I El. E1 EEEVIV'E . Vm ‘III‘VV‘I' ‘ EVEEEEE EEVEV‘“"’EE"“E‘ 'EEEEEE V‘ ‘EEEEVE'E IE "IIIIII ‘V" M5 'VVVV'VVV: I'MLE. "'V ‘VVVI‘VVVVV VVV VVVV.VVV.VVVVVV 'VV. IEVVVEIV'V ..,I .V. III! . E E E‘E'E EEE'E‘E'EEVE EEVIEEV. EVV EVE... E“ .IIVE’ EEEE‘E EEEEEEIV V ‘EHUEEEEE EEV. WE MW . , 5.- _..~ .o—~ —, 2:.- I '- r - "a?“ "Va?“ . 3,... y - r V' V. .. I?H%"WMWEEEWVEEE I. V VV V... .V VVVVV - E ' E” III! '2' ..'V . IEVVE'VEEE"'E‘E§IE ‘I IVV‘EIIIVVVIVV . E". .5EVEEEEEE, ‘EEE’EE IIIhm EE'ELE . .‘VIV‘ _’. ,VE'EV EVVVVEVVV Vs. E EV V, V )4; 'EE EEEEEEVE {III NEEIEV 'EEIE‘E EEEEEEEIE EIEE EEE'EEEEEEEE EEIEEIE IIt. .IEEEI EV’LVILVE'VEEEEV' "EH1 E.. EEEEE' VIIV' . . .EEEEVEEE'IIE EEEEVEE‘EEEEEEI‘ V .V ’ E V; ”E“ ~ I1. . E VE,EflV“,“ .IVE EVE? " I V. “ EIME III EEI'EE I.HII iEE .fi. 5 Efiséggfifigfié7. V .IVV V . VVIII’ EEEE'E'. - ... ..E EE,E'E EE'tg‘ELEIVE‘E I EIEVI'E'EE EE‘IJ I‘AL. 5': .FEEE'EEE ‘4’ :EfEJu' tr ‘ :‘ 1:23;? ‘: E I éiEEEEEEEiEEE EIE‘I;E'I‘{E(1EEEEEJiiEEEHELEE‘tEEZE”:SVEEEtE”EEH I %g‘ its? - . I... ‘ .‘ j . EVEEEEEIEE‘EE'ZEEEEEE‘1EFEE‘VEIIJE‘EE‘QI“‘EVVEIEgiEEE EIIEEV . 44.335.31.15, : 7 I:‘ V EEVEV’V‘EEV‘VEEE’VVEIV‘ VEEiV E‘EEE‘E V EEVVV IV; I. .EEVV .. 'V's' I I s :r!‘:0":-E.IE 0".‘E ’IV’EVV'E'IEVi-V'VVIVVEVE ""51 V51; ' ’V‘. . .VEVVVV Vi ;EEEE EEJ E \V I E E EEO? . '9‘ E. V 5 3“ E5: '. “Huh, fIVEEE‘VE'ZV‘EVEE‘VEEE‘EfiE'VE EEEEEE...EVV.VV,E;EEVEEEVVEI';EEEVVIEEVI€'V c 73:. 93.3. 1 3:3, ZIEEEXVEVEVVIVEEVEFTEE}. ‘VVIVIIVEIEVEVIE'IVEVV I EE'EEVVE‘VVVE'VVVIVE'I.EIEVIV'VE‘I. EL. . EEEEEEEEE EEEEE I E _ z. EEEEIE.IIE§§I :: . |.;:I I ..- pI;;;}V_.‘ El : I”; : :‘V‘,' .E (.'EEQ .;;i '3 LEE Esz IE“ ”:I giffi“ :1 3&411“? NEEE; gEEq - .EE' EEE E‘Eu .l .' ,V V FEE“? 5219?} ‘7"- “I .E-I EEE'EEE EE‘ \EEEEV EEE E‘ E" ! WEEE I V I E E“ . 39M : E“- L: . . VEEE‘EKEEEE ‘1‘ EEE E" E” ;:E EEEE EEE‘ 7'1 h9EEE‘EE'EEf" EE’EEEE EE‘EEEEEE EEEVEEE‘EEEEE E; “’3 g, . E I EEEEEEE HEEE'EEE‘ I; EE ‘EEEEEEEE EIEEE'E. EEEEE‘E E E 5.514. . E'. : EEE EEEEEIEEE 011‘! ”I.” EV'VEEEEJ' EEIEEE'IE.L“ 55"? z . ”(P V I EVEVEVEEEEWE ' 'A‘ . . .5. 9. ._ . - ' u . I EVVVV I.V.'V‘IV'€.I?I.V;5VEVV "I! VEEh I ‘ . EEEEE"EEEEE:ED:-'1I VEIIt-ZI EEEEE'EEETE :EEEIzEEE,E '3 VI 1’?” EE’VEEEE ”E “W , EEEEE “1' EEEEEE EEI'V'. 'V5 EEEE HIEEEIEEI EEEE IE“; . i . .- "EEE EIV‘EV TE‘.E.E- IEEEH ‘E'VEE'EEE EEEE IEE'“ E:'?‘; Eif‘ =.V 635‘. EVEVEEVI -'-VV IEIEV 5M MI} .3". {VVVVVVVIVVVV EE , .. . . I -. V 0“ . “'3. g . 1941 v - "- L-zr ' u .. V4 O‘: E . E ‘I'E‘V. ‘ ' 5E VE'EVVVSEVE EEEEEEE 4:13;; ..- VVV‘ ‘ . . EQEEiLIE‘gE‘I ' . - Va ..- 1. _-I I" . ‘IIENI E‘EIENE'E 'EEE. EL} ' ‘2E.EV':1.:V'V':E'E'EEVE'E‘VVV'EJIE'VIIVI ' VV E-EEI VE EV‘EEE' VV ‘3 v .u-‘ EVVI I. EEEEEEIEEE (E. “E ”L.“ E EE “EE . E:EE'EE II‘ E. .IVI.I E I ' ‘1,” ’EEES‘E‘ VEI .EEE' ‘ E IIII VJEE'E‘ LE E-hE ‘EVI‘ I‘IIrV .:.2 EEE IVII’VVIIVEEIII .;:II-IV'.E -‘I"IEA':IE1V.."EV" ”IEEIEVE Elt EV. gig“ -EV.I,V ;VI:L,‘§VEV:".uEa‘xVA ,., IEEEE: EE‘» VEV'E'V "EEVV‘1EEEVE' . EEEE‘E' EE E EEEEE ‘ EEEJ “‘EE ”ka EEVEE‘EEEEVEEEEEE‘EEEME'EIE‘VEE E ’ ".‘V.'V ‘ I "V‘. :.‘-.V'1T‘V :V'Z V E' ‘V E}: V‘;;E EEEEVE V .‘EEVVV E. IEE3 '.EE :‘EE . VII;I.VI_V.IVI.;;VI'.:VFI : .: 'E'V ‘EIEIEEEEEE' V VEVVIEV V .EE, 5E EgEiEI EEIEEEE EEE 1‘“. a“, EVEEEEVEEEfEErEEEEEEEVEQI'EE :I. .V.Vl‘.‘, V II \ E~IV ,5'.".‘.VV El‘l ”E, .1 ' I ' ‘.V,‘ '5 .' N: I-gIII;-:VE;.‘V:~I.1'I'IVIL'I‘V ’EEEE EI'EEVEV V ' EE ,,..g,.EgII'VEEI.1EI..V'VE‘:I‘E“ I VVVEVE EEEE V‘E ,JEVVEI; EEEEEE IV‘E'I I EE EE Vnwh'“ufiE M.”. .VV I EHM”EW.EfipJEWME ELPEJWLkbwfi '3 VII I VIVV"i.‘:;.I'V w; ‘VVVVIVVVI‘ .I I IIIVVuI EEEI ’EE‘ ”EVE’VV IE IEI'I'V'V'VI EfipsiEEE " V ' IVVVVEE‘EE :ziVEEI‘E EEEEE EEE'EE ' EfEE EEEEEEE'EEEEEEEEEEE I E :Vétt‘VlE‘EEVEIVV.._ EV ‘. E.IVE’EIE‘E,IV‘_.E"E ‘ EEEEE EE .V; EE.5; E‘EEVI [EE VE‘EEEEE EVVI EEE EEVE E 'EEEEE EE? EVVVVIEEV, EE'VI‘EVEE ‘EE‘ VE'EEII le. IVI EEEVVVVEIEEEEE Er EEEJE E.VEEE~EEEEEEEEEEEEEE EEEEEEE. EEEE '.. 'lE‘EE'E-IEI'YEII - I ,I’ V ' .." V‘ EVEEIE .V"“'E"E «"E EEE'EE’ EEES EE . .éEVEEIEEEVEEE.‘,.~: IIIEIEEE ‘EEE EE E E'EEIEVEE EIEI VEE EEE EEE'EEE EEIEEEEVEE-EEIVJV EEEEE Lid ‘3 in .V... .,.V.IV,..V» I’" .. a: I "fi’V/f',‘ VVV'VV'V EE'EEE IIEEIE IWV. I I E‘EVI' EEEIEE .V E EEVVE .1 O llllllllllllllllllllllllllllllll!!!ll||lll||l|ll|llllllll|l|H|l L 3 1293 1 115 4858 L I 4 gAII. Y Michigan State University . This is to certify that the thesis entitled PARENT AND ADULT MEDIATION OF TELEVISION PROGRAMMING AND PEER PRESENCE: THEIR EFFECTS ON CHILDREN presented by Brian Frederick Fontes has been accepted towards fulfillment of the requirements for Ph.D. degree in _C_thnnniear.ion Arts and Sciences--Mass Media L4; ,9. (M! Major professor Date [/I// 2}/ 7 7 0-7 639 (:> Copyright by BRIAN FREDERICK FONTES 1977 PARENT AND ADULT MEDIATION OF TELEVISION PROGRAMMING AND PEER PRESENCE: THEIR EFFECTS ON CHILDREN By Brian Frederick Fontes A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Communication Arts and Sciences--Mass Media 1977 x’t ABSTRACT PARENT AND ADULT MEDIATION OF TELEVISION PROGRAMMING AND PEER PRESENCE: THEIR EFFECTS ON CHILDREN By Brian F. Fontes BACKGROUND Research has shown that comments made by experimenters before and during the viewing of anti-social television programming can reduce post- viewing aggression and can also increase the amount of information gained from educational programming. Media campaigns emphasize that parents can effectively mediate television programming by talking with their children about what is occurring in TV programming. This study examines the effect of mother and adult comments and peer presence on children's ability to (a) recall pictorial content, (b) recall consequences of violent behavior, (c) recall verbal content, (d) perceive violence as inappropriate, and (e) demonstrate post-viewing cooperation. METHOD An experimental post-test only control group design was employed. Subjects were randomly assigned to one of the following conditions: (1) mother providing comments about TV programming while viewing with her child, (2) mother viewing with her child but providing no comments, (3) a female adult providing comments about TV programming while viewing with a child, (4) a female adult viewing with a child but providing no comments, (5) a child viewing with a peer, with no comments provided, and (6) a child viewing alone. Brian F. Fontes Comments were made before, during, and after the viewing of the stimulus. Comments described what occurred in the stimulus and condemned violence. The stimulus depicted shooting scenes taken from two movies shown on network television during 1976-77. Children (N=l32) 6-8 years of age participated in the experiment. Each treatment condition contained 11 males and 11 females. After viewing the stimulus each subject played a game in which he/ she could either help or not help someone. A post-viewing cooperation measure of duration and frequency for helping and non-helping behavior was obtained. Subjects were then interviewed to determine how much of the stimulus could be recalled and if they perceived violence occurring in the stimulus as inappropriate. One:say ANOVA.was used to test each dependent measure. Where F ratios were significant, post hoc comparisons determined which condi- tion(s) contributed to their significance. RESULTS Subjects participating in the parent, adult or combined parent/ adult "comment" conditions are able to recall more: (1) pictorial con- tent, (2) consequences of violent behavior, and (3) verbal content than ' and "alone" those subjects participating in the "no comment," "peer,' conditions. Those subjects participating in the combined parent/adult "comment" conditions perceive more violent behavior as inappropriate than those in the "no comment" conditions. There are no differences between parent and adult "comment" conditions and parent and adult "no comment," Brian F. Fontes "peer," and "alone" conditions for the dependent measure of perceived inappropriateness of violent behavior. There are no differences be- tween treatment conditions or combinations of treatment conditions for the dependent measure of post-viewing cooperation. Subjects participating in the parent and combined parent/adult "come ment" conditions are able to recall more: (1) pictorial content, (2) consequences of violent behavior, and (3) verbal content than those sub- jects participating in the parent and combined parent/adult "no comment" conditions. There are no differences between treatment conditions or combinations of treatment conditions for the dependent measure of post-viewing coopera- tion when controlling for sex. CONCLUSIONS Findings of this study support previous research indicating that comments can increase the amount of information children can recall from television programming. They do not, however, support research demon- strating that comments are effective in medicating post-viewing behavior. Comments did not increase the amount of children's post-viewing coopera- tion. Because the measure of post-viewing cooperation differs from measures used in previous research, the extent to which this research can support previous research on the effects of comments on post-viewing behavior is limited. DEDICATED TO Jennifer, Michael, Carolann Jason, Kristin ii ACKNOWLEDGEMENTS I want to acknowledge and thank the members of my committee for the time, assistance and support they have given me throughout my doctoral program at Michigan State University. Professor John D. Abel, guidance and dissertation committee chairman, deserves special recognition for his encouragement of verbal mediation of television content as a dissertation topic. I am especially grateful for his academic guidance, example of professionalism, support and personal friendship. Appreciation is also extended to Professors Thomas Baldwin, Martin Block and Margaret Buboltz for their help and assistance while serving on my committee. On a personal level I thank Thomas Baldwin, chairman of the Ph.D. Program in the Mass Media, and Robert Schlater, chairman of the Depart- ment of Telecommunication at MSU, for their support of my academic interests. Appreciation is also extended to my personal friends and fellow graduate students for their encouragement. I am particularly grateful to my parents, Mr. and Mrs. Frank R. Fontes, and my grandparents, Mr. and Mrs. Frederick A. Cedros, for their love and encouragement of my academic interests. Finally, I thank those who helped in the technical aspects of this study, and especially Jaynewzenaty for her time and concern. I am very thankful to Lois Pierson, Ann Spalding and Shellie Abel for their typing and editing abilities in preparing this dissertation. iii TABLE OF CONTENTS CHAPTER I. STATEMENT OF THE PROBLEM Background of Study Purpose of Study The Problem Theoretical Framework and Hypotheses Limitations of the Scope of the Problem CHAPTER II. LITERATURE REVIEW Children and Television Viewing TV Violence Models of the Effects of TV on Children The Catharsis Model The Arousal Model Comments and Post-Viewing Behavior The Facilitation Model Incidental Learning Social Learning Verbal Comments and Learning from TV Theories of Child Development Developmental and Learning Theories -- Moral Development Child Development and Verbal Communication Need for Research on Verbal Mediation of TV Programming The Social Environment iv 16 20 22 23 28 (“30 33 36 39 40 Page Family Effects 41 Peer Effects 47 CHAPTER III. METHODOLOGY Anti-Social Stimulus 51 The Design 52 Independent Variable 52 Operationalization of Dependent Variables 54 Recall of Content 54 Recall of Consequences of Anti-Social Behavior 54 Inappropriateness of Anti-Social Behavior 55 Post-Viewing Cooperation 55 Control Variables 56 The Viewing Situation and Order of Events 57 Order of Events 58 Sampling 60 Pretest 62 Statistical Analyses and Hypothesis Formation 63 Data Analysis 78 CHAPTER IV. FINDINGS Description of Sample 81 Description of Measures 82 ANOVA Results 86 Test of Hypothesis 1 104 Hypothesis 1A 105 Hypothesis 13 105 vi Hypothesis 1C Hypothesis 1D Hypothesis 2 Hypothesis 2A Hypothesis ZB Hypothesis 2C Hypothesis 2D Hypothesis 3 Hypothesis 3A Hypothesis BB Hypothesis 3C Hypothesis 3D Hypothesis 4 Hypothesis 4A Hypothesis 4B Hypothesis 4C Hypothesis 4D Hypothesis 5 Hypothesis 5A Hypothesis SB Hypothesis 5C Hypothesis 5D Tests of Hypotheses, Controlling for Sex Hypothesis 1, 1A, 18, 1C, 1D Recall of Pictorial Content (Hypothesis 1) Recall of Consequences of Anti-Social Behavior (Hypothesis 1A) 108 109 110 110 111 112 113 114 114 115 116 117 118 118 119 120 121 122 122 123 124 124 125 125 vii Recall of Verbal Content (Hypothesis lB) Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 1C) Post-Viewing Cooperation (Hypothesis 1D) Hypotheses 2, 2A, 23, 2C, 2D Recall of Pictorial Content (Hypothesis 2) Recall of Consequences of Anti-Social Behavior (Hypothesis 2A) Recall of Verbal Content (Hypothesis 2B) Perceived Inappropriateness of Anti—Social Behavior (Hypothesis 2C) Post-Viewing Cooperation (Hypothesis 2D) Hypotheses 3, 3A, 3B, 3C, 3D Recall of Pictorial Content (Hypothesis 3) Recall of Consequences of Anti-Social Behavior (Hypothesis 3A) Recall of Verbal Content (Hypothesis 3B) Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 3C) Post-Viewing Cooperation (Hypothesis 3D) Hypotheses 4, 4A, 4B, 4C, 4D Recall of Pictorial Content (Hypothesis 4) Recall of Consequences of Anti-Social Behavior (Hypothesis 4A) Recall of Verbal Content (Hypothesis 4B) Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 4C) Post-Viewing Cooperation (Hypothesis 4D) Hypotheses 5, 5A, SB, 5C, 5D Recall of Pictorial Content (Hypothesis 5) 126 126 128 128 128 129 129 129 131 132 132 132 132 133 135 135 135 136 136 136 138 138 viii Page Recall of Consequences of Anti-Social Behavior (Hypothesis 5A) 138 Recall of Verbal Content (Hypothesis 5B) 139 Perceived Inappropriateness of Anti-Social Behavior (Hypothesis SC) 139 Post-Viewing Cooperation (Hypothesis SD) 139 Summary 141 CHAPTER V. SUMMARY CONCLUSIONS AND DISCUSSION Statement of the Problem and Study Design 146 Findings: General Observations 148 Findings: Dependent Measures of Recall 149 Conclusion 151 Discussion 151 Findings: Dependent Measures of Perceived Inappropriateness of Anti-Social Behavior and Post-Viewing Cooperation 158 Conclusions 159 Discussion: Perceived Inappropriateness of Anti-Social Behavior 160 Discussion: Post-Viewing Cooperation 164 Findings: Hypothesis Tests 166 Limitations of Research 173 Conclusions and Implications for Future Research 174 Policy Implications 175 Implications for Future Research 178 APPENDICES Appendix A -- Script of Segments 182 Appendix B -— Comments Made by Parent/Adult 190 ix Appendix C -- Instrument Appendix D -- Letters to Parents and Consent Forms LIST OF REFERENCES Page 192 206 209 Table II III IV VI VII VIII IX XI LIST OF TABLES Correlation Matrix of Dependent Measures. Analysis of Variance Tests on Dependent Measures/ Total Sample. Analysis of Variance Tests on Dependent Measures/ Males. Analysis of Variance Tests on Dependent Measures/ Females. Summary Table: Overall F Tests for Dependent Measures. Hypotheses 1, 1A, 1B, 1C, 1D: Tukey Contrasts Between "Mother Comment" and "Mother No Comment" Conditions and Levels of Significance. Hypotheses 2, 2A, 2B, 2C, 2D: Tukey Contrasts Between "Female Adult Comment" and "Female Adult No Comment" Conditions and Levels of Significance Hypotheses 3, 3A, 3B, 3C, 3D: Scheffé Contrasts Between "Comment" and "No Comment" Conditions and Levels of Significance. Hypotheses 4, 4A, 4B, 4C, 4D: Scheffé Contrasts Between "Comment" Conditions and "Peer Viewing" Conditions and Levels of Significance. Hypotheses 5, SA, SB, SC, SD: Scheffé Contrasts Between "Comment" Conditions and "Viewing Alone" Conditions and Levels of Significance. Hypotheses 1, 1A, 1B, 1C, 1D: Controlling for Sex: Tukey Contrasts Between "Mother Comment" and "Mother No Comment Conditions and Levels of Significance. Page 85 87-91 92-96 97—101 102-103 108 112 116 120 124 126-127 Table XII XIII XIV XV xi Hypotheses 2, 2A, 2B, 2C, 2D: Controlling for Sex: Tukey Contrasts Between "Female Adult Comment" and "Female Adult No Comment" Conditions and Levels of Significance. Hypotheses 3, 3A, 3B, 3C, 3D: Controlling for Sex: Scheffé Contrasts Between "Comment" and "No Comment" Conditions and Levels of Significance. Hypotheses 4, 4A, 4B, 4C, 4D: Controlling for Sex: Scheffé Contrasts Between "Comment" and "Peer Viewing" Conditions and Levels of Significance. Hypotheses 5, SA, SB, 5C, 5D: Controlling for Sex: Scheffé Contrasts Between "Comment" and "Viewing Alone" Conditions and Levels of Significance. 