movie information--adolescents/peers. Although the proposed relationships were borne out again for the media use data relating adolescent media use with that of their peers (see Table 20), none of the zero-order correlations between adolescent use and peer use was statistically significant in the movie information categories. This means that the relationships between adolescent and peer media use for getting movie information were essentially zero. 85 TABLE 19 Product-Moment Correlations Between Adolescent Television Viewing and Parent Television Viewing Item Zero-order Correlation Partial Correlation* National news .30b .08 Local news .12 .09 Sports .34C .19a Movies .16 .11 Variety shows .04 .22 Cartoons .28b .1 5 Police and adventure .30b .18 Comedy .34C .07 Soap operas .25b .22a Summated scale .31b .00 *Controlling for adolescent perceptions of parent viewing. 3p < .05 bp < .01 (:p < .001 86 TABLE 20 Product-Mom ent Correlations Between Adolescent Media Behavior and Peer Media Behavior for Movie Information S_ot_1_r_cg Zero-order Correlation Partial Correlation* Newspapers .00 - .16 Television .10 . 00 Radio .14 .02 Magazines .18 .09 Other people .09 - .03 *Controlling for adolescent perception of peer behavior. Correlations of media use for music information--adolescents/peers. Removing the effects of adolescent perceptions of peer media use for music information reduced the relationship between adolescent use and peer use for getting information from radio (see Table 21). The zero-order correlation between adolescents and peers for radio use for music information (r = .33) was statistically significant at the .001 level. The partial correlation (r = .10) was not statistically significant. Correlations of newspaper use--adolescent/ peers. For the government and politics, economy, sports, entertainment, and crime items of the newspaper- reading index, removing the effects of adolescent perceptions of peers reduced the relationship between adolescent newspaper reading and peer newspaper reading (see Table 22). The effects of perceptions were particularly noticeable for reading sports in the newspaper. The zero-order correlation between adolescent reading of sports and peer reading of sports in the newspaper was .27 87 TABLE 21 Product-Moment Correlations Between Adolescent Media Behavior and Peer Media Behavior for Music Information m Zero-order Correlation Partial Correlation* Newspapers .17 - .05 Television .07 - .04 Radio .33C .10 Magazines .16 - .01 Other people .14 .09 *Controlling for adolescent perceptions of peer behavior. Cp <.001 (p < .01). Controlling for perceptions of peers, the correlation was .01, indicating that perceptions accounted for much of the relationship. A similar situation existed for reading government and political news in the newspaper. The situation was different, however, for reading economic news, entertainment news, and crime news. For those items, significant zero-order correlations were reduced when perceptions were controlled, but perceptions apparently accounted for the relationship to a lesser extent. 88 TABLE 22 Product-Mom ent Correlations Between Adolescent Newspaper Reading and Peer Newspaper Reading Item Zero-order Correlation Partial Correlation* Government and politics .18a .03 Econom y . 32b .18 Sports .2713 .01 Advertising .14 . 03 Comics - .05 - .12 Entertainment .28b .13 Crime .33C .10 Summated scale .17 .04 *Controlling for adolescent perceptions of parent reading. 3p < .05 bp < .01 Cp <.001 Correlations of television viewink-adolescents/peers. Once again, the proposed relationships were indicated when perceptions of peer television viewing were controlled (see Table 23). This was the case for the summated index and for the following items: sports, variety shows, police and adventure shows, and soap operas. In no case, however, was the relationship between adolescent use and peer use completely eliminated. For example, the zero-order correlation between adolescent viewing and peer viewing of police and adventure shows was .39 (p < .001). Removing the effects of adolescent perceptions of peer viewing of police and adventure shows reduced the relationship, but only to .20 (p < .05). This would indicate that similarities between adolescents and peers on this particular item exist for reasons other than perceptions of peers. The same holds true for the viewing of soap operas. TABLE 23 Product-M om ent Correlations Between Adolescent Television Viewing and Peer Television Viewing It_er_n_ Zero-order Correlation Partial Correlation* National news .00 .12 Local news - .04 .18 Sports .28b .12 Movies .12 .01 Variety shows .28b .11 Cartoons .08 .07 Police and adventure .39C .20a Comedy .04 .12 Soap operas .556 .32b Summated scale 2613 .11 *Controlling for adolescent perceptions of peer viewing. 3p < .05 bp < .01 Cp < .001 90 Additional Analysis As noted before, similarities existed between parents and adolescents for watching comedy and national news on television and for overall television viewing. Furthermore, perceptions of parent viewing contributed to the similarities. It is possible, however, that the lack of opportunity afforded when a household had one television contributed to the similarities. In other words, with another set the adolescent might have chosen to watch something different from the parent. Because of this possibility, additional analyses were conducted for watching comedy shows and national news and for overall television viewing. Although many reasons exist for owning more than one television sets, often a second TV allows members of a household to vary their viewing. An extra set, of course, provides the opportunity for viewing different programs. Extra sets, therefore, create the possibility of reducing parent-child correlations. Separate partial correlation analyses were performed among parent-adolescent pairs where the parent reported having more than one television in the household. It was believed that the parent-adolescent zero- order correlations would decrease in these households because the adolescent would have the opportunity to watch something different from the parent. Table 24 shows the parent-child zero-order correlations for national news viewing and overall television viewing dropped among multi-set homes, but not significantly, while the correlations for watching comedy shows actually increased. For all three viewing behaviors, perceptions of parent behavior contributed to the parent-child correlation in multi-set homes. 91 TABLE 24 Product-Mom ent Correlations Between Adolescent Television Viewing and Parent Television Viewing--Multi-Set Homes Item Zero-order Correlation Partial Correlation* National news .23b (.30 overall) .04 Comedy .37C (.34 overall) .05 Summated scale .29b (.31 overall) - .05 *Controlling for adolescent perceptions of parent viewing. bp < .01 Cp < .001 Summary of Results Tests of the hypotheses. Adolescents accurately perceived the media behaviors of their parents to a greater extent than they accurately perceived the media behaviors of their peers for the following: 1. getting music information from television and other people, 2. reading comics in the newspaper, I 3. watching national and local news, movies, and comedy on television, and 4. overall television viewing. Adolescents were more similar to their parents than they were to their peers when: -- watching national news and comedy shows on television. They were more similar to peers than parents in: 1. getting music information from radio, and 2. watching soap operas on television. 