130-131 133-134 136-137 140-141 CHAPTER I STATEMENT OF THE PROBLEM Background of Study The issue of anti-social television programming and its effects on children is the concern of the broadcast industry, consumer groups, Congress, and, more recently, the American Medical Association. For the past twenty-five years, researchers have been examining the effects of television on the family and especially on children. In the late forties and early fifties, purchasing a television set was thought of as a means to bring the family together in the home set- ting (Riley, Cantwell & Rutteger, 1949; Hamilton & Lawless, 1956). It became apparent, however, that viewing television was largely a non- interactive activity (Maccoby, 1951). Television was shown to contribute to "parallel" rather than "interactive" social activity within the family (Hamilton & Lawless, 1956). These early studies focused on the impact of television on the family as a whole. Comparisons were made of family activities before and after the purchase of a television set. By the mid-1950's, however, the research emphasis was changing from TV's impact on the family structure to television programming content's impact on children's behavior. As early as 1954, Senator Estes Kefauver, then Chairman of the Senate Subcommittee on Juvenile Delinquency, questioned the display of violent behavior in television entertainment programming. The National Association of Broadcasters promised that it would develop research to measure the effects of television programming on children (Senate Subcommittee to Investigate Juvenile Delinquency, 1954). In 1961 the Senate Subcommittee on Juvenile Delinquency, then chaired by Senator Thomas Dodd, again questioned the use of violence in children's television. The dodd hearings revealed that the earlier promise by the NAB to conduct research into the effects of television programming on children had not been fulfilled. Testimony provided during the hearings indicated that violence in programming remained rampant (Subcommittee to Investigate Juvenile Delinquency, 1961; Stenier, 1963; Johnson, 1967). In 1961, representatives from ABC, NBC, and CBS agreed to develop industry- wide research addressing the issue of the effects of TV violence on child- ren. In 1962, the Joint Committee for Research on Television and Children was co-sponsored by the broadcast industry and the United States Depart- ment of Health, Education and welfare. This committee requested research proposals from the scientific community. By 1964, however, it was appar- ent that few of these proposals were funded. In reality only three pro- jects were started and of these, only two were completed (Hartley, 1964; Feshback, 1971; Liebert, Neale & Davison, 1973). In 1968, the National Commission on the Causes and Prevention of Violence held hearings and questioned network executives about the pro- mised research. Network representatives at this time argued that the broadcast industry should not be responsible for conducting research ex- amining the effects of programming. These representatives suggested that research should be conducted through more objective channels. In addi- tion, the broadcast industry was skeptical of previous research conducted by social scientists due to the inadequacies and limitations of experi- mental research designs (Baker & Ball, 1969). In 1969, Senator John 0. Pastore, then Chairman of the Senate Sub- committee on Communication of the Senate Commerce Committee, expressed concern over the lack of information establishing a possible relationship between televised violence and anti-social behavior of viewers, especially young viewers. Senator Pastore wrote to Robert Linch, then Secretary of Health, Education and Welfare, concerning the feasibility of this cause and effect relationship. Secretary Linch directed William H. Stewart, then Surgeon General, to select a committee to review the evidence on the effects of television on children. One million dollars was appropriated to fund projects examining the effects of televised violence on children. These grants were under the auspices of the National Institute of Mental Health (Cisin, et a1., 1972). In January 1972, Jesse Steinfeld, then Surgeon General, released the committee's full report. The evidence indicated a causal relationship between viewing televised violence and post-viewing aggression; however, the wording of the report led to misunderstanding, an example of which is the classic New York Times article entitled, "TV Violence Held Unharmful to Youth" (Gould, 1972). In March 1972, Senator Pastore held hearings in an attempt to clari- fy the findings of the committee report (hearings before Senate Subcom- mittee on Communication, 1972). He concurred with the statement that there was enough evidence in existence to support a causal relationship between viewing televised violence and post-viewing aggression. Pastore concluded saying, "... and we ought to take it from there (p. 152)." Since 1972 several studies funded by government agencies and the broadcast industry have examined various aspects of television programs ming on children. The Surgeon General's study and subsequent research has demonstrated a variety of variables affecting the impact of televised violence on children. The 1974 hearings on violence and television examined the progress of the 1972 Pastore hearings. Dr. Eli Rubenstein expressed concern over the slow progress of reducing violence in television programming. During these hearings Dr. Liebert and others recommended that efforts be made to foster socially relevant television (U.S. Senate Hearings, 1974). One possible effort to develop socially relevant television would be to study parent-child and adult-child interaction and examine its effects on children's perceptions of television programming. Reports stemming from evaluations of educational programming have shown that parent-child interaction relating to program content tends to increase children's‘ ability to learn that content (Friedrich & Stein, 1975; Ball & Bogatz, 1970). Other studies have shown that adult-child interaction while view- ing anti-social programming tends to reduce post-viewing aggression (Baron, 1972). In the past few years researchers have been calling for more studies to examine how parent-child interaction may affect children's perceptions of anti-social television programming (Leifer, 1973; Smart & Minet, 1976). It is against the background of several Congressional hearings and volumes of data indicating that anti-social television programming affects post-viewing aggression, that this experiment was designed. Purpose of Study The purpose of this study is to test means by which the effects of anti-social programming on children may be mediated. The viewing of televised anti-social behavior and its effect on children has been a debated issue since the early 1950's. Anti-social behavior in television programming still exists today. At present, violence appearing in children's Saturday and Sunday morning programs is on the increase (Gerbner & Gross, 1976). Although the broadcast industry has made at- tempts to reduce violence in programs aired during the early evening hours, violence in post-9 p.m. programming is on the increase. Further, millions of children are viewing television past 9 p.m. (Neilsen, 1976). One obvious solution to the problem posited by Gerbner & Gross would be to remove violence from television programming; however, after twenty-five years it appears unlikely the broadcast industry would remove violence from television entertainment fare (Gerbner & Gross, 1976). Therefore, alternative means must be developed to counter the possi- ble effects anti-social television programming may have on children. The Problem This research is designed to provide an alternative and more immedi- ate means by which the effect of televised violence on children may be reduced. [:ensorious comments about televised violence have been known to reduce the effect of TV violence on childre (Berkowitz & Green, 1967; Berkowitz & Rawlings, 1963; Green & Stoner, 1972). As a result of evalu- ations of televised educational programs, parents have been found to facilitate knowledge gained by children by interacting with them about programming content (Ball & Bogatz, 1970). This study examines the ef- fects of the presence of parents, adults and peers as well as parent and adult verbal mediation of anti-social television programming on children's post-viewing behavior and also their ability to recall program content and consequences of violent behavior and perceived inappropriateness of anti-social behavior. Definitions relevant to the problem and the research hypotheses are as follows: 1) The viewing conditions consist of a mother viewing with her child and providing comments, a mother viewing with her child and providing no comments, a female adult viewing with a child and providing comments, 3 female adult view- ing with a child and providing no comments, a child view- ing with a peer and a child viewing alone. 2) Comments refer to describing the content of programs and judging the anti-social behavior as wrong, bad or inap- propriate. 3) Anti-social programming is limited to shooting scences taken from programs shown on commercial broadcast stations received in East Lansing, Michigan. Theoretical Framework and Research Hypotheses The theoretical framework for this study draws upon the facilitation and arousal models posited by Watt and Krull (1977). The facilitation model is based on the premise that children learn and/or model behaviors from viewing television programming. The rationale common to the facili- tation model is that the violent content of television programming facilitates the appearance of aggressive behavior. The arousal model has arisen out of a concern with the relationship between a person's physiological state and his cognitive state (watt and Krull, 1977, p. 99). watt and Krull describe how the arousal model operates ... "Its operation involves two distinct stages: first, the person is placed in a drive-enhanced state by some arousal procedure; then the social condition in which he finds himself determines his behavior (watt and Krull, pp. 99-100). The arousal model predicts a positive relationship between viewing anti-social programming and levels of aggressive behavior. Berkowitz and his associates have demonstrated that censorious comments directed toward anti-social content in films tended to reduce aggressive behavior in previously frustrated or otherwise aroused viewers. As a result of this finding by Berkowitz and others, this study proposes to examine the extent to which parental and adult verbal mediation of televised anti-social programming affects children's ability to recall programming content, consequences of anti—social be- havior, perceived inappropriateness of anti-social behavior and post- viewing behavior. The research hypotheses are as follows: Hypothesis 1: Children viewing anti-social programming with their mothers providing comments will be able to: (l) recall more pictorial content, (2) recall more consequences of anti- social behavior, (3) perceive violence as inappropriate and (4) will exhibit more post-viewing cooperation than those children viewing with their mothers providing no comments. Hypothesis 2: Children viewing anti-social programming with a female adult providing comments will be able to: (l) recall more pictorial content, (2) recall more consequences of anti- social behavior, (3) perceive violence as inappropriate and (4) will exhibit more post-viewing cooperation than those children viewing with a female adult providing no comments. Hypothesis 3: Children viewing anti-social programming with their mothers or a female adult providing comments will be able to: (l) recall more pictorial content, (2) recall more consequences of anti-social behavior, (3) perceive violence as inappropriate and (4) will exhibit more post-viewing co- operation than those children viewing with their mothers or a female adult providing no comments. Hypothesis 4: Children viewing anti-social programming with their mothers or a female adult providing comments will be able to: (l) recall more pictorial content, (2) recall more consequences of anti-social behavior, (3) perceive violence as inappropriate and (4) will exhibit more post-viewing co- operation than those children viewing with a peer. Hypothesis 5: Children viewing anti-social programming with their mothers or a female adult providing comments will be able to: (l) recall more programming content, (2) recall more consequences of anti-social behavior, (3) perceive vio- lence as inappropriate and (4) will exhibit more post-viewing cooperation than those children viewing alone. Limitations of the Scope of the Problem 1) 2) 3) 4) Only one type of anti-social behavior is examined -- shooting. The subjects will be six and seven years olds. Parents and adults will be female. The research design is experimental and, therefore, generali- zation of the results will be limited. CHAPTER II LITERATURE REVIEW Children and Television Viewing In the early 1950's, drama occupied most of television broadcast time. The largest subgroup of drama was crime-detective, providing one out of six minutes of programming in 1953 (Smythe, 1954). Smythe (1954) analyzed one television week in New York in 1953 and found that there were 3,421 violent acts and threats, an increase of 15 percent over the previous year. Entertainment programs accounted for 98 percent of this violence. Smythe also found tha hours when children might be expected to view were more violent than the rest of the television week. \ \. / TV Violence [After twenty years, television programming still contains anti- social behavior Gerbner and Gross (1976), in their annual violence index, concluded that there has been no overall reduction in the amount of violence in television programming in the past three years. During the 1975-76 programming season the National Association of Broadcasters introduced "family hour" (7-9 p.m. Eastern and Pacific Standard Time, and 6-8 p.m. Central and Mountain Standard Time). During this time Gerbner and Gross recorded a reduction in violence. This decline in violence during "family hour" has been matched by an increase in vio- lence during weekend children's programming and a two-year increase in past 9 p.m. Violence.' Gerbner and Gross (1976) describe television content as follows: 10 The overall rate of violence episodes, 8 per hour, is, if anything, higher than at any time since 1969. The violence saturation of weekend children's programs declined from the 1969 high, but increased from the 1974 low to 16 per hour, double that of overall programming. Between six and seven out of every ten leading characters (eight and nine for children) are still involved in some violence. Between one and two out of every ten are still involved in killing. (Page 187) @ the exception of "family hour," violence in television program- ming and especially children's programming is on the increa323\ Neilsen (1976) has reported that there are over 20 milli6fi>children aged 2 to 17 viewing television after 9 p.m. (TV Guide, 1976). In another study 29 percent of the children aged 4 to 12 were viewing past 9 p.m. on weekends (Fontes, et a1., 1976). These findings indicate that children are being exposed to periods of TV programming which contain an increase in violence. Children have some control over which television programs they view. Hannenman, et a1.,(197S) stated that "research on control of children's television behavior by parents indicates that parents generally do not control their children's TV behavior (p. 13)." Atkin (1972) reported data which suggests "that parents are not always aware of the television fare that their children are exposing themselves to (p. 1)." Television has been used as a device or means by which children are kept "quiet and out of mischief (Sears, et a1., 1957, p. 289)." Singer and Singer (1971) have found that generally parents pay little attention to content, extent, and impact of TV on their children's behavior.i:éreenberg and Dervin (1970) reported that children from lower income families devoted more time to television viewing and received less guidance about TV viewing from pareHEQB Greenberg, et a1.,(l972) reported that mothers made lower estimates than 11 their children about the amount of TV violence viewed, amount of total viewing, andZEEe extent to which TV is perceived as reality by childreg;:§ In addition, mothers made higher estimates of their role in program se- lection. Lyle and Hoffman (1972) in a study with first-, sixth-, and tenth-graders found that "although the majority of the first-grade mothers interviewed stated that they did try to guide program selection for their young children, few indicated that they attempted to restrict amount of viewing. About one-third of the students themselves said that their parents tried to control their viewing, either 'now' or 'when they were younger' (p. 134)." Rossiter and Robertson (1975) examined parental con- trols over TV viewing by asking parents and children about TV viewing behavior. They found that parents claimed 1) significantly lower viewing by their children; 2) stricter household rules governing viewing;z££:more co-viewing with children, and; 4) greater parent-child interaction than reported by the child themselves. In addition, viewing time and pro— grams selected were largely a decision of the child (Hess and Goldman, 1962; Lyle and Hoffman, 1972). Working-class parents reported that their children were heavy viewers of TV programming when compared to middle- and upper-class parents. In short, children are frequently viewing tele- vision without parental supervisiog The ability of children to distinguish television programming from reality is associated with age (Lyle and Hoffman, 1972). Hirsch (1969) i:£§und that children aged six are more~1ikely to say television represents reality than 11 and 13 year old Chiqufsl Flapan (1968) reported that children's range and degree of understanding of feelings, intention, and thought of movie characters increased between the ages of 6 and 12. Children and teenagers from lower—income families and black families 12 expressed greater belief that television accurately portrays real life (Greenberg and Dervin, 1970). Young children perceived selected TV pro- grams to be more violent when compared to mothers' perception of that same program (Abel and Beninson, 1976). Collins, Berndt and Hess (1974) reported that children in kindergarten and second grade tended to re- member aggression alone or with its consequences while older children remembered motives as well as consequences. S:Eéfsummary, younger children are more likely to perceive TV as real when compared to older children. In addition, they are capable of per— ceiving programming as violent and can recall aggressive acts and their consequences, but are unable to comprehend the motives of aggre . This section illustrates the point that violence does exist in tele- vision programming and children are being exposed to this violence with varying degrees of parental supervision. These children have abilities to perceive programming as violent, and children six years of age or younger are more likely to perceive television as real than any other age group. Models of the Effect of TV on Children What effect does exposure to anti-social programming have on child- ren? In answering this question, researchers have developed three basic models: 1) catharsis, 2) arousal, and 3) facilitation.l Defined, these models are as follows: catharsis model -- the reduction of aggression through vicarious participation in television violence; arousal model -— television programming serves as an agent of cognitive arousal, generating 1For the purpose of this study, the labels and definitions of these models are the same as those posited by Krull and Watt (1973) and watt and Krull (1977). 13 a predisposition to action which is shaped by situational variables; and facilitation model -- the learning or modeling of aggressive behavior from televised violence. The catharsis model has fallen into disrepute, while the arousal and facilitation models have served as the basis for the vast majority of research into the effects of televised violence on children. The Catharsis Model The catharsis model's main advocates have been Seymour Feshbach and Robert Singer. Feshbach (1961) presented his model during the time of the Dodd Hearings. His was one of the two completed studies funded by the Joint Committee for Research on Television and Children. The basic pursuit of this model is that as children view violent programming they become vicarious participants, and this serves as an outlet for their aggressive feelings. The result is reduced observable post-viewing ag- gression. Feshbach and Singer (1971) reported an association for boys in boys' schools between exposure to violent television programming and lessened aggressiveness toward peers. Those that viewed non-violent programming became more aggressive. They concluded that the findings support their hypothesis that TV violence can serve as a substitute for aggressive behavior, which serve as a means to assist children in controlling their aggression. Liebert, et a1., (1973) responded to this study by pointing out its weaknesses. One of the main weaknesses is "the differential liking and/ or initial preferences for the two TV diets may have led to frustration or resentment and thus to greater aggression on the part of the control 14 group (p. 350)." They also pointed out methodological and design weak- nesses. Wells (1973) essentially replicated Feshbach and Singer's study and found increased verbal aggression against peers in the non-violent film condition and increased physical aggression against peers in the violent film condition. These findings are contrary to those of Feshbach and Singer. Konecni (1975) found some support for the catharsis model. The dis- tinction was made between levels of arousal (anger) and amount of ag- gression expressed. The level of anger was reported as increasing; however, aggressive behavior was not stimulated. Manning and Taylor (1975) found support for Konecni's study. Their study contained both an aggressive measure and a hostility measure. Manning and Taylor reported that hostility (emotion) was reduced and aggressive behavior was not stimulated after viewing the violent film. Drabman and Thomas (1975) found that televised violence may have a dual effect: 1) exacerbate some children's violent behavior; and 2) teach others to tolerate aggressive behavior. The latter findings fit the general premise of the catharsis model. This approach had had few followers over the years. The premise of the catharsis model is not generally accepted. An overwhelming amount of research reports that televised violence does have "some" effect on children's aggressive behavior. The Arousal Model The arousal model's premise is that television programming serves as an agent of cognitive arousal, generating a predisposition to action which 15 is shaped by situational factors. It is the "arousal" aspect of tele- vised violence that instigates post-viewing aggression and not the programming itself (Tannenbaum, 1972). A difficult objective of this approach is the measurement of "arousal." The assumption is made that emotional arousal is not specific to a particular emotional state (Schachter, 1967). Arousal is typically indexed by physiological mea- sures, for example, heartbeat rate, blood pressure, skin temperature, and galvanic skin response. Zuckerman (1971) has demonstrated that there are dramatic individual differences in the specific types of reactivity. Some individuals may demonstrate arousal by increased heart rate, while others show varied skin temperatures or respiration rates. These vari- ations indicate a lack of consistency in the measure of arousal, and makes comparisons between and within groups very difficult. Another assumption of this model, according to Tannenbaum (1972), is that "heightened arousal, no matter how it is perceived, should lead to higher levels of response, no matter what kind of response behavior is called for (p. 307)." Most of the experimental research to date examines an aggression measure response. Causal inferences are then made from induced anger, or frustration, to post-viewing aggression. Studies have varied the film content, while retaining a post-viewing aggression mea- sure (Zillmann, 1969; Tannenbaum, 1972). Schachter (1964, 1967) provides the addition of a cognitive compo- nent to the arousal model. This addition postulates an interaction between cognitive and physiological factors in determining emotional behavior. The assumption that "arousal" is a general emotion continues to be made, however, the individual utilizes cognitive information to exhibit specific behavioral responses. The individual cognitively labels 16 the arousal state and the assumption is made that behavioral responses are congruent with the labeled "aroused" state. Tannenbaum (1972) has tested this cognitive component. Subjects were angered by the experi- menter giving shocks to them prior to viewing a violent film. After viewing the film, subjects were allowed to give shocks to the experimen- ter and other people. The experimenter received significantly more shocks from the angered group than from the non-angered group. Tannen- baum and others proposing the arousal model have traditionally used college students as their subjects. Those accepting a Piagetian per- spective to child development accept cognitive development as a function of age; therefore, different results would be expected to occur when subjects are from different cognitive levels. In short, anti-social programming is seen as a possible catalyst for aggressive behavior in previously aroused subjects. One important aspect of the arousal model that has been incorporated in a few studies is the role situational factors have in generating a predisposition for ag- gressive behavior. Comments and Post-Viewing Behavior The following studies demonstrate how verbal comments can affect post-viewing behavior. This body of literature is reviewed to provide support for the hypothesis of verbal comments affecting post-viewing cooperation. Many studies have been conducted under the theoretical framework of the arousal model (Berkowitz and Green, 1966; Green and Berkowitz, 1967; Hartmann, 1969; Doob and Kinshenbaum, 1973; Zillmann and Johnson, 1973; Hanratty, et a1., 1969; Baron and Ball, 1974). All of these studies de- monstrate that subjects who were frustrated, angered or otherwise aroused 17 behaved more aggressively after viewing a violent film when compared to equally aroused subjects who viewed non-violent films. A few studies have attempted to modify the viewing situation by providing verbal com— ments about the film to the subjects. In all cases where comments were made condemning the anti-social behavior viewed by the subjects mea- sured post-viewing aggression was lower compared to subjects in condi- tions where verbal comments praised the anti-social behavior. Berkowitz and Rawlings (1963) angered subjects prior to the viewing of a seven-minute fight scene from the movie Champion in which one of the actors was severely beaten. The authors introduced an additional cogni- tive component, justification of anti-social behavior. The authors achieved the introduction of "justification" by presenting different synopses prior to viewing the film. In one condition the actor who was beaten was depicted as a "scoundrel," and in another condition he was described as a person who was shaped by his environment and was beginning to show remorse for his misdeeds. Subjects who viewed the film in the unjustified condition ("scoundre1" condition) displayed more aggression toward the person that had angered them than those exposed to the justi— fied condition. The justified/unjustified verbal comments provide a means to mediate the aroused state of viewers. This research project incorporates censorious comments and measures the effects these comments will have on post-viewing behavior as well as the learning of program content and the perceived inappropriateness of anti-social behavior. Lefcourt, et a1., (1966) found that subjects who had been exposed to censorious comments made by an experimenter about a violent film were not as likely to increase intensity of electric shocks they administered as were the group of subjects not exposed to censorious comments. In this l8 experiment subjects were tested to determine their level of aggression prior to viewing a segment of the film Rebel Without a Cause. Also prior to viewing the film, subjects were told: 1) that the film they were about to see would be enjoyable, one with plenty of action or 2) that they would be seeing a film about a bunch of hoodlums that ought to be thrown in jail. The experimenter also made comments during the viewing situation either praising or condemning the actions of the aggressors in the film. Those who heard the censorious comments (condemnation of behavior) demonstrated less post—viewing aggression than those who did not hear the censorious comments. Berkowitz and Green (1967) essentially replicated Berkowitz and Rawlings' (1963) study. Subjects who viewed a justified violent film administered more post-viewing shocks than those who viewed an unjusti- fied violent film. When the confederate who was receiving the shocks was identified with the victim in the film a greater amount of post-viewing aggression was found. The authors conclude that available target persons associated with the victim of observed violence received more attacks from angered individuals than other possible targets lacking this asso- ciation. Both of these studies demonstrate the impact of mediating com- ments on post-viewing aggression, a variable that will be included in this study. Green and Stoner (1972) found support for earlier arousal studies using justified/unjustified comments. In this experiment subjects were previously angered by receiving electrical shocks. Prior to viewing a beating scene between two professional fighers in the movie Champion, subjects were told that the fighter in the movie had been savagely beaten by the other fighter in a previous match and wanted to get even,or the 19 subjects were told that no hard feelings existed between the two fighters and that each was just doing his work. In this experiment, vengeful vio- lence (unjustified) produced more aggression than professional violence (justified) for previously angered subjects. Again, the results indi- cated that verbal mediating comments lessened post-viewing aggression. Parke and Ewall (1975) investigated the relationship between verbal cues and motor aggression. In their study college males were verbally reinforced for selecting and speaking either aggressive, neutral or help— ful words. On a subsequent test for nonverbal aggression in which sub- jects had an opportunity to shock another individual, those reinforced with aggressive verbalizations were more aggressive than those in the neutral-word group. The authors conclude that these findings were con- sistent with the view that both the elicitation and inhibition of non- verbal aggression can be regulated by verbal cues. The above studies demonstrate that censorious verbal comments can reduce post-viewing aggression. The problem with the arousal model is the difficulty encountered when attempting to measure arousal level in subjects. In addition, the age of the subjects, type of program stimu- lus, and measures of aggression introduce a host of variables contribut- ing to the difficulty of studying the arousal model. With the exception of censorious comments, the arousal model would be impractical in the home viewing situation. It simply would be too difficult for parents to determine the level of "arousal" of their children prior to viewing television. Finally, research conducted utilizing the arousal model has focused on aggressive behavior as the dependent measure. Onacould ask how arousal, censorious comments, and violent program content affect 20 children's ability to recall content, to recall consequences of anti- social behavior, and to perceive inappropriateness of anti-social pro— gramming. Although the arousal model adequately describes how verbal comments can reduce post-viewing aggression, it does not answer the questions of recall and perceived inappropriateness. Answers to these questions are found in the body of research utilizing the facilitation model. The Facilitation Model The basic premise of the facilitation model is that violent content of television programming facilitates the demonstration of aggressive behavior in the viewer. The word "facilitation" was developed by Krull and Watt (1973) to incorporate tenets of the social learning, incidental learning, and modeling-imitation theories under one theoretical per- spective. The facilitation model implies that learning of problem- solving techniques takes place by viewing television. Watt and Krull (1977) place the facilitation model at the cognitive end of an emotive theory-cognitive theory continuum. (An example of emotive theory would be the catharsis theory.) Another distinction made by Watt and Krull (1977) is that the learning process tends to be long-term. By this, Watt and Krull mean that the long-term effect of viewing violence is a slow accretion of learned violent methods. In the facilitation model the pro- gram attribute of overt violent content is important, since this model. depends on a learning/modeling process. Bandura and Walters and their associates have examined children and television viewing utilizing learn- ing theory. The theory behind Bandura's (1973) work assumes the level and type of aggressive behaviors displayed by children are not instinctual, but 21 learned. Frustration and emotional arousal facilitate aggression, but are not conditions necessary for its occurrence. Bandura suggests that neither instinct theory nor the hypothesis that aggression is always attributable to prior frustration seems to fit the evidence as well as social learning theory. From the perspective of social learning theory, an observable aggressive behavior depends on the prior learning of that behavior, the presence of environmental stimuli which would elicit ag- gressive behavior, and the anticipated consequences of aggressive behav- ior. These three factors are dependent on what has been learned from direct experience and/or from observing others in real life, or in tele- vision programming. This approach takes the individual into considera- tion because the process of "learning" depends on the mental processes of the individual, prior learning and current stimuli. Each of these factors has its effect on the individual's beliefs, feelings and know- ledge. In short, the issue that Bandura raises is one of acquiring aggres— sive behaviors. These aggressive behaviors may be learned by directly observing others either in real life situations or on television and by personal experiences. Weiss (1969), after viewing the literature, stated that "there is little doubt that, by displaying forms of aggres- sion or models of criminal and violent behavior, the media are 'teaching' and people are 'learning' (p. 77)." The performance of learned aggres- sive behavior is dependent on individual characteristics and environmen- tal stimuli. Although aggressive behavior may be learned by viewing television the performance of aggressive behavior may be mediated by the environ- ment. Environmental factors provide a practical means