92 Reconceptualization of modelirg. Removing the effects of adolescent perceptions of parental behavior reduced the zero-order correlations between adolescent behavior and parental behavior in all instances involving media use behaviors. It must be noted, however, that many of the zero-order correlations were not statistically significant, meaning no relationship existed between adult behavior and adolescent behavior to begin with. To some extent this can be attributed to the small sample size which, in turn, reduced variance. This methodological issue notwithstanding, the proposed reconceptualization was supported by the data, although the degree of reduction varied. ‘ Examples follow. 1. When the effects of perceptions were removed, the adolescent-parent correlations for reading comics and crime in the newspaper decreased almost totally. 2. Controlling for perceptions reduced the adolescent-parent correlations for reading advertising in newspapers, but the relationship was still statistically significant. 3. Adolescent perceptions of parental behavior accounted for much of the relationship between the parent and the adolescent child with respect to watching national news and comedy on television. 4. Perceptions accounted for some of the relationship between parent and adolescent in regard to watching sports, cartoons, police and adventure shows, and soap operas. But perceptions alone did not account for those relationships. Similar relationships were observed for peer-adolescent correlations, although the aforementioned methodological issue also existed here. Even so, in every instance removing the effects of perceptions of peers reduced the peer- adolescent correlations for media use behavior. Some examples follow. 93 1. Controlling for adolescent perceptions of peer reading of government and political news, sports, and crime news in the newspaper nearly eliminated the relationship between adolescent reading and peer reading of these items. 2. Removing the effects of perceptions of peer reading of economic news and entertainment news in the newspaper reduced .the correlation between adolescent and peer behavior, but did not approach eliminating it. 94 NOTES 1Alan M. Rubin, "Television Use by Children and Adolescents," Human Communication Research 5 (Winter 1979): 109-120. 2w” pp. 112113. 3Ernest Q. Campbell, "Adolescent Socialization," in Handbook of Socialization Theory and Research, ed., David Goslin (Chicago: Rand McNally, 1969), pp. 821-859. l‘John D. Abel, "The Family and Child Television Viewing," Journal of Marriage and the Famgy 38 (May 1976): 331-335; L. Erwin Atwood, "Perception of Television Program Preferences among Teenagers and Their Parents," Journal of Broadcastifl 12 (Fall 1968): 377-388; and Children and Newspapers (New York: Newspaper Advertising Bureau, 1980). CHAPTER V DISCUSSION This chapter includes interpretations of the results and implications of the research. Incorporated in both sections are suggestions for future research. Interpretation of Results In instances where significant differences existed between adolescents' accuracy of parents and peers, adolescents accurately perceived parents more often. Significant differences where adolescent accuracy of parent was greater than that of the peer occur for eight media use items. The opposite was the case for one media item. On the other hand, when significant differences existed for adolescent agreement with parents and peers, the differences occurred twice in both directions. Adolescents agreed more with peers than with parents for two items and agreed more with parents than peers for two items. Furthermore, where adolescents' accuracy of parental behavior was greater than adolescent accuracy of peer behavior, almost always agreement with parents was not greater than agreement with peers. Accuracy was generally not followed by imitative or similar behavior. In cases where adolescent accuracy of parents was greater than adolescent accuracy of peers but no differences existed in agreement, observational opportunity might be at work. That is, adolescents have a better Chance to observe their parents' media behaviors in the home than they have to observe those of their peers. Subsequently, they can be more accurate about the adults' behaviors than the peers'. 95 96 For example, adolescents' accuracy of parental use of television to get music information (r = .49) was significantly greater than their accuracy of peer use of television to get music information (r = .19, Z = 2.14, p < .02). But adolescent agreement (similarity) with parental use of television for music information (r = .25) was not significantly greater than agreement with peers (r = .07, Z = 1.36, p > .05). In this case, adolescents could better predict their parents' behaviors than they could their peers' behaviors, but they were not necessarily going to agree with parental behavior more than with peer behavior. This was the case for almost all the media use behaviors. Adolescents were more accurate in their perceptions of parents than they were of peers, but not more similar; or adolescents are more accurate in their perceptions of peers than parents, but not more similar. Two exceptions existed: watching national news and comedy on television. Adolescents' accuracy of their parents' watching national news was greater than accuracy of their peers. Their perceptions of parental viewing of comedy shows was also more accurate than their perceptions of peer viewing of comedy shows. And in both instances, adolescents agreed more with their parents than they did with their peers. Still, this evidence did not itself indicate that adolescents model their parents' watching national news and comedy. Additional evidence from the reconceptualization of modeling indicated this possibility; that is, adolescent perceptions of parental viewing of national news and comedy did intervene with parental viewing and adolescent viewing. The zero-order correlations between adolescents and parents for viewing of national news and comedy--both statistically significant--were substantially reduced when the effects of 97 perceptions were removed. This indicated that adolescents behaved similarly to their parents based on accurate perceptions. This interpretation corresponds partially with that in a previous study by Chaffee, McLeod, and Atkin.1 Their findings indicated that as children got older, they behaved similarly to their mothers in terms of watching television news. That is, TV news viewing among ninth-graders was more similar to that of parents than TV news viewing among sixth-graders, and the amount of news viewing increased with age. But for viewing of comedy, the situation was reversed. Sixth-graders were more similar to parents than ninth-graders, and the amount of time spent watching comedies decreased as the child got older. Chaffee et a1. developed this scenario: older children, desiring to be more adult-like, increased their viewing of TV news and decreased their viewing of comedies. Parents, meanwhile, relied on the television "expert" sixth grader and watched comedies in a similar fashion (reverse modeling). The present study posits a similar interpretation to Chaffee et al. for watching television news (national news here). But evidence here supports an inference for viewing comedy shows different from the previous study. Adolescents accurately perceived their parents' viewing and behaved in a manner similar to their parents. And the perceptions contribute to the relationship between the behaviors. The evidence did not, however, indicate a scenario similar to that of Chaffee et al. Though parent-child similarities in viewing national news were greater for older children than younger ones (see Table 25), the frequency of viewing national news did not increase for older children (see Table 26). And, unlike the previous study, parent-child similarities in viewing comedies was greater for older children than for younger ones. Like the earlier study, though, viewing comedy shows was greater among younger children. 