by which the 22 potential effect of televised violence on children may be controlled. The question is raised: what factors are effective in mediating the effect of television programming on children? Research supporting the arousal model demonstrated how censorious comments provided by the experimenter reduced post-viewing aggression. The facilitation model also takes into consideration factors affecting learning from television. The fact that children are capable of learning from television has been demonstrated by many researchers. Much of this learning is in terms of incidental learning and social learning, modeling or imitative behavior. Incidental Learning The incidental learning perspective contains a body of literature demonstrating that television contributes to children's acquisition of general knowledge (Schramm, Lyle and Parker, 1961). The initial knowledge gained by children who view television is not maintained as they grow older. The learning that takes place is related closely to the principle content of television programming (Himmelweit, Oppenheim and Vince, 1958; Furu, 1971, 1967). Incidental learning as defined, is the learning of stimulus material not essential for completion of a specific task. Inci- dental learning is found to increase with age until the child is about eleven years old, after which incidental learning is known to decrease (Stevenson, 1954; Maccoby and Hagen, 1965; Siegel and Stevenson, 1966). The principal measure of incidental learning in media studies has been the ability to recall the visual and verbal content of the stimulus. In light of the facilitation model, incidental learning research considers the cognitive developmental level of children. In addition, factors in 23 the social environment and program content are known to have an effect on the degree of incidental learning. These factors are the age of child, spatial relationship of objects appearing on the television screen, and color of programming content. (Hagen, 1967; Siegel and Van Cara, 1971; Hawkins; 1973; Wheeler and Dusek, 1973; Hale and Tawell, 1974.) Social Learning Another vast body of literature examines social learning theory, imitation and modeling. The following studies demonstrate children's ability to learn from television. These studies show the role of child- ren's cognitive capabilities and environmental cues in the learning processes. Maccoby and Wilson (1957) found that seventh—graders tended to remember the words and actions of their object of identification better bered aggressive content of programming better than girls when the aggres- when the behavior was relevant to the viewer. For example, boys remem- sor was a boy. The results suggest viewers often identify with a movie character and this in part determines what the viewer learns from the film. The findings of Bandura, Ross and Ross (1961) suggests that identi- fication with a model increases the likelihood of imitating the behavior of the model. Nursery school children exposed to aggressive models repro— duced many of the aggressive behaviors performed by the models. Imitation was differentially influenced by the sex of the model. Boys tended to identify with the male model more than girls. They conclude that subjects given an opportunity to observe aggressive models later reproduced a good deal of the verbal and physical aggression exhibited by the model. They 24 also suggest that the observation of adult models creates justification or permission to imitate the behavior of the model. Siegel (1968) found that children generalize from the content of the mass media to their social world. Children exposed to an aggressive radio drama about a taxi driver used more aggressive language in completing stories about taxi drivers and attributed more aggressiveness to taxi drivers in general. The media are then capable of presenting children with aggressive models from which they may generalize to their social environment. Flanders (1968) found that children may well imitate cartoons, filmed adults, or live models. The likelihood of imitation is increased when the model is rewarded. Flanders concluded that an observed behavior is more likely to be imitated by a child if the imitative behavior is rein- forced or if the observed model's behavior is reinforced. Hanratty, et a1. (1969) demonstrated that imitative aggression occurs without prior frustration or anger. The amount of aggression exhibited by kindergarten boys appears dependent on both the presence of aggressive modeling cues and the type of available victim. Goranson (1970) found that aggressive behavior may be learned by children from exposure to realistic aggression on television. If the aggressive TV sequence is practiced by children there is a greater chance of retaining that aggressive behavior. The performance of aggressive ‘$\ acts learned from television is partially dependent on the similarity /§U between the setting of the media violence and cues in later situations, and by the perception that aggression is effective and will not result in punishment. This study examined environmental situations as factors affecting the modeling of aggressive behaviors viewed on television. 25 Leifer, et al. (1971) examined the cognitive ability of young child- ren (five to seven years of age) to reconstruct sequences of an entertain- ment film and to understand the emotions of the characters. Their findings are similar to the incidental learning perspective in that the ability to reconstruct sequences and correctly identify emotions is dependent upon age. There was no evidence that children understood the feelings and motivations of same-sex models better than opposite-sex models. This research also suggests that although children may model behavior they are incapable of understanding the emotions and motivations of the models. Stein and Bryan (1972) examined rule violation behavior in third- and fourth-grade girls. These girls were exposed to television portrayals of a model obeying the rules of a game and another model advocating cheat- ing. Girls who heard the model verbalize rule conformity were more accurate in recalling the rules of the game compared to those exposed to the model who cheated. The results suggested that rule violations were affected by the model's level of success. The authors conclude that behavioral transgressions of the children increased when the child was presented with either preachers or practitioners of transgression. Much of television entertainment fare portrays aggression as transgression. Meyer (1973), found that television characters provided socially desirable social models, especially for females. Children perceived their favorite TV characters behavior in a socially desirable manner when presented with conflict situations. Meyer also found a sizable minority of males had favorite TV characters who provided undesirable models of behaviors. These boys perceived that the violent means of conflict re- solution demonstrated by their favorite TV characters was appropriate and 26 worthy of imitation. Meyer reported that parents of these children fre- quently would neither recommend nor condone a violent response, but these children see themselves as making violent responses which they judge as being right. These findings demonstrate that children select violent TV characters as social role models. Collins (1974) found that if the "bad guy" had some saving qualities, children were more likely to imitate his/her behavior. Children exposed to television programs depicting a model coping constructively with a conflict situation tended to be more helpful toward unseen peers involved in a problem-solving task than children exposed to a model demonstrating aggressive behaviors. Collins concludes that television models affect children's willingness to help or hurt other children. Television models can instigate helpful and aggressive behaviors in children. Baron (1974) examined qualities of the individual that affect the degree to which modeling is likely to occur. Baron found that a general relationship between self-esteem and modeling from television was sup- ported. High self—esteem males tended to model aggressive behaviors to a significantly greater degree than females or low self-esteem males. Sex differences in modeling are found only with regard to aggressive be- havior. Baron demonstrates that elements of the individual's personality affect the types of behavior that may be modeled. This research demon- Strates personality characteristics affecting modeling behavior. Dollinger and Thelen (1975) support Bandura's theory that model con- sequences are crucial to vicarious learning. Children seeing pro-social and anti-social behavior rewarded were more likely to display imitative responses than children not seeing the adult model's behavior rewarded. 27 Fouts and Liikanen (1975) assessed the effects of age and develop- mental level on the use of imitation by children. Young children, aged five, at a higher developmental level imitated televised models more than their less mature agemates, whereas older children, aged eight, with a higher developmental level would imitate less than their less mature agemates. Their findings suggest that imitative behavior is also depend- ent on the attributes of the subject, in this case, developmental level and age. In summary, the facilitation model examines to what extent aggressive behaviors may be learned by watching violent television programs. The review of the literature included research that demonstrates children's ability to learn non-essential material contained in programming as well as behaviors to be learned, modeled or imitated. The research findings suggest that the degree to which children learn from television program- ming is dependent on many factors, e.g., age, cognitive ability of children, children's social environment and similarity between televi- sion's programming settings and children's real life settings. The fact that children's social environment has an effect on learning is important in terms of providing practical means to counter the effect of televised violence on children. Reviewing the literature utilizing the arousal model has shown how censorious verbal comments made by the experimenter reduced post-viewing aggression. In a similar manner a few studies have shown that parents or adults are capable of facilitating children's ability to learn from television by interacting with children in the viewing situation. 28 Verbal Comments and Learning from TV The following studies demonstrate how verbal comments can increase a child's ability to learn from television programming. This body of liter- ature is reviewed to provide support for the dependent measures of recall of pictorial content, consequences of anti—social behavior and verbal content. Hicks (1968) found support for the effect of co-viewers on illicit- ing appropriate behaviors from children. In an experiment with children seven years of age Hicks found that positive and negative verbal sanctions by an adult co-viewer produced corresponding increases and decreases in imitative aggressive behavior when the adult co-viewer remained with the children who had viewed a film of an adult male behaving aggressively against a Bobo doll. The children were then observed in a play setting. If the observer remained in the play room children's behavior corre- sponded to the verbal sanctions. Where negative verbal comments were made the children displayed less imitative aggressive behavior and when the filmed behavior was positively sanctioned the children displayed more imitative aggressive behavior. The imitative behavior was less likely to occur when the adult co—viewer was absent during the play period. The author concludes that the adult co-viewer provides cues as to the appro- priate behavior in given situations. In 1970 and 1971, Ball and Bogatz published findings of their eval- uation of "Sesame Street." One of the main findings of their studies was that the viewing of "Sesame Street" resulted in gains in some skills, but the range of skills for which there were gains was far wider for child- ren whose viewing was associated with outside encouragement to view. The authors found in their evaluation of "Sesame Street" that children who 29 rehearsed the concepts they watched on "Sesame Street" with their mothers, learned more than children who watched by themselves. McLeod, Atkin and Chaffee (1972) found that parents can mitigate some of the harmful influences of television violence by interpreting the vio— lent acts as they are portrayed. Children were asked how often their parents told them that: (l) the content in television programming is "just pretend," (2) the things appearing in TV programming are not like things happening in real life, and (3) there are better ways than violence to solve problems. The more parents provided these types of comments the lower the level of children's self—reported aggressive behavior. Atkin and Gantz (1974) discovered that parental explanation and am- plification of news program content tends to increase the child's acqui- sition of current events knowledge. Gelfand and Hartmann (1975) suggest that parents may mediate the response of children to television sports. The authors conclude that models, whatever their behavior, play a major role in determining the nature of boys' responses to frustration induced by competition. The approving or admiring ways in which broadcasters comment on unsportsman— like behaviors may contribute to undesirable behaviors in children. If the model of reaction to competition is the parent, he or she may deter- mine the way the child will confront competitive situations. The parent may even mediate the response of the child to television sports by pro- viding a model of response to what is watched. Friedrich and Stein (1975) examined how interacting with children about the content of a television program increases children's scores on verbal labeling and nonverbal helping behaviors. As a result of their experiment the authors concluded that verbal labeling is more important 30 for learning or acquisition of observed content, while role play training is more important for performance of learned behavior. Another important aspect of this study was the ability of children to model behavior in situations similar and dissimilar to those situations portrayed in the program. Walling (1976) conducted an experiment with first graders. Children were assigned to three groups at random; interaction with parents, no interaction and a control group. After one week, the children were inter- viewed in order to assess their learning of problem resolution from three television shows. Walling found that children who interacted with their parents were significantly better able to complete problems which paral— led those they had viewed on television, than children who did not inter- act with parents during the viewing and children who did not watch television at all. In summary, the above literature demonstrates that verbal comments by parents and unrelated adults about television programming content in the viewing situation with the child facilitates learning. Two perspec- tives on learning by children will now be discussed. These perspectives will provide a theoretical understanding of how verbal comments can mediate what children learn from television. Theories of Child Development This study draws upon two theoretical frameworks in the study of child development, cognitive development and learning theories. Cognitive development theory provides an understanding of the mental capabilities of children six to eight years old. Learning theory examines the impor- tance of environmental factors in determing children's behavior. 31 Piaget (1948) is best known from his development of cognitive devel- opment theory. Piaget is also known for his age-stage theory which describes four main stages of child development. The sensorimotor stage is said to be the first two years of life. During this stage. motor activity becomes coordinated and children begin to notice events outside their own bodies. It is in this stage that goal-directed activity is beginning to develop. By the end of this stage some degree of problem- solving is possible. During the second, preOperational, stage children conceptualize more, construct more complex thoughts and images, and group objects into classes. Children can use symbols, but they cannot organize them into broad concepts. Typically, children have no mental representation of sequences of events. During the third, concrete oper- ational, stage reasoning processes begin to be based on logic. Children are capable of classifying objects not only by labeling objects but also by mentally combining objects within a class. During the final, formal operations, stage truly logical thinking emerges. Children now take the final step toward abstract thinking and conceptualization. They now consider general laws and think about what is hypothetically possible as well as what is real (Piaget, 1948). Children participating in this study are in the transition period from preoperational stage to concrete operational stage. During these two stages, children use logic and reason in an elementary way and apply them to the manipulation of concrete objects, but not to verbal proposi- tions (Mussen, 1963). Therefore, children in these stages may have difficulty in cognitively: (l) ordering the events occurring in tele- vision programming, and (2) processing the abstractions of program plots or stories. 32 The cognitive development theory provides explanations for limita- tions on information a child can cognitively process. In addition, this theoretical perspective concentrates on the child's mental processes and capabilities. Although cognitive development theory emphasizes the mental/cognitive structures of the child, environmental factors must also be considered in a child's learning process. Learning theory examines the importance of factors external to the child in facilitating a child's development. Sears, et a1. (1957) accept the notion that the child has innate drives; however, these drives are modified by the child's socializing environment and become secondary motivational drives. These secondary motivational drives determine the individual's behavioral patterns. Children as young as six years old are capable of internal control of motivational drives (Sears, 1957). To Sears, child development is dependent on the dyadic interaction between the child and his/her caregivers. As the child develops, he relates himself more and more to the environment. Sears assumes that identifi- cation with caregivers becomes a motivational system which drives the child to behave as expected. The child's outward behavior reflects the product of the social environment. Learning theorists examine stimulat- ing factors which have an immediate behavioral impact upon the child, regardless of whether or not they enter within the child's cognitive threshold. These theorists consider only the latest social experiences as the major behavioral determinants which replace or reinforce previous behavior. These experiences establish the essential patterns for deal- ing with later life events. One important factor in the child's environment which influences social experiences and identification is verbal communication. The fact 33 that newborn infants make overt movements in synchronized response to adult speech suggests that, from the moment of birth, there is a complex behavioral interaction between infant and caregiver (Condon and Sander, 1974). As the child develops, language and verbal communication skills have an increasingly important role in processing information to be learned. An eclectic approach using both cognitive development and learning theories provides insight into children's learning capabilities (mental/ cognitive processes) and environmental factors that have an impact on what children learn (identification, verbal communication). Television programming and parents are part of a child's social environment. Pro- gramming content, as part of the child's environment, can be learned by the child and parents are capable of mediating information to be learned (Atkin and Gantz, 1974; Friedrich and Stein, 1975). The extent to which a child can process information to be learned is dependent upon his/her stage of cognitive development (Piaget, 1948). Therefore, both theore- tical approaches are important in understanding the role which parents, adults and verbal comments have on children's ability to learn from tele- vision programming. Developmental and LearninggTheories -- Moral Development Just as Piaget (1948) has the age-stage theory of cognitive develop— ment, Piaget and Kohlberg have stage theories of moral development. Piaget perceives three forces as interacting to produce moral change: adult constraint, peer-group cooperation, and changing character of the child's mind (maturation). These forces determine the child's progress through the first two stages of development. The first stage, moral realism, is from birth until about age seven. Judgment by a child of a 34 deviant act is in terms of the damage done by the act. According to Rogers (1977) ". . . children at this stage see adults as dominant and omnipotent and see rules as unchanging and to be obeyed without questions (p. 287)." In the second stage, rules are seen in terms of their intent and what they are supposed to accomplish (restitution). Children still judge people absolutely alike. Later in this stage, children understand the need for the differential application of principles (equity) to ensure real justice. In the third stage, beginning in late childhood, children come to accept and follow the morals and values of the society. The final stage occurs during adolescence. During this stage a child's moral code is based on principles that have been thought through and tested. Personal conscience, not external authority, becomes the final decisive factor of moral behavior. Kohlberg (1975) subdivides moral development into three main stages. In the first, preconventional, stage the child responds to cultural rules and to labels of good and bad, right and wrong. The child's responses to these rules are in terms of consequences of action (punishment or reward) or of the physical power of those who establish the rules. During the second, conventional, stage the child perceives rules as having worth in their own right, regardless of their consequences. The child's attitude is not just one of conformity to social order, but of loyalty to it and identification with persons involved in it. During the third, post- conventional, stage the child tried to define moral principles and values, apart from authority of persons holding these values. Kohlberg (1964) also discusses moral judgments as existing separately from.moral behavior. He also mentions that moral judgments and behaviors are positively correlated. Training in moral judgment and behavior begins 35 at birth in the home via parental models and parent-child interactions. An important determining factor in the demonstration of moral judgments and behaviors is the child's stage of moral development. Turiel (1964) demonstrated that children are capable of assimilating adult moral reasoning at one level above their functioning stage more readily than moral reasoning two levels above their functioning stage. Kohlberg (1964) states that the development of higher levels of moral thought is depend- ent upon earlier attainment of each preceding moral stage. Children participating in this experiment are in Kohlberg's preconventional stage of moral development. Environmental factors such as parent-child verbal interaction have an impact on which moral judgments and behaviors are to be learned. Learning theory's approach to moral development focuses on these environ- mental factors. The learning theory approach to acquisition of morality and con- science formation supposes moral development to be a matter of good habits, learned through example, teaching and reward; or a matter of bad habits that are prevented from being practiced. Research on morality has been influenced by ideas of learning through conditioning; for example, by parental reward of "good behavior" and punishment of "bad behavior" (Breckenridge and Vincent, 1966). It is assumed by these individuals that "moral behavior" learned at home would be generalized to unsuper- vised situations outside the home. Research indicates a relationship between the modes of discipline in the home and the capacity for moral decision making (Glueck and Glueck, 1950; Bandura and Walters, 1959). Peck and Havighurst (1960) also con- cluded that parental nurturance and influence play a major role in the 36 development of moral judgment and behavior. Further discussion on famil- ial effect on moral development is included under "Family Effects" later in this chapter. In review, the developmental approach to moral develOpment discusses the importance of moral judgments made by a child being based on immediate consequences of an act and the capability of children to comprehend good and bad, right and wrong. In addition, learning theory provides an under- standing of the impact of environmental factors, e.g., parent-child inter- action, on children's moral judgment and behavior. Child Development and Verbal Communication Two theoretical frameworks in the study of child development, cogni- tive development and social learning, are important in understanding the role verbal comments have on children's ability to recall programming content. Cognitive development theorists focus on the stages of mental development individuals go through from birth to adulthood. Children, six to eight years of age, participating in this study are in the chrono— logical age groupings that would be associated with the transition from preoperational to concrete operational thought (Piaget, 1948). Piaget (1948) has listed structures that separate stages of cognitive development. Two structures that are particularly important to the study of children and television are preceptual boundedness and centration. Perceptual boundedness is a major characteristic of children in the pre- operational stage of development, and refers to children's ability to respond to their perceived environment. In this stage, the preoperational child's conceptual framework is based in his/her own subjective percep— tions and judgments are not logical. The concrete operational child is 37 capable of logical operations, and has an orderly conceptual framework which he/she can systematically bring to bear on the environment. Centration refers to the ability of children to focus on a limited amount of available information. The preoperational child focuses on a few dimensions of each stimulus and thus only partially comprehends the total environment. The concrete operational child is capable of decen- tration, the ability to conceive of a spatial, temporal, and causal world of events existing independently of the child. Piaget (1948) mentions that the amount of interaction a child has with others repeatedly forces the child to re-evaluate his/her frame of reference or conceptualization, thus facilitating the transition from pre- operational to concrete operational stages. Language becomes an increas- ingly important means of interaction with others as the child grows older. Jones and Gerard (1967) discuss the importanxaof human language as a means to facilitate a child's definition of the environment in which he/ she lives. Elkind (1969) found that verbal communication increased per- ceptual discrimination in children aged five to seven years. Belmont and Butterfield (1969) found that verbally describing certain aspects of material presented to children increased the amount of material they were able to recall. According to the cognitive development theoretical framework, then, verbal interactions are important and necessary for children to remove their perceptual boundedness and centration. Learning theory also examines the effectiveness of verbal comments on children's ability to learn from television programming content. Learning theorists place emphasis on the environment for the stimulat- ing factors which have an immediate behavioral impact on the child. 38 Environmental symbols and events are essential stimulators for internally motivated overt behaviors. Sears (1957) discussed the importance of ver- bal mediation as a source of controlling a child's behavior as he/she grows older. Verbal comments have the potential of mediating environ- mental cues a child perceives. In addition, Rau (1964) found that the source of verbal mediation has an influence on the effectivess of com— ments. Parents and adult authority figures, e.g., teachers, have the greatest impact on children in the early grade school years. Aronfreed (1969) states that verbal mediation focuses a child's at- tention and cognitive representation more efficiently to enchance his/her learning of specific behaviors. He goes on to say ". . . the use of a verbal medium will enhance learning not only because it can focus the child's attention, but also because it introduces a very powerful mode of representation of the form of observed behaviors and its critical features for the control of outcomes (p. 275)." Sigel (1964) states that the comprehension of language "determines how the environment is disciminated, what objects can be integrated, and what kinds of abstract concepts can be invoked (p. 233)." Evidence also suggests that verbal comments are effective in changing moral judgment and behavior. Grusec and Skubinski (1970) found that verbal comments increased the helping behavior of early grade-school girls. Wolf and Cheyne (1972) found that verbalizing altru- istic standards increased prosocial verbalized judgments in children. In short, verbal comments mediate the environment to which the child responds. In this study, it is assumed that verbal comments made by mothers and an adult will be effective in mediating television programming con— tent. An eclectic cognitive development/learning theory approach provides an understanding of why verbal comments can effectively mediate TV 39 programming. Verbal comments can reduce preceptual boundedness and increase decentration. In addition, comments can mediate environmental cues in terms of material and behaviors to be learned. Need for Research on Verbal Mediation of TV Programming In a paper presented by Singer and Singer (1974) to the American Psychological Association, the authors encourage further research on how parents and adults can mediate the effect of television programming. The authors designed an experiment to determine the effect of viewing "Mister Roger's Neighborhood" with and without an adult intermediary on children's subsequent spontaneous imaginative play. The authors found that parents generally paid little attention to extent, content and im- pact of their children's viewing behavior. Mothers of children who had exhibited a greater degree of spontaneous and creative play following viewing of "Mister Roger's Neighborhood" were found to report a self ori- entation built around an emphasis on independence and internally oriented values. These mothers tended to take a more active role in monitoring the TV viewing of their children and placed greater emphasis on viewing of non-aggressive type programs than did mothers of children who had exhibited a lower degree of spontaneous and creative play behavior. The authors conclude that viewing of such "low key" educational programs as "Mister Roger's Neighborhood" seems to be most effective when viewing is done with an adult present to act as an intermediary. In addition, the authors stress the implications of their findings to other types of child TV programming. Liefer, et a1. (1974) also stress the critical need to develop tech- niques which allow parents to have an impact as co-observers on aggres- sive programming viewed by children. 40 Walling (1976) says that parental interaction does not create learn- ing from television . . . "but it does increase and guide the potential learning which may already be taking place within the normal habits of the child. Educators and parents might both benefit by forming a "partnership" and using the home viewing situation to complement and supplement the child's formal education. Moreover, learning through inter- action with a significant adult can be extended beyond the parent to the teacher who utilizes television creatively in order to com- plement traditional classroom study . . . entertainment programs on prime time TV are particularly well suited to communication education curriculum at all grade levels (p. 24)." There is a need for further research into effective means by which parents and adults can mediate television programming content. The research that exists indicates parents and adults are effective as medi- ators of television programming. Walling's comment indicate the import- ance of parental interaction with children about TV content, and in addition, the importance of interaction by teachers and other educators. In short, it is the child's social environment that provides the most im- mediate means by which the impact of anti—social television programming may be mediated. The Social Environment The findings of the above research demonstrate how verbal comments by parents and adults in the television viewing situation can reduce post- viewing aggression and increase learning of programming content. In this study, two aspects of the social environment will be examined: comments made by a female adult and mothers, and the presence of peers. Parents are the most important socialization agents for young children. As the child grows older the peer group challenges the family as a source of socialization. The following discussion is divided into two sections, 41 family type and its influence on television viewing, and peers' influence on TV viewing. Family Effects A discussion of family interaction patterns is presented here as a rationale for using verbal comments as a means of mediating television con- tent. It is important to remind the reader that this study will not col- lect data on family type and the purpose of presenting this information is to show how verbal comments by parents have the potential of affecting children's perception of television programming. Television is an important influence in the socialization process (Himmelweit, et a1., 1958 and Schramm, et a1., 1961). Children under five spend approximately 23.5 hours per week viewing television and in a seven- teen year period the typical teenager has watched 15,000 hours of TV. (Newsweek 2-21-77, p. 63). It is also thought that the mass media and particularly television must pass through a variety of sociological filtenh the family being one such filter. What, then, is the effect of the family on children's television viewing? The manner in which parents facilitate the development of moral be— havior in their children should be particularly relevant to the study of children's ability to evaluate anti-social behavior in television program— ming. Parents who have enabled their children to distinguish "right" from "wrong" or "good" from "evil" have developed in their children the ability to evaluate television content. Moral development has been a major concern on the part of the child developmentalists and psychologists. Parental practices are very import- ant in the child's process of internalizing moral values. This internal- ized moral value has often been referred to as "conscience" (Sears, et a1., 42 1957). The process of internalization has been defined by many researchems (Sears et a1., 1965, p. 199; Aronfreed, 1969, p. 264—265; Hoffman, 1970, p. 262; Staub, 1975, p. 118; Hoffman, 1975, p. 234). At least three com- mon elements exist in all definitions of the interanlization process: (1) the requirement of the adoption of social norms; (2) social norms pro- vide internal direction to the behavior of the child; and (3) internal direction is achieved by self-generated consequences for the social behavior of the child. The most important aspect of the concept of internalization is pre- cisely the notion of "inner-directedness." The child who has been raised to be independent of outer-directed stimulation for the control of his/ her behavior, should also be more resistant to the imitation or modeling of behaviors that run counter to his/her values. Hoffman (1970) mentions that one aspect of internalization is precisely "the amount of resistance offered to pressures and temptations to behave counter to the standard (p. 286-287)." In short, the child who is inner-directed should discount or criti- cize models that are inconsistent with his/her internalized normative standards. In contrast to the child with a high level of moral internalization, Hoffman (1975) describes the child with a low level of moral internali- zation as "behaving properly through fear of external sanctions and evaluating moral action on the bases of the likelihood of external pun— ishment (p. 232)." This type of child is sensitive to external initia- tion and performance. In brief, the child with a low level of moral internalization tends to be outer-directed. The child's behavior is shaped by forces external to the child. 