98 TABLE 25 Correlations Between Adolescents and Parents for Viewing National News and Comedy Shows, by Age 13-15 1_6_ 17-18 National news .26 .12 .58 Comedy shows .15 .35 .46 n=25 n:31 n=29 TABLE 26 Adolescent and Parent Frequency of Watching National News and Comedy Shows, by Age 13-15 1_6 17-18 National news 4.94 - 7.08 4.61 - 7.64 4.76 - 8.14 Comedy shows 7.88 - 4.80 7.42 - 6.14 6.34 - 5.59 Though the direction of the aforementioned relationships and their changes indicated a modeling inference, it cannot be said conclusively that modeling exists for these two media use items, primarily because the zero-order correlations, though statistically significant, are not great enough to indicate great similarities in behavior. At .30 and .34 for national news and comedy shows, respectively, the relationships were moderate at best.2 Evidence that might add strength to the modeling interpretation would be indications of interpersonal contact with parents or peers, whether it be the social settings in which an adolescent watches television or reads a newspaper or the conversations an adolescent has had about the media with others.3 Another way 99 to obtain stronger evidence of modeling would be to include open-ended questions that would require adolescents to go into detail about reasons they watch or read particular items. In a similar vein, personal interviews or focus groups could add depth to the data obtained from a similar study. Finally, some measurement indicating that adolescents do observe the media behaviors of their parents would add strength to a modeling interpretation. For example, asking teens when their parents read a newspaper or watch television could serve this purpose. This conservative interpretation notwithstanding, the zero-order correlations found here are higher than those in two previous studies of adolescent media modeling.“ The reason for reporting these differences is not to compare this study favorably with previous research, but to indicate the possibility of a modeling interpretation with the addition of still higher zero- order correlations or additional measures. As mentioned earlier, watching national news and comedy shows were the only two items for which adolescents accurately perceived and similarly behaved like their parents to a greater extent than their peers. In no instance was this situation reversed; that is, adolescents did not behave similarly and perceive accurately their peers to a greater extent than parents. But similar behaviors and accurate perceptions did exist for parents gig peers on some media use items. For example, the correlations between adolescents and parents for comics and crime news reading in the newspaper were both significant at the .05 level, indicating similarities. Removing the effects of adolescent perception of parents' reading comics and crime news reduced the relationship and indicated that adolescent perceptions contributed to the similarities. 100 Likewise, the correlations between adolescents and peers for getting music information from the radio decreased when adolescent perceptions of peers were controlled. This, too, implied perceptions were contributing to the similarity between adolescents and peers, this time in their use of radio to get music information. In the same manner, adolescent perceptions contributed to the similarities with peers for reading government and politics, sports, and crime news in newspapers. But does this type of relationship represent modeling? In this instance, probably not. Again, the parent-cth similarities for reading crime news and comics were not great enough to indicate imitative behavior. Nor was the peer- child similarity for getting music information from the radio. On the other hand, several instances existed for which the adolescent perceptions had little effect on adolescent-peer similarities: reading economic and entertainment news in newspapers, and viewing sports, variety shows, police and adventure shows, and soap operas on television. The adolescent-peer similarities for these media use items resulted from other reasons than the perceptions adolescents had of their peers. For example, similar socioeconomic background could have contributed to the similarity. Even when the similarities were great, they did not result from accurate perceptions, making modeling an unlikely interpretation. In conclusion, then, perceptions of peer and parent media behaviors contributed to the similarities adolescents had with peer and parent media behaviors. This indicated that adolescents were performing similar behaviors based on accurate perceptions. But because the degree of similarity was not great for most media use items, evidence of modeling, or imitative behavior, was inconclusive. Where the degree of similarity was high, the effects of the perceptions were not as great. 101 Implications of the Study This section addresses the questions raised in the first Chapter. There, implications of the research were presented in the general research framework of comparing parental and peer influence on adolescents. And specific research implications for mass media issues were raised as well. Here the implications will be discussed in light of the study's evidence and narrow scope. Adolescent culture versus adult culture? The primary general research question dealt with the existence of a youth culture that was independent of and antagonistic toward adults. The evidence presented in this study indicated such was not the case. Adolescents perceived parental media behaviors as accurately, if not more so, as peer media behaviors. And they used the media no more similarly as peers than as parents. In fact, for most items where significant differences existed in perceptions or similar behavior, the difference was in the direction of parents. Adolescents' accuracy of and similarity with parental media behavior, though not overwhelming, was still strong enough to argue against the existence of a youth culture. Instead, adolescence probably serves as a transition period in which young people make some decisions autonomously and others through the influence of parents and peers. Implications for public policy. Among the narrow issues this study was meant to address was its usefulness for public policy questions dealing with adolescent media use. Perhaps most pertinent today among these issues is the effect of television advertising of beer and wine on teen-aged alcohol consumption. Although this study did not address this issue specifically, some of its findings may contribute to that discussion. For example, because most the beer and wine advertisements appear on sports programs, it could be instructive to see whether the perception of peer and parental television viewing of sports contributed to the similarities between adolescent and parents and adolescents 102 and peers. In this study, perception of another's viewing contributed to the similarities in television sports watching adolescents had with parents and peers. In neither case, however, was the similarity great enough to posit a modeling interpretation. Still, the direction of the relationships and their changes, and previous research on parental and peer influence on adolescent alcohol consumption, have implications for future research. New studies should consider potential interrelationships among sports viewing, reading, and participating; alcohol consumption among teen-agers; and the influence of others. No attempt was made in this study to gauge effects on media use by education policy such as newspaper-in—education participation. But results of this study imply implementation of such a program could be difficult, assuming such a program incorporated the participation of parents. Adolescents in the sample are not avid newspaper readers and have little similarity with parents' newspaper reading. Of course, one of the purposes of the newspaper-in- education program is to foster newspaper reading, so one could reason that an N-I-E program would, if nothing else, improve newspaper-reading habits among adolescents. The similarities adolescents have with parents in terms of television use and the contributions adolescent perceptions made to the similarities indicate TV might be a useful educational tool in the home. Before implementing such a policy, however, obtaining additional evidence of modeling is recommended. For example, measurement of whether adolescents and their parents discuss television programs and measurement of the times when adolescents and parents watch television might add to the validity of the findings of this study. The combined findings, in turn, would hold promise for intervention strategies whereby parents could be encouraged and trained to discuss with their teen-aged children television programming content. 