43 Hoffman (1975) states that parental practices and interactions with children about social behavior are the most likely sources of influence to greatly affect moral development. Hoffman goes on to say that the importance of parental practices is emphasized because parents are the initial source of influence on their children. Sears, et al. (1957); Bandura and Walters (1959); Allensmith (1960); Aronfreed (1961); Burton, Maccoby and Allensmith (1961); and Hoffman and Saltzstein (1967) have shown that a relationship exists between parental disciplinary practices and the internalized control of their children's social behavior. Aronfreed (1976) and Hoffman (1975) identify two types of discipli- nary interactions between parents and their children, induction and sen- sitization. Aronfreed describes these two types of disciplines as follows: "Induction types of discipline have in common . . . that they tend to make the child's control of its behavior independent of external contingencies. In contrast, disciplinary habits of direct physical and verbal attack may be characterized as sensitization because they tend merely to sensitize the child to the anticipation of punishment." (1969, p. 309-310) Hoffman also mentions that induction techniques include the use of reasoning and explanation. Induction techniques appeal to the child's pride and achievement. Techniques used in sensitization include physical punishment, depri- vation of material objects or privileges, the direct applications of force, or threat of any of these (Hoffman, 1970, p. 285). The basic difference between the inductive and sensitization modes of parental discipline is that the former mode is communication oriented while the latter mode is based on the exercise of actual and implied power. Ac- cording to Aronfreed (1969) induction, being based on communication provides 44 the cognitive structure with which the child will be able to categorize his/her social experiences. "Cognitive structure facilitates internal- ized suppression by serving as an intrinsic mediator of anxiety which can intercede before the commission of a punished act (Aronfreed, 1969, -p. 276)." Sensitization, by requiring constant surveillance, and by being united to the disciplinary event, reduces the length of time during which the child experiences anxiety for the wrongdoing. The child that has internalized moral standards can be expected to be in a better position to reject televised anti-social behaviors compared to those with a low degree of moral development. Here again, the differ- entiation between "right" and "wrong" is based on either the internal conviction of the child or on external motivations for desired behavior. MbLeod, Chaffee, and Eswara (1966) stressed the importance of family communication patterns and typologies in communication research. These authors were interested in studying "the possibilities of changing per- sonality, through the change in the person's interaction patterns (p. 4)." These researchers "felt that families were likely to differ in the degree to which they emphasize harmony in interpersonal relations and impose restrictions on the expression of anger (p. 5)." According to McLeod and Chaffee (1972) there are two uncorrelated dimensions of communication structure in families. These dimensions are concept-orientation and socio-orientation. In concept oriented families, "the child is stimulated to express his ideas and to challenge others' beliefs. He is frequently exposed to both sides of an issue, and takes part in controversial discussions (McLeod, Chaffee and Wackman, 1967, p. 2)." 45 In socio—oriented families, "the child is encouraged to maintain harmonious personal relations with his parents and others. He may be advised to give in on arguments, avoid controversy, repress anger, and generally keep away from trouble (McLeod, Chaffee and Wackman, 1967, p. 1)." These two orientations interact to form a fourfold typology of fame ily communication patterns. To form these typologies each orientation is dichotomized into "high" and "low." The typologies are as follows: Laissez—faire families emphasize neither type of orientation. Children are not prohibited from challenging parental views, however, they are not exposed to independent and contending ideas. Protective families stress socio-oriented relations. These families emphasize harmonious and hierarchical socio-relations but do not stress exposure to controversy or expression of ideas. Pluralistic families stress concept orientation and not harmon- ious, hierarchical social relations. The child is exposed to controversial material and is free to develop new ideas. The child is able to make decisions without endangering social relations with his family. Consensual families emphasize both harmonious, hierarchical relations and controversy or expression of ideas. The child is encouraged to learn his parents' ideas and adopt their values. Stone and Chaffee (1970) say that "students from concept-oriented homes have been found less persuasible and more sensitive to the amount of information substantiating arguments in a persuasive message. Persons from a socio-oriented background have said they would be more likely to abandon play if a neighbor friend were to say he thought it was a poor idea (p. 240)." Referring to the typologies above, children of plural- istic families would be more independent in their judgment of social situations. The child in this family pattern would also be expected to 46 be more resistant to persuasive messages. Stone and Chaffee (1970), found that youngsters from highly socio-oriented homes were more persu- aded than others to change their opinion with respect to a social truism when attacked by a highly credible source. McLeod and Chaffee (1972) describe the socio—oriented person as one who asks "Who is involved? Do I like them? Do they like me? The concept— oriented person asks: What is involved? Is it a good idea? How does it compare to what I know? (p. 87)." McLeod, Atkin and Chaffee (1972), utilizing the above family typolo— gies, have found suggestive evidence of the role of family interaction patterns in mediating the relationship between televised anti-social be- havior and aggressive behavior of children. Protective families have children who are the heaviest television viewers, and are quite high in violence viewing. Children from pluralistic families have been found to be very low consumers of television content, and specifically violent television programming. Protective families tend to have highly aggres- sive children, while pluralistic families raise passive children. The consensual and laissez-faire types have been found to be somewhere in between. These findings suggest the possible indirect role of parental practices in reducing the effect of televised violence on children's aggressive behavior. Atkin (1972) reported that: "The relationship between violence viewing and aggressive be- havior in homes where the parent tried to teach the child not to act aggressively was compared to homes where a more laissez- faire attitude was implemented. The relationship between vio- lence viewing and aggressive behavior was much stronger in the half of the . . . samples where no emphasis was placed on nonviolent behavior . . . while only a slight positive rela— tionship was found where the parents did emphasize nonvio- lence (p. 2)." 47 In review, two theoretical approaches have been discussed with regard to parent-child interaction patterns and their effects on social behavior. Some implications of each approach have been indicated for the relationship between the viewing of anti-social television program— ming and aggressive behavior on the part of the child. The parent-child interaction styles of both research trends are congruent. Concept orientation is associated with reasoning, explana— tion and an informational orientation in general. Power assertion or sensitization is associated with the social power orientation where external pressures towards conformity are the rules. The above discussion provides the rationale for using verbal comments as a means of mediating television programming. The research cited above demonstrates the effectiveness of parent-child verbal communication in mediating a child's social behavior. The discussion on verbal interaction affecting moral judgment and behavior provide support for the dependent measure of perceived inappropriateness of anti—social behavior. This section on "family effects" provides the overall rationale of why paren- tal comments should be effective in mediating television content. The following section will examine the effects peers have on children. Peer Effects The distinction must be made between "peer" and "peer group." The "peer group" as a collective of "peers" follows the notion that the whole is greater than the sum of the individual parts. It is the "group's" values, goals and so forth that are more important than the individuals who comprise the group. Peer group formation begins to develop in the early teens when individuals seek the security of homogeneous collections 48 (Blitsten, 1971). Children used in this study are six to eight years of age and what is commonly thought of as "peer group influence" is not appropriate for these subjects. Young children are capable of developing relationships with peers. Peer in this context is defined as those children in the same age range or developmental stage (Piaget, 1955). Relationships seem to involve an actual exchange of energy, or an investment of energy into the person or object or activity to which one becomes related. The energy that is involved is transformed into a felt need to act with respect to, or on behalf of, the object to which one becomes related. Relationships are described as "ties" or "bonds." Blitsten says "their meanings are vari- ous and change with respect to the same object, but in all cases someone or something or some activity has significance that is distinct from the significance of one's own organism or one's own personal attributes (p. 59)." Key elements of relationship formation are communication and inter— action patterns. Blitsten (1971) suggests that the nature of verbal com- munication has an important effect on the development of relationships. Through communication children are capable of manipulating the environ- ment in order to provide for felt needs. Children then develop "bonds," "ties" or relationships with individuals that provide for their needs and security. A few studies have examined the effect of peer models and com- ments by peers on children's behavior. Some surprising results have been reported, but Piaget cautions that even when children talk to one another and play together they do not listen and do not expect consistent an— swers and responses. In the context of television, children tend to model peer behavior more than responding to verbal comments by peers. 49 Walters and Parke (1964) found that kindergarten-age boys who were exposed to a film in which a peer was punished for playing with certain toys did not play with similar toys when the experimenter placed prohibi— tions on playing. When the experimenter lifted his prohibitions on the toys, children who had seen the model punished played with toys as readily as other children. Hicks (1965) found that children who observed a male peer performing aggressive behavior on film demonstrated more aggressive imitative beha- viors immediately after viewing compared to children exposed to female peer and adult models. Over time, however, the adult male model behaviors were imitated more often than male peer's behavior. This is particularly important due to the large number of male (adult) models appearing in aggressive roles in television programming. Baron (1972) conducted research based on the hypothesis that "a peer's censuring remarks would be highly effective in counteracting the influence of the model under conditions where observers' level of attrac- tion toward the peer was relatively high, but would be generally low (p. 266)." Baron found that censure of the model by a disliked peer was more effective in reducing the amount of aggression than censure by a liked peer. Baron concluded that this result was attributed to: (l) "The failure of the censuring peer to demonstrate in his own behavior the type of restraint he urged, and (2) the tendency on the part of subjects to assume that this individual (the censuring peer) closely reciprocated their level of liking (p. 266)." Wolf and Cheyne (1972) found that live and televised models were most effective in facilitating conforming and disinhibiting behaviors in second- and third-grade boys, compared to a verbal mode of communication. 50 The models were same sex peers. The effects of the deviant peer model were more stable over time than the effects of the conforming peer model. Combined with Hick's study, deviating peers have a greater over-time effect than conforming peers, and deviating adults have a greater over- time effect on modeled deviant behavior than deviating peers. Wolf (1973) found that early grade school boys who viewed televised portrayals of peers deviating demonstrated more deviant behavior than those who viewed models displaying no deviant behavior. Behavior por- trayed by a peer may elicit the subject's attention and thereby facili- tate the coding of the observed information for future use. The above studies indicate that children are susceptible to peer models appearing in television programming. Baron's findings are con— sistent with Piaget's prediction of inconsistent behavior of children. Based on the information about parent-child interaction patterns affecting children's ability to accept or question sources of infor- mation, how will peers in a viewing situation affect one another in terms of perception of programming content? The answer to this question and questions about verbal comments affecting learning and aggression will be addressed in this study. CHAPTER III METHODOLOGY This experiment was designed with the assumption that mothers can mediate the effect of anti-social programming by providing comments about TV programming content and by simply being present in the viewing situa- tion. It was also anticipated that non-parent adults would mediate the effect of anti-social programming on children in the same manner as parents, but parents would be more effective. In addition, it was anti- cipated that viewing anti-social programming in a peer situation without an adult or parent present may influence children's perception of the anti-social programming. This experiment examines the effects of viewing televised anti-social behavior with peers and with a female adult/mother present with and with— out providing comments, on children's: (l) ability to recall content, (2) ability to recall consequences of anti-social behavior, (3) perceived inappropriateness of anti-social behavior, and (4) post-viewing cooperation. This chapter will discuss the stimulus for the experiment, the design with independent, dependent and control variables operationalized, sampl- ing, and statistical analysis and hypothesis formation. Anti-Social Stimulus The stimulus for this experiment consisted of two videotaped segments of "Dirty Harry" and "Magnum Force." Both movies were aired on the CBS and ABC television networks during the 1976-77 season. Clint Eastwood, the star in both movies, appears in both segments. The total length of 51 52 the stimulus is ten minutes. The two segments are as follows: (1) Bus hijacking - a man takes over a school bus and assaults the driver. In an effort to escape Eastwood's pursuit, he wrecks the bus. A chase scene follows involving shooting. The hijacker takes a child hostage and East- wood eventually shoots the hijacker; and (2) Gangster shoot-out - police attempt to serve warrants for the arrest of one of the gangsters. The gangsters, thinking police are really other gangsters in disguise, start shooting the police. This scene demonstrates the consequences of shooting, e.g., bullets go through wood, glass, car metal, in addition to hurting people. (A script of the videotaped segments is in Appendix A.) O The Desigg This study employs an experimental post-test only control group design utilizing the following one-by-six fixed effects model of analysis of variance. Parent Parent . Adult Adult Peers Viewing Comment No Comment Comment No Comment Viewing Alone n - 22 n - 22 n - 22 n - 22 n - 22 n - 22 N - 132 Independent variable The independent variable for this study is the viewing situation in which the child is placed. There are six viewing conditions: 1. Mother providing censorious and descriptive comments about anti-social behavior while viewing with her child. 53 2. Mother present, but giving no comment about anti~social content. 3. A female adult providing censorious and descriptive comments about anti-social behavior while viewing with a child. 4. A female adult present, but giving no comment about anti-social content. 5. The child viewing with a peer, no female adult or mothers present, and no comments given about the content. 6. The child viewing alone with no comments given about the contents (Control group). Comments were given before, during and after the viewing of anti- social behavior. The exact comments were selected from comments made by mothers and fathers who viewed the videotaped violence and are contained in Appendix B. In summary, prior to the beginning of the videotape the mother/female adult provided comments about how wrong it is to shoot people and that shooting causes much damage. During both segments the mother/female adult was making comments about the content of the segments. The mother/ female adult was pointing out things that were happening on the screen to the child. Between the segments the mother/female adult was again providing comments about the inappropriateness of shooting. Finally, at the conclusion of the viewing situation the mother/female adult told the child that it was wrong to shoot people. The two segments were separated by eight seconds which provided sufficient time for comments to be made. The same time delay between segments existed for all conditions. Comments were made available to the mother upon arrival. The mothers were allowed to study the sheet of paper which contained the comments 54 for five minutes before participating in the study. During the viewing situation they were also allowed to hold the sheet of paper with the comments. They were encouraged to restrict their comments to those that appeared on the paper and also the wording of the comments was to be the same as on the paper. Operationalization of Dependent Variables After viewing the anti-social behavior stimulus all subjects were personally interviewed. The subjects were measured on the following dependent variables: (1) ability to recall pictorial and verbal content of both segments, (2) ability to recall consequences of anti-social behavior in both segments, (3) perceived inappropriateness of anti-social behavior, and (4) post-viewing cooperation. Recall of Content Specific questions were asked about the pictorial and verbal content in both segments viewed. Pictorial content refers to the content of the segments the child was able to see. The verbal content refers to that which may be recalled by hearing the videotape segments. In formulating the instrument, researchers independently viewed the videotape and dev- eloped questions pertaining to its content. The final instrument was developed from questions formulated by these individuals and a copy of the instrument is in Appendix C. Photographs were taken of specific actors and scenes appearing in both segments to serve as references for questions asked. Recall of Consequences of Anti—Social Behavior Questions were asked relating to consequences of specific shooting and wreckless driving incidences occurring in both segments (See 55 Appendix C). Again, formulation of the instrument was based on indep- endent viewing and questions were developed by researchers and other individuals. Consequences of anti-social behavior refers to the effects caused by shooting or driving. The use of observations or judgments made by researchers to develop questions for the dependent measures of recall of pictorial and verbal content and the consequences of anti—social behavior satisfy the neces- sary requirements for content validation. Inappropriateness of Anti-Social Behavior This dependent variable is a measure of the child's percgption of the "rightness" or "wrongness" of specific incidences of anti-social behavior occurring in the segments (See Appendix C). "Right" refers to behavior that is perceived correct, justified or appropriate, while "wrong" refers to behavior that is incorrect, unjustified or inapprop- riate. It is assumed by the experimenter that shooting is wrong unless absolutely necessary. The anti-social incidences that occurred in the segments were explained to a policeman and a criminal investigator with a district attorney's office to determine if the anti-social incidences were "right" or "wrong." It was against this standard that children's responses were compared. Post-Viewing Cooperation The final dependent variable involves a measure of post-viewing cooperation (See Appendix C). All children participated in this post- viewing cooperation measure. The child was seated at a response box modeled after the one employed by Liebert and Baron (1972). The box displayed two identical buttons. The words "NO HELP" appeared above 56 the button on the left, while the work "HELP" appeared above the button on the right. Each button was connected to a clock which measured, in seconds, the amount of time each button was depressed. A research assistant sat behind the response box and counted, with the use of a hand counter, the number of times each button was depressed. The assis- tant explained the gaming situation to the subjects. The explanation was as follows: Today we are going to play a game. Now I want you to think . . . if somebody needed your help would you help them or would you not help them. (POINTING TO "HELP" AND "NO HELP"). This button is for if you would HELP them and this is for if you would not help them. You can push either button, it doesn't have to be the same one each time, as many times as you like for as long as you want. When I say go, I want you to answer the qeustion I will ask you by pushing one of these buttons. Now remember you can press either button (POINTING TO THE BUTTONS) as many times as you want for as long as you want. The questions are: if someone needed your help, would you help them or not help them. When I say "again" I want you answer the question again. Are you ready? If someone needed your help would you help them or not help them? Go! (10 seconds) Again! (10 seconds) Again! (10 seconds) Again! (10 seconds) Again! (10 seconds). Thank you. Now if you would please follow me (Take child to interviewer). This lady would like to ask you some questions. This game allowed for a measure of the frequency and duration of "HELP" and "NO HELP." Control Variables This experiment controlled for the sex of the subjects. Sex has been a traditional control variable in studies on childhood development. Developmental research indicates that girls develop a moral conscience at an earlier age than boys (Kohlberg, 1964). In addition, research has also indicated that boys typically engage in more aggressive play be- havior than girls. Boys also behave more aggressively after viewing 57 anti-social programming than girls (Harris and Siebel, 1975). The sex of both the participating parent and the adult co-viewer was female. In the early years of development mothers spend more time with children than do fathers. The subjects in this experiment were six, seven and eight years of age. These age groups were first and second graders. The research literature has indicated that children in this age group perceive tele- vision content to be real more than any other age group (Hirsch, 1969). The Viewing Situation and Order of Events The viewing situation is diagrammed as follows: l " Videotape recorder l____ l TV monitor [2] [1] Chairs (1) Mother/female adult (2) Child The mother/female adult was seated to the right of the child. It is assumed that this seating arrangement would maximize the likelihood of child-mother and child-female adult conversation. All comments were monitored by the experimenter to ensure comments made by mothers and a female adult were those developed for this study. All mothers and female adults stated the prescribed comments exactly. The same seating arrangement existed for the mother/female adult "no comment" conditions as well as the "peer viewing" condition. In the "viewing alone" condition the mother/female adult chair was removed. 58 Order of Events The order of events is described for the most complex situation, the "mother comment" condition. The "mother comment" condition began when the mother arrived at the research rooms. The mother's name was checked off a master list of those who were participating in the study and a check was made to ensure that the child listed was the correct child. The mother was then presented with the sheets of paper that contained the comments and was told she was to make these comments while viewing the videotape with her child. While the mother was reading the comments for approximately five minutes, the experimenter went to the classroom to bring the child to the experiment. The mother and child were free to greet each other for approximately thirty seconds to a minute. This time was allowed so that the child would feel more at ease prior to the viewing situation. The experimenter then escorted the mother and child into the viewing situation. The mother and child were seated in the order described above. After the mother and child were seated and before the beginning of the videotape the experimenter told the child the following: In a minute you will see two short stories on the TV before you. The first story is about a man who tries to take a bus and another man comes and chases him. There is a lot of shoot- ing that takes place. Then the TV will go black for a few seconds before the second story begins. The second story is about some men in an office and the police come and there is a lot of shooting that takes place between the men in the office and the police. The experimenter then started the videotape. At the conclusion of the videotape the experimenter allowed a few seconds to pass before entering the room. This allowed time for the mother/female adult to provide the final comments. After the videotape was turned off the child was asked if he/she was scared or frightened by 59 what was just seen. Regardless of the child's response the following comments were made: What you have just seen on TV is not real. Just as you have teachers who have the job of teaching you how to read, spell, and add and subtract the people you have seen on TV are paid to make us believe that what they do is real. That is their job. They are called actors. Are you still afraid? (None of the children said they were afraid at this point.) Now if you would come with me there is a lady who wants to play a game with you and another lady who wants to ask you some questions. I'll talk with your mom in just a minute. Let's go. The child was then escorted to the gaming situation. On returning to the "check—in" table the second mother was processed like the first, only this time the mother was told that she would be participating in the "no comment" condition. The second mother was told ESE to provide comments while viewing the videotape with her child. The first mother and the experimenter left the room and at that time the experimenter .explained the experiment to the mother. The mother was told that the results would be made available to the school if she wanted a copy of them. At this time the first mother left and the experimenter went to the classroom to get the child of the second mother. After the second videotape session began the experimenter returned the child who had participated in the first session back to the classroom. On the return to the classroom the child was asked if he/she rode a bus to school. If the child answered "yes" then a discussion followed about how safe school buses are, and that the incident about the bus hijacking was just pretend and not real. The "adult," "peer," and "alone” conditions followed in the same manner except the five minute study time was not necessary as the mother was not participating in these conditions. In the "adult" con- ditions the child was introduced to the female adult prior to comments 60 made by the experimenter. The order of events stayed the same for all treatment conditions. Sampling Each treatment condition in this experiment contained twenty-two subjects, eleven males and eleven females, for a total of one hundred and thirty-two children. Due to efficiency and economy, subjects were selected from schools in the Lansing, Michigan, metropolitan area. Participating schools were: Sycamore and Midway elementary schools in Holt, Michigan, and Resurrection School in Lansing, Michigan. Sycamore and Midway are public elementary schools with grades kindergarten through six. Resurrection is a private Catholic elementary school with grades kindergarten through eight. All first- and second-grade students in these schools carried home to their parents a letter explaining the study and a consent form to be signed by the parents (See Appendix D). The consent form also contained additional questions to determine if the mother or female guardian would be willing to participate in the study with the child, and a telephone number where the mother/guardian could be reached. The response to the first letter was approximately fifty percent of the projected sample size of ninety- six. From the returned consent forms a list of participating students, mother/guardians, phone numbers and teachers was compiled. Approximately four days later a second letter and accompanying consent form was sent home with the children who did not previously return the first consent form. This second letter netted a sample size of 132, 36 more than initially planned. Information as to the student's name, mother/ guardian, phone number and teacher's name was added to the list origin- ally compiled. 61 Next, a list of dates and times the research would be conducted at the various schools was developed. Participating mothers were then contacted by telephone and a date and time to come to the school was established. Mothers were scheduled at fifteen minute intervals, allow- ing a five minute preparation time for reading comments that some mothers provided, and ten minutes for viewing the videotape with the child. If the mother could not be contacted in two telephone calls or otherwise was unable to participate in the study, her child was placed in one of the other four treatment conditions. All mothers except one partici- pated in the study at their appointed time. The one mother who did not keep her appointed time was attending another child who had emergency surgery. Arrangements were made for four additional mothers to part- icipate with their children if they were needed. One situation did occur where a mother and child participated and it was assumed by the experi- menter that the child was a girl by the spelling of the name. The child was a boy and this made for twelve mother/male child subjects in the mother comment condition, and ten mother/female subjects in the same treatment condition. One of the mother/male subjects was randomly deleted, and an additional mother/female subject participated in the study. This allowed for equal sample sizes for males and females in the mother comment condition. Subjects were randomly assigned to treatment conditions. For the mother comment/no comment conditions at the beginning of each day a coin was tossed for the first mother/child participating. The mother called the toss of the coin. If the first viewing situation was a comment condition then the second viewing situation was a no comment condition, and so forth throughout the day. The remaining subjects 62 were randomly assigned to the "adult," "peer" and "alone" conditions. A table of random numbers was used in this selection process. For each subject a number between one and four was randomly selected. After the four groups were filled each group was randomly assigned a treatment condition. In the peer viewing condition twice as many students were selected. In the selection of peers the first two available students randomly assigned to this condition were used. After the children viewed the videotape a coin was tossed by the experimenter and the correctly selected toss was included in the study while the other child was debriefed and sent back to the classroom. Pretest Six subjects were used in the pretest conducted prior to data collection. As a result of the pretest the instructions for the game to measure cooperation were changed from Liebert and Baron's description to the present instructions described earlier. Too much information was contained in Liebert and Baron's instructions and the children did not fully understand the game. The instrument was changed to include incidences that occurred in the segments that were not so obvious, for example the appearance of a factory worker, and the fact that the bus ran into a pile of sand. In addition, only questions pertaining to what occurred in the segments were included in the final instrument. The pretest was beneficial for the research assistants, game instructor and interviewer, in terms of organizing the order of the experiment which allowed for efficient use of time. The pretest allowed the female adult viewer to become familiar with comments and the videotape stimulus. 