103 Implications for research. Perhaps the most important implication of this study is that it demonstrated the importance of perceptions in determining how the behaviors of others influenced adolescent behaviors. That is not to say that perceptions alone influenced adolescent behavior; instead, findings in this study indicated that accurate perceptions contribute to the similarities between adolescent behavior and behavior of others. The new conceptualization provides stricter criteria for the modeling inference, in essence applying a more rigorous standard to modeling research. Requiring more than an imitative behavior, however, could have drawbacks in terms of solving the problems of whether behaviors actually are observed. These questions of observational. learning raise the global issue of whether media behaviors can be modeled at all. In the case of watching television, adolescent observation of parents seems entirely possible. For example, the two behaviors for which support for modeling exists--watching national news and comedy shows on television--are typically behaviors adolescents can observe parents perform. But it is difficult to imagine adolescents observing their peers perform these behaviors. Furthermore, newspaper reading would also seem to be a difficult behavior to observe. One must envision an adolescent peering over the parent's shoulder to observe the parent reading a murder story. These problems notwithstanding, empirical support for the reconceptualization in this study provides a basis for new research schemes in the future. Some of these have already been discussed, but most important is the incorporation of a measurement of interpersonal communication with the parent or peer model. Implications for management. Consistent with other research, both academic and industrial, this study indicated that adolescents use television and radio to a greater extent than print media.5 Furthermore, it indicated the 104 possibility of adolescent modeling of parental television viewing, but not modeling of newspaper reading. In the short term, this bodes well for electronic media and not as well for print media. The findings suggest that the television industry in several ways has transcended an age gap between adolescents and their parents, while the newspaper industry has failed to develop readership of younger people in the home. The popularity of high quality, family-oriented situation comedies such as "The Bill Cosby Show" speaks to this point. Whether these short-term implications have long-term ramifications is left to speculation at this point. This study suggests that media use habits of adolescents portend another generation heavily reliant on electronic media and not on print media. 105 NOTES 1Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parent- Adolescent Similarities in Television Use," paper presented to the Radio-TV Division of the Association for Education in Journalism, Washington, DC, August 1970. 2Jacob Cohen and Patricia Cohen, Applied Multiple Reggession/ Correlation Analysis for the Behavioral Sciences (Hillsdale, NJ: Lawrence Erlbaum Associates, 1975). 3Steven H. Chaffee and Albert R. Tims, "Interpersonal Factors in Adolescent Television Use," Journal of Social Issues 32 (1976): 98-115. 4Chaffee, McLeod, and Atkin, "Parent-Adolescent Similarities"; and Steven H. Chaffee, Jack M. McLeod and Charles K. Atkin, "Parental Influence of Adolescent Media Use," American Behavioral Scientist 14 (January/February 1971): 323-340. 5Leo Bogart, Press and Public (Hillsdale, NJ: Lawrence Erlbaum Associates, 1981). APPENDIX A PROCEDURES PRIOR TO DATA COLLECTION APPENDIX A PROCEDURES PRIOR TO DATA COLLECTION To collect data for the study, clearance for using human subjects was needed. The necessary material to obtain approval from the Michigan State University Committee on Research Involving Human Subjects was submitted on June 27, 1984. It included an abstract of the project, a description of and rationale for the proposed populations, an analysis of risks and benefits from the research, the consent procedure, a signed statement from the major professor on the research, and a copy of the research proposal. The Committee approved the project on July 3, 1984. The questionnaire was pre-tested on July 12, 1984, when it was administered to 40 participants of a high school newspaper workshop at Michigan State University. The purpose of the pretest was three-fold: (a) to find out how long it took to complete the questionnaire, (2) to discover difficult or unclear wording, and (3) to determine an adequate number of steps in the measurement scales being used. The questionnaire took as few as 10 minutes and as many as 20 minutes to complete. Changes were made in the wording of the questionnaire for easier reading and clarity. And the number of steps in measurement scales was increased in an attempt to increase variance. 106 APPENDIX B QUESTIONNAIRES INSTRUCTIONS You are being asked to participate in a study that will compare teen-agers with their best friends and parents regarding the media they use. In order to do that, we need responses from all three groups, so we are asking not only your participation, but also your cooperation. In return, we are willing to contribute funds to you or your group. Here's what you have to do: 5) Complete the questionnaire labeled "Media Use Questionnaire--Adolescent." Have the person you consider to be your best friend complete the questionnaire labeled "Media Use Questionnaire--Best Friend.” Have one of your parents complete the questionnaire labeled "Media Use Questionnaire--Parent." Provide in the spaces below your name and telephone number and your best friend' 5 name and telephone number. This information will be used only to spot check some of the responses that are provided. In no wa will the information be used to personally identify respondents in the study. Return the three completed questionnairesenulthe spot- -check information in the brown envelope to your post advisor by . Your post (or you) will receive $5 for successful completion of this task. If you have any questions, please feel free to contact me (Jim Bernstein) at 485-7290. Your name and number Your best friend's name and number 107 Instructions: QUESTION. 108 ‘ MEDIA USE QUESTIONNAIRE - ADOLESCENT You are being asked to complete this questionnaire to help us find out what media teen-agers use. This is not a test and you won't be graded on your replies. Answer each question as honestly as you can, but if you are not sure about a partic- ular answer check or write the answer that come closest to what you think you know. ‘ Take as much time as you need and BE SURE TO ANSWER EVERY Do not pay attention to the numbers you see in parentheses. They are only included to assist us in processing your answers. You may begin now. Don't stop until you have answered all the questions. _-'__—'y-—"_._\ nr.‘ _ . ID No. 109 With 0 as "never" and 10 as "always," indicate on a0-to-10 scale how much you use the following when you want information about movies (circle the most appropriate answer). a . newspapers 0 1 2 3 4 S 6 7 8 9 10 b. television 0 1 2 3 4 S 6 7 8 9 10 c . radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 '3 4 s 6 7 8 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 With 0 as "never" and 10 as "always," indicate on a O-to-IO scale how much you use the following when you want information about music or musical performers (circle the most appropriate answer). a . newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 ' 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 110 With 0 as l'never" and101as "always," indicate on a O-to-loscale how often you read the following in the newspaper (circle the most appropriate answer). a. news about government and politics 0 1 2 3 4 5 6 7 8 9 10 b. news about the economy 0 1 2 3 4 5 6 7 8 9 10 c. sports news 0 1 2 3 4 5 6 7 8 9 10 d. advertisements O 1 2 3 4 5 6 7 8 9 10 e. comics 0 1 2 3 4 5 6 7 8 10 f. entertainment news 0 1 2 3 4 5 6 7 8 9 10 9. crime news 0 1 2 3 4 5 6 7 8 9 10 With 0 as "never" and10 as "always," indicate on 0-to-1OScale how much you watch the following on television (circle the most appropriate answer). a. national news 0 1 2 3 4 5 6 7 8 9 10 b. local news 0 1 2 3 4 5 6 7 8 9 10 c. sports events 0 1 2 3 4 5 6 7 8 9 10 d. movies 0 1 2 3 4 5 6 7 8 9 10 e. variety shows 0 1 2 3 4 5 6 7 8 9 10 f. cartoons 0 1 2 3 4 5 6 7 8 9 10 9. police and adventure shows (for example, "The A Team" and "Knight Rider") 0 1 2 3 4 5 6 7 8 9 10 h. comedy shows 0 1 2 3 4 5 6 7 8 9 10 i. soap operas O 1 2 3 4 5 6 7 8 9 10 111 Here are some statements a person might say about newspapers. Based on the newspapers you read, check whether you strongly agree, agree, disagree or strongly disagree with each statement. Strongly Agree (1) a. newspapers ignore controversial issues b. newspapers will not take a stand on some issues because of fear of offending people c. newspapers do not give all sides of a story d. I don't believe many stories in newspapers e. newspapers do not provide in-depth coverage on important issues f. newspapers are influenced by special interest groups 9. newspapers are influenced by advertisers h. I am skeptical of newspapers that are monopolies i. newspapers give only the news they want to give Neutral (3) Strongly Disagree Disagree (4) (5) 112 NON, WE WOULD LIKE YOU TO ANSWER THE FOLLOWING QUESTIONS AS THEY PERTAIN TO YOUR BEST FRIEND. .....fi. -._—.~...;;' — -‘ 113 With 0 as "never" and10 as "always," indicate on a0- to- 10 scale how much your best friend uses the following when he/she wants information about movies (circle the most appropriate answer). a. newspapers 0 2 3 4 6 8 9 10 b. television 0 2 3 4 5 8 9 10 c. radio 0 2 3 4 6 8 9 10 d. magazines 0 -2 3 4 6 8 9 10 e. another person or persons 0 2 3 4 6 8 9 10 With 0 as "never" and 10 as "always," indicate on a O-to-IO scale how much your best friend uses the following when he/she wants information about music or musical performers (circle the most appropriate answer). a. newspapers 0 2 3 4 6 8 9 10 b. television 0 2 3 4 6 8 9 10 c. radio 0 2 3 4 6 8 9 10 d. magazines 2 3 4 6 8 9 10 e. another person or persons 0 2 3 4 6 8 9 10 114 With 0 as "never" and 10 as "always," indicate on a 0-to-10 scale how often your best friend reads the following in the newspaper (circle the most appropriate answer). a. news about government and politics 0 1 2 3 4 5 6 7 8 9 10 b. news about the economy 0 1 2 3 4 5 6 7 8 9 10 c. sports news 0 1 2 3 4 5 6 7 8 9 10 d. advertisements 0 1 2 3 4 5 6 7 8 9 10 e. comics 0 1 2 3 4 5 6 7 8 9 10 f. entertainment news 0 l 2 3 4 5 6 7 8 9 10 9. crime news 0 1 2 3 4 5 6 7 8 9 10‘ With 0 as "never" and 10as "always," indicate on aO-to-lo scale how much your best friend watches the following on television (circle the most appropriate answer). a. national news 0 1 2 3 4 5 6 7 8 9 10 b. local news 0 1 2 3 4 5 6 7 8 9 10 c. sports events 0 1 2 3 4 5 6 7 8 9 10 d. movies 0 1 2 3 4 5 6 7 8 9 10 e. variety shows 0 1 2 3 4 5 6 7 8 9 10 f. cartoons 0 1 2 3 4 5 6 7 8 9 10 9. police and adventure shows (for example, "The A Team" and "Knight Rider") 0 1 2 3 4 5 6 7 8 9 10 h. comedy shows 0 1 2 3 4 5 6 7 8 9 10 i. soap operas 0 1 2 3 4 5 6 7 8 9 10 ANSWER QUESTION 10 AS YOU THINK YOUR'BEST FRIEND 110010. — 115 10. Here are some statements a person might say about newspapers. Based on the newspapers you read, check whether you strongly agree, agree, disagree or strongly disagree with each statement. Strongly Strongly Agree Agree Neutral Disagree Disagree (1) (2) (3) (4) (5) a. newspapers ignore controversial issues b. newspapers will not take a stand on some issues because of fear of offending people c. newspapers do not give all sides of a story d. I don't believe many stories in newspapers e. newspapers do not provide in-depth coverage on important issues f. newspapers are influenced by special interest groups 9. newspapers are influenced by advertisers h. I am skeptical of newspapers that are monopolies i. newspapers give only the news they want to give 116 NOW, WE WOULD LIKE YOU TO ANSWER QUESTIONS 11-15 AS THEY PERTAIN TO YOUR PARENTS. 11. 12. 117 With 0 as "never" and 10 as "always," indicate on aO-to-IO scale how much your parents use the following when they want information about movies (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 10 b. television 0 1 2 3 4 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 9 10 e. answer person or persons 0 1 2 3 4 5 6 7 8 9 10 With 0 as "never" and10 as "always," indicate on a o-to-loscale how much your parents use the following when they want information about music or musical performers (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 13. 14. 118 With 0 as "never" and10 as "always," indicate on a 0-to-10 scale how often your parents read the following in the newspaper (circle the most appropriate answer). a. news about government and politics 0 1 2 3 4 5 6 7 8 9 10 b. news about economy 0 1 2 3 4 5 6 7 8 9 10 c. sports news 0 1 2 3 4 5 6 7 8 10 d. advertisements 0 1 2 3 4 5 6 7 8 9 10 6. comics 0 1 2 3 4 5 6 7 8 9 10 f. entertainment news 0 1 2 3 4 5 6 7 8 9 10 9. crime news 0 1 2 3 4 5 6 7 8 9 10 With 0 as "never" and10 as "always," indicate on a 0—to-1OScale how much your parents watch the following on television (circle the most appropriate answer). a. national news 0 1 2 3 4 5 6 7 8 9 10 b. local news 0 1 2 3 4 S 6 7 8 9 10 c. sports events 0 1 2 3 4 5 6 7 8 9 10 d. movies 0 1 2 3 4 5 6 7 8 9 10 e. variety shows 0 1 2 3 4 5 6 7 8 9 10 f. cartoons 0 1 2 3 4 5 6 7 8 9 10 9. police and adventure shows (for example, "The A Team" and "Knight Rider") 0 1 2 3 4 5 6 7 8 9 10 h. comedy shows 0 1 2 3 4 5 6 7 8 9 10 1.50aPOperas 012345678910 ANSWER QUESTION 15 AS YOU THINK YOUR PARENTS WOULD.I 119 15. Here are some statements a person might say about newspapers. Based on the newspapers you read, check whether you strongly agree, agree, disagree or strongly disagree with each statement. Strongly Strongly Agree Agree Neutral Disagree Disagree (1) (2) (3) (4) (5) a. newspapers ignore controversial issues b. newspapers will not take a stand on some issues because of fear of offending people c. newspapers do not give all sides of a story d. I don't believe many stories in newspapers e. newspapers do not provide in-depth coverage on important issues f. newspapers are influenced by special interest groups‘ 9. newspapers are influenced by advertisers h. I am skeptical of newspapers that are monopolies i. newspapers give only the news they want to give HIVDWEK ll'll'. KLI'IHLIVINU UUtDIlUIVD HD 111111 VtKIHUV 1U IUU. 120 16. Different people have different reasons for watching TV. Here are some reasons other people gave us for watching TV. Please tell us how much each reason is like you. Check the best answer for you. a lot a little not much not at all I'watch TV... a. because it relaxes me b. so I won't be alone c. so I can learn about things happening in the world d. because it's a habit a. when I have nothing better to do f. because it helps me learn things about myself 9. because it's thrilling h. so I can forget about school and homework 1. because it's a pleasant rest j. when there's no one to talk to or play with k. so I can learn how to do things I haven't done before l. because I just like to watch m. because it passes the time away n. so I could learn about what could happen to me 0. because it excites me p. so I can get away from the rest of the family q. because it calms me down when I am angry r. because it makes me feel less lonely 18. 19. 20. 21. 16. (continued) l2l a lot a little not much not at all I'watch TV... 5. because it teaches me things I don't learn in school t. because it gives me something to do u. because it shows how other people deal with the same problems I have v. because it stirs me up w. so I can get away from ' what I'm doing . How many movies have you seen since school started? number of movies How many records or tapes have you bought since school started? number of records or tapes Does your family have a daily newspaper delivered to your home? Yes No Don't know Does your home have cable television? Yes No Don't know How many working television sets do you have in your home? number of working sets 122 22. In what grade are you? grade 23. What is your age? years ~ 24. What are your parents' jobs? father's job mother's job 25. What is your sex? Male Female Thank you for your help Instructions: QUESTION. 123 MEDIA USE QUESTIONNAIRE - BEST FRIEND You are being asked to complete this questionnaire to help us find out what media teen-agers use. This is not a test and you won't be graded on your replies. Answer each question as honestly as you can, but if you are not sure about a partic- ular answer check or write the answer that come closest to what you think you know. Take as much time as you need and BE SURE TO ANSWER EVERY Do ndfpay attention to the numbers you see in parentheses. They are only included to assist us in processing your answers. You may begin now. Don't st0p until you have answered all the questions. 124 ID No. With 0 as "never" and 10 as "always," indicate on a0-to-10 scale how much you use the following when you want information about movies (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4 5 6 7 8 9 10 c . radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 With 0 as "never" and 10 as "always," indicate on a 0-to-10 scale how much you use the following when you want information about music or musical performers (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4. 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d . magazines 0 1 2 3 4 5 6 7 8 9 10 e. another person or persons 0 l 2 3 4 5 6 7 8 9 10 125 With 0 as "never" andlxlas "always," indicate on a 0-to-10$cale how often you regd the following in the newspaper (circle the most appropriate answer . a. news about government and politics 0 1 2 3 4 5 6 7 8 9 10 b. news about the economy 0 1 2 3 4 5 6 7 8 9 10 c. sports news 0 1 2 3 4 5 6 7 8 9 10 d. advertisements 0 I 2 3 4 5 6 7 8 9 10 e. comics 0 1 2 3 4 5 6 7 8 9 10 f. entertainment news 0 1 2 3 4 5 6 7 8 9 10 9. crime news 0 1 2 3 4 5 6 , 7 8 9 10 With 0 as "never" and10 as "always," indicate on.0-to-105cale how much you watch the following on television (circle the most appropriate answer). a. national news 0 1 2 3 4 5 6 7 8 9 10 b. local news 0 I 2 3 5 6 7 8 9 10 c. 'sports events 0 1 2 3 4 5 6 7 8 9 10 d. movies 0 1 2 3 4 5 6 7 8 9 10 e. variety shows 0 1 2 3 4 5 6 7 8 9 10 f. cartoons 0 1 2 3 4 5 6 7 8 9 10 9. police and adventure shows (for example, "The A Team" and "Knight Rider") 0 1 2 3 4 5 6 7 8 9 10 h. comedy shows 0 1 2 3 4 5 6 7 8 9 10 i. soap operas 0 1 2 3 4 5 6 7 8 9 10 le Here are some statements a person might say about newspapers. Based on the newspapers you read, check whether you strongly agree, agree, disagree or strongly disagree with each statement. Strongly Agree (1) a. newspapers ignore controversial issues b. newspapers will not take a stand on some issues because of fear of offending people c. newspapers do not give all sides of a story d. I don't believe many stories in newspapers e. newspapers do not provide in-depth coverage on important issues f. newspapers are influenced by special interest groups 9. newspapers are influenced by advertisers h. I am skeptical of newspapers that are monopolies i. newspapers give only the news they want to give Agree (2) 11328191 (3) Strongly Disagree Disagree (4) (5) Instructions: 127 MEDIA USE QUESTIONNAIRE - PARENT You are being asked to complete this questionnaire as part of a study by MSU's School of Journalism of adolescent media use and how it relates to parental and peer media use. Your child has been asked to participate as part of his or her Explorer post activities. Answer each question as honestly as you can, but if you are not sure of a particular answer, check or write the answer that comes closest to what you think you know. Your responses will be held in strictest confidence. If you have any questions, feel free to contact Jim Bernstein at 485-7290. 128 ID No. With 0 as "never" and 10 as "always," indicate on a0-to-10 scale how much you use the following when you want information about movies (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 With 085 "never" and 10 as "always," indicate on a O-to-lO scale how much you use the following when you want information about music or musical performers (circle the most appropriate answer). a. newspapers 0 1 2 3 4 5 6 7 8 9 10 b. television 0 1 2 3 4. 5 6 7 8 9 10 c. radio 0 1 2 3 4 5 6 7 8 9 10 d. magazines 0 1 2 3 4 5 6 7 8 9 10 e. another person or persons 0 1 2 3 4 5 6 7 8 9 10 129 With 0 as "never" andlxlas "always," indicate on a.0-to-105cale how often you read the following in the newspaper (circle the most appropriate answer). news about government and politics news about the economy sports news advertisements comics entertainment news crime news OOOOOOO NNNNNN N wwwwwww With 0 as “never" and10 as "always," indicate you watch the following on television (circle a. b. national news local news sports events movies variety shows cartoons police and adventure shows (for example, “The A Team" and "Knight Rider") comedy shows soap operas 000000 1 1 1 NNNN 3 3 3 ####### on.0-to-105cale how much mmmmmmm \J\l\l oooooooooooooo {00040404040 10 10 10 IO 10 10 10 the most appropriate answer). 4 ##### 5 5 5 6 01010101 7 7 7 8 8 8 8 8 8 9 9 9 10 10 IO 10 10 IO 10 IO 10 130 Here are some statements a person might say about newspapers. Based on the newspapers you read, check whether you strongly agree, agree, disagree or strongly disagree with each statement. Strongly Strongly Agree Agree Neutral Disagree Disagree (1) (2) (3) (4) (5) a. newspapers ignore controversial issues b. newspapers will not take a stand on some issues because of fear of offending people c. newspapers do not give all sides of a story d. I don't believe many stories in newspapers e. newspapers do not provide in-depth coverage on important issues f. newspapers are influenced by special interest groups 9. newspapers are influenced by advertisers h. I am skeptical of newspapers that are monopolies i. newspapers give only the news they want to give 131 6. Do you subscribe to a daily newspaper? Yes No Don't know 7. Do you subscribe to cable television? Yes No Don't know 8. How many working television sets do you have in your household? number of working sets 9. How much education have you completed? through 8th grade college degree 9th through 12th graduate work high school degree graduate degree some college 10. What is your age? years 11. What is your total annual household income? dollars 12. What are the occupations of the main wage earners in your household? Person #1 Person #2 13. What is your sex? Male Female Thank you for your help. BIBLIOGRAPHY BIBLIOGRAPHY Abel, John D. "The Family and Child Television Viewing." Journal of Marriage and Family 38 (May 1976): 331-335. Adler, Keith, and Larry Haight. "Communication Satisfaction and General Satisfaction in the Doctor-Patient Relationship: A Conceptual Analysis and Exploratory Study." A paper presented to the International Communication Association, 1979. Adler, Richard P. (Principal Researcher). Research of the Effects of Television Advertisingon Children, Natioggl Science Foundation Report. Washington: Government Printing Office, 1978. Atkin, Charles K. "Children's Social Learning from Television Advertising: Research Evidence on Observational Modeling of Product Consumption." Advances in Consumer Research 3 (1976): 513-519. Atkin, Charles K., John Hocking and Martin Block. "Teenage Drinking: Does Advertising Make a Difference?" Journal of Communication 34 (Spring 1984): 157-167. Atwood, L. Erwin. "Perception of Television Programming Preferences Among Teenagers and Their Parents." Journal of Broadcastigg 12 (Fall 1968): 377-388. Avery, Robert. "Adolescents' Use of the Mass Media." American Behavioral Scientist 23 (September/October 1979): 53-70. Bachman, Jerald G. "Premature Affluence: Do High School Students Earn Too Much?" Economic Outlook USA (Summer 1983): 64-67. Bandura, Albert. Social Learning Theory. Morristown, NJ: General Learning Press, 1971. Bandura, Albert, Dorothea Ross and Sheila Ross. "Imitation of Film-Mediated Aggressive Models." Journal of Abnormal Psychology. 66 (January 1963): 3-11. Bandura, Albert, Dorothea Ross and Sheila Ross. "Transmission of Aggression Through Imitation and Aggressive Models." In Readings in Social Psychology, pp. 162-175. Edited by Jonathan L. Freeman, J. Merrill Carlsmith and David O. Sears. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1971. 132 133 Biddle, Bruce J., Barbara J. Bank and Marjorie M. Marlin. "Parental and Peer Influence on Adolescents." Social Forces 58 (June 1980): 1057-1079. Blake, Reed H., and Orval Nelson. "Mass Media Exposure as a Dimension of the Teenage Innovator." Paper presented to the Rocky Mountain Social Science Association, Laramie, WY, 1973. Blalock, Hubert M., Jr. Causal Inferences in Nonexperimental Research. New York: W.W. Norton 6: Company, Inc., 1964. Blumler, Jay. "The Role of Theory in Uses and Gratifications." Political Communication Review 3 (1978). Bogart, Leo. Press and Public. Hillsdale, NJ: Lawrence Erlbaum Associates, 1981. ' Brown, J.R. "Children's Use of Television." In Children and Television, pp. 116- _— 136. Edited by Ray Brown. Beverly Hills: SagefiPublications, 1 76. Burgoon, Judee K., Michael Burgoon and Charles K. Atkin. The World of the Working Journalist. New York: Newspaper Advertising Bureau, 1982. Burgoon, Michael, Charles K. Atkin, Judee K. Burgoon and James Gaudino. "The Pennywhistle Press: Forming the Reading Habit by Tailoring Newspaper Content to Children." Unpublished paper, Michigan State University, 1982. Campbell, Ernest Q. "Adolescent Socialization." In Handbook of Socialization Theory and Research, pp. 821-859. Edited by David A. Goslin. Chicago: Rand McNally and Company, 1969. Campbell, Laurence R. "What Teenagers Read in the Hometown Daily Newspaper." Paper prepared for the Quill and Scroll Foundation, Iowa City, IA, March 1970. Chaffee, Steven.H., and Jack M. McLeod. "Adolescent Television Use in the Family Context." In Television and Social Behavior, Vol. 3, pp. 149-172. Edited by‘ George A. Comstock and Eli A. Rubinstein. Washington: Government Printing Office, 1971. Chaffee, Steven H., and Jack M. McLeod. "Consumer Decisions and Information Use." In Consumgr Behavior: Theoretical Sources, pp. 385-415. Edited by Scott Ward and Thomas S. Robertson. Englewood Cliffs, NJ: Prentice- Hall, Inc., 1973. Chaffee, Steven H., Jack M. McLeod and Charles K. Atkin. "Parent-Adolescent Similarities in Television Use." Paper presented to the Association for Education in Journalism, Washington, 1970. Chaffee, Steven H., Jack M. McLeod and Charles K. Atkin. "Parental Influences on Adolescent Media Use." American Behavioral Scientist 14 (January/February 1971): 323-341. 134 Chaffee, Steven H., Jack M. McLeod and Jose L. Guerrero. "Origins and Implications of the Co-orientation Approach in Communication Research." Paper presented to the Association for Education in Journalism, Berkeley, CA, 1969. ' Chaffee, Steven H., and Albert R. Tims. "The Development of News Media Use in Adolescence: Implications for Political Cognitions." Paper presented to the International Communication Association, Boston, 1982. Chaffee, Steven H., and Albert R. Tims. "Interpersonal Factors in Adolescent Television Use." Journal of Social Issues 32 (Winter 1976): 98-115. Chaffee, Steven H., L. Scott Ward and Leonard P. Tipton. "Mass Communication and Political Socialization." Journalism Quarterly 47 (Winter 1970): 647- 659. Children and Newspapers. New York: Newspaper Advertising Bureau, 1980. Churchill, Gilbert, Jr., and George P. Moschis. "Television and Interpersonal Influences on Adolescent Consumer Learning." Journal of Consumer Research 6 (June 1979): 23-35. Clarke, Peter. "Does Teen News Attract Boys to Newspapers?" Journalism Quarterly 45 (Spring 1968): 7-13. Clarke, Peter. "Parental Socialization Values and Children's Newspaper Reading." Journalism Quartem 42(Autumn 1965): 539-546. Cohen, Jacob, and Patricia Cohen. Applied Multiple Regression/ch'firElation Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates, 1975. Coleman, James L. The Adolescent Socig. New York: Free Press, 1961. Comstock, George, Steven Chaffee, Natan Katzman, Maxwell McCombs and Donald Roberts. Television and Human Behavior. New York: Columbia University Press, 1978. Conway, M. Margaret, A. Jay Stevens and Robert G. Smith. "The Relation Between Media Use and Children's Civic Awareness." Journalism Quarterlj 52( Autumn 1975): 531-538. Dembo, Richard, and Robin McCron. "Social Factors in Media Use." In Children gnd Television. pp. 137-166. Edited by Ray Brown. Beverly Hills: Sage Publications, 1976. Editor 6: Publisher International Yearbook. New York: Editor 6: Publisher Co., 1983. Elam, Stanley M. "Anti-Democratic Attitudes of High School Seniors in the Orwell Year." Phi Delta Kappan, January 1984, pp. 327-332. ———————i 135 Engel, James F., Roger D. Blackwell and David T. Kollat. Consumer Behavior, 3rd Ed. Hinsdale, IL: The Dryden Press, 1978. Forslund, Morris A., and Thomas J. Gustafson. "Influence of Peers and Parents and Sex Differences in Drinking by High School Students." Quarterly Journal of Studies on Alcohol 31 (December 1970): 868-875. Greenberg, Bradley S. "Gratifications of Television Viewing and Their Correlates for British Children." In The Uses of Mass Communication, pp. 71-92. Edited by Jay Blumler and Elihu Katz. Beverly Hills: Sage Publications, 1974. Haller, A.O., Joseph Woelfel and Edward L. Fisk. "The Wisconsin Significant Other Battery." A report from the University of Wisconsin, Madison, July 1968. Herriott, Robert E. "Some Social Determinants of Educational Aspiration." Harvard Education Review 33 (Spring 1963): 157-163. Hill, Reuben, J. Mayone Stycos and Kurt W. Back. The Family and Pomlation Control: A Pugto Rican Experiment in Social Change. Chapel Hill: University of North Carolina Press, 1959. Himmelweit, Hilde, A.N. Oppenheim and Pamela Vince. Television and the Child. London: Oxford University Press, 1958. Johnstone, John W.C. "Social Integration and Mass Media Use Among Adolescents: A Case Study." In Uses of Mass Communication, pp. 35-47. Edited by Jay Blumler and Elihu Katz. Beverly Hills: Sage Publications, 1974. Jones, Landon. Great Expectations: America and_the Baby Boom Generation. New York: Coward, McCann and Geoghegan, 1980. Kandel, Denise. "Inter- and Intragenerational Influences on Adolescent Vlarijuana Use." Journal of Social Issues 30 (Spring 1974): 107-135. Kandel, Denise, and Gerald S. Lesser. "Parental and Peer Influences on Educational Plans of Adolescents." American Sociological Revyei 34 (April 1969): 213-223. Kaplan, Abraham. The Conduct of Inquiry. New York: Chandler Publishing Company, 1964. Katz, Elihu. "Mass Communication Research and the Study of Popular Culture." Studies in Public Communication 2 (Summer 1959): 1-6. Kerlinger, Fred N. Foundations of Behavioral Research, 2nd ed. New York: Holt, Rinehart and Winston, Inc., 1973. Kohlberg, Lawrence. "The Cognitive Development Approach to Socialization." In Handbook of Socializatign Theory and Research, pp. 347-480. Edited by David A. Goslin. Chicago: Rand McNally and Company, 1969. 136 Larkin, Ernest F., and Gerald L. Grotta. "The Newspaper as a Source of Consumer Information for Young Adults." Journal of Advertising 6 (Fall 1977): 5-10. Lehnert, Eileen. "A Market Segmentation Study of the Youth Market's Perception of the Ideal Newspaper." Paper presented to the Association for Education in Journalism, Boston, 1980. ‘ Liska, Allen. The Consistency Controversy. New York: John Wiley and Sons, 1975. Lyle, Jack, and Heidi R. Hoffman. "Children's Use of Television and Other Media." In Television and Social Behavior, Television in Day-to-Day Life: Patterns of Use, pp. 129-256. Edited by Eli Rubinstein, George Comstock and John Murray. Washington: Government Printing Office, 1972. Maccoby, Eleanor. "Why Do Children Watch Television?" Public Opinion Quarterly 18 (Fall 1954): 239-244. Mass Media in the Family Setting. New York: Newspaper Advertising Bureau, 1980. McGuire, William J. "Psychological Motives and Communication Gratifications." In The Uses of Mass Communication, pp. 167-196. Edited by Jay Blumler and Elihu Katz. Beverly Hills: Sage Publications, 1974. McLeod, Jack M., Charles K. Atkin and Steven H. Chaffee. "Adolescents, Parents and Television Use: Adolescent Self-Report Measures from Maryland and Wisconsin Samples." In Television and Socii Behavior, Vol. 3, pp. 173-238. Edited by George A. Comstock and Eli Rubinstein. Washington: Government Printing Office, 1971. McLeod, Jack M., and Steven Chaffee. "The Construction of Social Reality. In The Social Influence Processes, pp. 50-99. Edited by James T. Tedeschi. Chicago: Aldine-Atherton, Inc., 1972. McLeod, Jack M., Steven H. Chaffee and Daniel B. Wackman. "Family Communication: An Updated Report." Paper presented to the Association for Education in Journalism, Boulder, CO, 1967. McLeod, Jack M., and Garrett J. O'Keefe Jr. "The Socialization Perspective and Communication Behavior." In _Cgrrent Perspectives in lass Communication Research, pp. 121-168. Edited by Gerald Kline and Phillip Tichenor. Beverly Hills: Sage Publications, 1972. Moore, Roy L., and George P. Moschis. "The Impact of Newspaper Reading on Adolescent Consumers." Newspaper Research Journal 2 (July 1981): 3-8. Moore, Roy 1..., and George P. Moschis. "A Longitudinal Analysis of Television Advertising Effects on Adolescents." Paper presented to the Association for Education in Journalism, Athens, OH, 1982. 137 Moore, Roy L., and George P. Moschis. "The Role of Family Communication in Consumer Learning." Journal of Communication 31 (Autumn 1981): 42-51. Moore, Roy L., Lowndes F. Stephens and George P. Moschis. "Mass Media and Interpersonal Influence in Adolescent Consumer Socialization." Paper presented to the International Communication Association, Portland, Oregon, 1976. Moschis, George P., and Gilbert A. Churchill Jr. "An Analysis of the Adolescent Consumer." Journal of Marketing 43 (Summer 1979): 40-48. Moschis, George P., and Gilbert A. Churchill Jr. "Consumer Socialization: A Theoretical and Empirical Analysis." Journal of Marketim Research 15 (November 1978): 277-291. Moschis, George P., and Roy L. Moore. "Decision Making Among the Young: A Socialization Perspective." Journal of Consumer Research 6 (September 1979): 101-112. Moschis, George P., and Roy L. Moore. "Family Communication and Consumer Socialization." In Advagces in Consumer Research, Vol. 6, pp. 359-363. Edited by William L. Wilkie. Ann Arbor: Association for Consumer Research, 1979. Murray, John P. Television and Youth: 25 Years of Research and Controversy. Boys Town, NE: The Boys Town Center for the Study of Youth Development, 1980. Newcomb, Theodore. "An Approach to the Study of Communication Acts." Psychological Review 60(January 1953): 393-404. Nunnally, Jum C. Psychometric Theory, 2nd Ed. New York: McGraw-Hill Book Company, 1978. Ray, Michael. "Psychological Theories and Interpretations in Learning." In Consumer_Byehavior:fiTheoretical Sources, pp. 45-117. Edited by Scott Ward and Thomas S. Robertson. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1973. Roberts, Donald F., and Wilbur Schramm. "Children's Learning from the Mass Media." In The Process and Effects of Mass Communication, pp. 595-611. Edited by Wilbur Schramm and Donald F. Roberts. Urbana, IL: University of Illinois Press, 1977. Rubin, Alan M. "Television Use by Children and Adolescents." Human Communication Research 5 (Winter 1979): 109-120. Scheff, Thomas J. "Toward a Sociological Model of Consensus." American Sociological Review 37 (February 1967): 32-45. Schramm, Wilbur, Jack Lyle and E.B. Parker. Television in the Lives of Our Children. Stanford, CA: Stanford University Press, 1961. 138 Schweitzer, John C., David Weaver and Gerald C. Stone. "Morning-Evening Newspaper Circulation: What Effect Do Appearance and Content Have?" Journalism Quarterly 54 (Winter 1977): 515-522. Stone, Gerald C., and Roger Wetherington Jr. "Confirming the Newspaper Reading Habit." Journalism Quarterly 56 (Winter 1979): 554-561. Stone, Vernon A., and Steven H. Chaffee. "Family Communication Patterns and Source-Message Orientation." Journalism Quarterly 47 (Summer 1970): 239-246. Surlin, Stuart H., Alan Wurtzel and Linda Whitener. "Parental Control of Children's Television Viewing Behavior: Support for the Reverse Modeling Principle." Paper presented to the International Communication Association, Chicago, 1978. Swanson, David L. "The Uses and Misuses of Uses and Gratifications." Human Communication Research 3 (Spring 1977): 214-221. Tagiuri, Renato, Jerome S. Bruner and R.R. Blake. "On the Relation Between Feelings and Perceptions of Feelings Among Members of Small Groups." In Readings in Social Psychology, 3rd Ed., pp. 110-116. Edited by Eleanor E. Maccoby, Theodore M. Newcomb and E.L. Hartley. New York: Holt, Rinehart and Winston, 1958. Tims, Albert R., and Steven H. Chaffee. "A Test of the Cumulative Model of Adolescent News Media Use." Paper presented to the Association for Education in Journalism, Corbalis, OR, 1983. Ward, Scott. "Consumer Socialization." Journal of Consumer Research 1 (September 1974): 1-16. Ward, Scott. "Effects of Television Advertising on Children and Adolescents." In Television and Social Behavior, Vol. 4, pp. 432-451. Edited by Eli A. Rubinstein, George A. Comstock and John P. Murray. Washington: Government Printing Office, 1971. Ward, Scott, and Thomas S. Robertson. "Adolescent Attitudes Toward TV Advertising: Preliminary Findings." In Television and Social Behavior, Vol. 4, pp. 526-542. Edited Eli A. Rubinstein, George A. Comstock and John P. Murray. Washington: Government Printing Office, 1971. Ward, Scott, and Daniel Wackman. "Family and Media Influences On Adolescent Consumer Learning." American Behavioral Scientist 14 (January/February 1971): 415-427. Ward, Scott, Daniel Wackman and Ellen Wartella. Children Learning to Buy: The Development of Consumer Informtion Processing Skills. Cambridge, MA: Marketing Science Institute, 1975. Women and Newspagaers in the 19705. New York: Newspaper Advertising Bureau, 1979. 139 Zigler, Edward, and Irvin L. Child. "Socialization." In The Handbook of Social Ps cholo , Vol. 3, pp. 450-589. Edited by Gardner Lindzey. Reading, MA: Addison-Wesley Publishing Co., 1969. nICHIan STATE UNIV. LIBRARIES 11111111111111111111111111111111111111111 31293107543534