63 Statistical Analyses and Hypothesis Formapion The data were analyzed utilizing a oneeway fixed effects model of analysis of variance. This technique is appropriate in determining if differences exist among treatment conditions. Analysis of variance alone will not tell which treatment conditions contribute to the sign— ificance of the overall F test. Comparisons between specific treatment conditions are applicable only to the situation where a preliminary analysis of variance and F test has shown overall significance. There is no requirement that post hoc comparisons be independent. In this experiment treatment conditions will be of equal size and, therefore, the Tukey post hoc technique was used when pair-wise differences between means were examined. In situations where averaged comparisons were made between groups of treatment conditions the Scheffe post hoc method was used. The Tukey and Scheffe post hoc techniques allow for comparisons among means to determine which differences between group means contri— bute to the significance of the overall F test obtained by a one-way analysis of variance. The Tukey post hoc technique is more powerful than the Scheffe technique when the number of subjects in each treatment con- dition are of equal size and when the researcher is interested only in pair-wise comparisons of means. The Scheffe post hoc technique has the advantage of simplicity and is more powerful when complex comparisons among means are made. The Scheffe technique is also known to be relatively insensitive to departures from normality and homogeneity of variance. The assumptions that are made for one-way fixed effects model of analysis of variance are as follows: (1) subjects are randomly drawn from a normally distributed population, (2) the population variances are 64 equal, (3) the observations between and within groups are independent, (4) the relationship between the variables is linear, and (5) the level of measurement is interval (Hays, 1974). In light of the dependent variables discussed above, the following questions will be asked: 1. Does the mean of the "mother comment" condition differ from the mean of the "mother no comment" condition? 2. Does the mean of the "female adult comment" condition differ from the mean of the "female adult no comment" condition? 3. Does the average of the means in the "female adult and mother comment" conditions differ from the average of the means in the "no comment" conditions? 4. Does the average of the means in the "female adult and mother comment" conditions differ from the mean of the "peer viewing" condition? 5. Does the average of the means in the "female adult and mother comment" conditions differ from the mean of the "viewing alone" condition? Statement of Testable Hypotheses Null Hypothesis 1: There is no difference in the mean scores between the "mother comment" and "mother no comment" conditions on subjects' ability to recall pictorial content. Symbolically: Ho : n1 = “2 Where: Mean score of subjects in "mother comment" condition. Mean score of subjects in "mother no comment" condition. Alternative Hypothesis 1: The mean scores of recall of pictorial content will be higher in the "mother comment" condition than in the "mother no comment" condition. Symbolically: Hl :p1> p2 Where: Legend is the same as for Null Hypothesis 1. 65 Null Hypothesis 1A: There is no difference in the mean scores between the "mother comment" and "mother no comment" conditions on subjects' ability to recall consequences of anti—social behavior. Symbolically: Ho : p1 = ”2 Where: Legend is the same as for Null Hypothesis 1 Alternative Hypothesis 1A: The mean scores of recall of consequences of anti-social behavior will be higher in the "mother comment" condition than in "mother no comment" condition. Symbolically: H1 : u1>p2 Where: Legend is the same as for Null Hypothesis 1. Null Hypothesis 1B: There is no difference in the mean scores between the "mother comment" and "mother no comment" conditions on subjects' ability to recall verbal content. Symbolically: Ho : p1 = ”2 Where: Legend is the same as for Null Hypothesis 1. Alternative Hypothesis 1B: The mean scores of recall of verbal content will be higher in the "mother comment" condition than in the "mother no comment" condition. Symbolically: H1 : pl>u2 Where: Legend is the same as for Null Hypothesis 1. 66 Null Hypothesis 1C: There is no difference in the mean scores between the "mother comment" and "mother no comment" conditions on subjects' perceived inappropriateness of anti-social behavior. Symbolically: Ho : "l = uz Where: Legend is the same as for Null Hypothesis 1. Alternative Hypothesis 1C: The mean scores of perceived inappropriate- ness of anti-social behavior will be higher in the "mother comment" condition than in the "mother no comment" condition. Symbolically: Hl : pl) uz Where: Legend is the same as for Null Hypothesis 1. Null Hypothesis 1D: There is no difference in the mean scores between the "mother comment" and "mother no comment" conditions on subjects' post-viewing cooperation. Symbolically: Ho : "l = "2 Where Legend is the same as for Null Hypothesis 1. Alternative Hypothesis 1D: The mean scores of post-viewing cooperation will be higher in the "mother comment" condition than in the "mother no comment" condition. Symbolically: H1 : pl) "2 Where: Legend is the same for Null Hypothesis 1. 67 Null Hypothesis 2: There is no difference in the mean scores between the "female adult comment" and "female adult no comment" conditions on subjects' ability to recall content. Symbolically: H0 : u3 = u4 Where: Mean score of subjects in "female adult comment" condition. Mean score of subjects in "female adult no comment" condition- Alternative Hypothesis 2: The mean scores of recall of pictorial content will be higher in the "female adult comment" condition than in the "female adult no comment" condition. Symbolically: Hl : p3)”4 Where: Legend is the same as for Null Hypothesis 2. Null Hypothesis 2A: There is no difference in the mean scores between the "female adult comment" and "female adult no comment" conditions on subjects' ability to recall consequences of anti-social behavior. Symbolically: H : u3 = u4 Where: Legend is the same as for Null Hypothesis 2. Alternative Hypothesis 2A: The mean score of recall of consequences of anti-social behavior will be higher in the "female adult comment" condition than in the "female adult no comment" condition. Symbolically: Hl : p3>n4 Where: Legend is the same as for Null Hypothesis 2. 68 Null Hypothesis 2B: There is no difference in the mean scores between "female adult comment" and "female adult no comment" conditions on subjects' ability to recall verbal content. Symbolically: HO : u3 = u4 Where: Legend is the same as for Null Hypothesis 2. Alternative Hypothesis 2B: The mean scores of recall of verbal content will be higher in the "female adult comment" condition than in the "female adult no comment" condition. Symbolically: Hl : p3) "4 Where: Legend is the same as for Null Hypothesis 2. Null Hypothesis 2C: There is no difference in the mean scores between the "female adult comment" and "female adult no comment" conditions on subjects' perceived inappropriateness of anti-social behavior. Symbolically: Ho : u3 = p4 Where: Legend is the same as for Null Hypothesis 2. Alternative Hypothesis 2C: The mean scores of perceived inappropriateness of anti-social behavior will be higher in the "female adult comment" condition than in the "female adult no comment" condition. Symbolically: Ho : u3> "4 Where: Legend is the same as for Null Hypothesis 2. 69 Null Hypothesis 2D: There is no difference in the mean score between the "female adult comment" and "female adult no comment" conditions on subjects' post—viewing cooperation. Symbolically: Ho : u3 = n4 Where: Legend is the same as for Null Hypothesis 2. Alternative Hypothesis 2D: The mean score of post-viewing cooperation will be higher in the "female adult comment" condition than in the "female adult no comment" condition. Symbolically: Hl : u3) u4 Where: Legend is the same as for Null Hypothesis 2. Null Hypothesis 3: There is no difference in the mean scores between the mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' ability to recall pictorial content. Symbolically: Ho : "1 + u3 = p2 + p4 2 2 Where: Mean score of subjects in "mother comment" condition. Mean score of subjects in "mother no comment" condition. 1 2 3 = Mean score of subjects in "female adult comment" condition. 4 Mean score of subjects in "female adult no comment" condition. Alternative Hypothesis 3: The mean scores of recall of pictorial content will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions: 70 Symbolically: Hl : M1 +-p.3 > uz +~p4 2 2 Where: Legend is the same as for Null Hypothesis 3. Null Hypothesis 3A: There is no difference in the mean scores between the mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' ability to recall consequences of anti— social behavior. Symbolically: HO : u +u3=u2+u 2 2 1 4 Where: Legend is the same as for Null Hypothesis 3. Alternative Hypothesis 3A: The mean scores of recall of consequences of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" condition. Symbolically: Hl : p1 +‘u3 > p2 +‘p4 2 2 Where: Legend is the same as for Null Hypothesis 3. Null Hypothesis 3B: There is no difference in the mean scores between the mother and female adult "comment" and mother and female adult "no comment" conditions on subjectS' ability to recall verbal content. Symbolically: Ho : p1 +'n3 = ”2 +‘u4 7—7—2.— Where: Legend is the same as for Null Hypothesis 3. 71 Alternative Hypothesis 3B: The mean score of recall of verbal content will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. Symbolically: H1 : p1 + p3).p2 + p4 2 2 Where: Legend is the same as for Null Hypothesis 3. Null Hypothesis 3C: There is no difference in the mean scores between the mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' perceived inappropriateness of anti- social behavior. Symbolically: Ho : "l +-p3 = p2 + p4 2 2 Where: Legend is the same as for Null Hypothesis 3. Alternative Hypothesis 3C: The mean scores of perceived inappropriateness of anti-social behavior will be higher in mother and female adult "comment" conditions than in the mothern and female adult "no comment" conditions. Symbolically: H1 : '11 + 113) 112 + pa 2 2 Where: Legend is the same as for Null Hypothesis 3. Null Hypothesis 3D: There is no difference in the mean scores between the mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' post-viewing cooperation. 72 Symbolically: Ho : pl +'p3 =7u2 + p4 2 2 Where: Legend is the same as for Null Hypothesis 3. Alternative Hypothesis 3D: The mean score of post—viewing cooperation will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. Symbolically: Hl : ”l + u3 ”2 + n4 ________.> ________ 2 2 Where: Legend is the same as for Null Hypothesis 3. Null Hypothesis 4: There is no difference in the mean scores between the mother and female adult "comment" conditions and "peer viewing" condition on subjects' ability to recall pictorial content. Symbolically: p1 +'u3 =4u5 2 Where: 1 = Mean score of subjects in "mother comment" condition. 2 = Mean score of subjects in "female adult comment" condition. 5 = Mean score of subjects in "peer viewing" condition. Alternative Hypothesis 4: The mean scores of recall of content will be higher in the mother and female adult "comment" conditionstfimuiin the "peer viewing" condition. Symbolically: H1 : pl +‘p3 > ps 2 Where: Legend is the same as for Null Hypothesis 4. 73 Null Hypothesis 4A: There is no difference in the mean scores between the mother and female adult "comment" conditions and "peer viewing" condition on subjects' ability to recall consequences of anti-social behavior. Symbolically: Ho : ”I +‘u3 = "5 2 Where: Legend is the same as for Null Hypothesis 4. Alternative Hypothesis 4A: The mean scores of recall of consequences of anti—social behavior will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. Symbolically: Hl : p1 + p3 2 >115 Where: Legend is the same as for Null Hypothesis 4. Null Hypothesis 4B: There is no difference in the mean scores between the mother and female adult "comment" and the "peer viewing" condition on subjects' ability to recall verbal content. Symbolically: Ho : p1 +‘u3 = p 2 5 Where: Legend is the same for Null Hypothesis 4. Alternative Hypothesis 4B: The mean scores of recall of verbal content will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. Symbolically: H1 : "l +‘u3 > ”5 2 74 Where: Legend is the same as for Null Hypothesis 4. Null Hypothesis 4C: There is no difference in the mean scores between the mother and female adult "comment" conditions and the "peer viewing" condition on subjects' perceived inappropriateness of anti-social behavior. Symbolically: HO :‘ul +,u3 = us 2 Where: Legend is the same as for Null Hypothesis 4. Alternative Hypothesis 4C: The mean scores of perceived inappropriateness of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. Symbolically: H : pi +.n 2 1 3>”5 Where: Legend is the same as for Null Hypothesis 4. Null Hypothesis 4D: There is no difference in the mean scores between the mother and female adult "comment" conditions and the "peer viewing" condition on subjects' post-viewing cooperation. Symbolically: HO : nl'+‘”3 =,u5 2 Where: Legend is the same as for Null Hypothesis 4. 75 Alternative Hypothesis 4D: The mean scores of post-viewing cooperation will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. Symbolically: Ho : p1 + n3 > L15 2 Where: Legend is the same as for Null Hypothesis 4 Null Hypothesis 5: There is no difference in the mean scores between the mother and female adult "comment" conditions and the "viewing alone" condition on subjects' ability to recall pictorial content. Symbolically: ”l +‘u3 = ”6 2 Where: 1 = Mean score of subjects in "mother comment" condition. 2 = Mean score of subjects in "female adult comment'condition. 3 = Mean score of subjects in "viewing alone" condition. Alternative Hypothesis 5: The mean scores of recall of pictorial content is higher in the mother and female adult "comment" conditions will be higher than in the "viewing alone" condition. Symbolically: pl + p3 >p6 2 Where: Legend is the same as for Null Hypothesis 5. Null Hypothesis 5A: There is no difference in the mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' ability to recall consequences of anti—social behavior. 76 Symbolically: H0 :‘pl +‘p3 = u 2 6 Where: Legend is the same as for Null Hypothesis 5. Alternative Hypothesis 5A: The mean scores of recall of consequences of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. Symbolically: H1 : ”l +‘u3 >‘p6 2 Where: Legend is the same as for Null Hypothesis 5. Null Hypothesis 5B: There is no difference in the mean scores between the mother and female adult "comment" and the "viewing alone" condition on subjects' ability to recall verbal content. Symbolically: HO : ”1 +.n3 = #6 2 Where: Legend is the same as for Null Hypothesis 5. Alternative Hypothesis 5B: The mean scores of recall of verbal content will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. Symbolically: H1 : ”l + 113 >u6 2 Where: Legend is the same as for Null Hypothesis 4. 77 Null Hypothesis SC: There is no difference in the mean scores between the mother and female adult "comment" conditions and the "viewing alone" condition on subjects' perceived inappropriateness of anti-social behavior. Symbolically: HO : p1 + p3 = "6 2 Where: Legend is the same as for Null Hypothesis 5. Alternative Hypothesis 5C: The mean scores of perceived inappropriate- ness of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. Symbolically: Hl : p1 +-p3 >‘p6 2 Where: Legend is the same as for Null Hypothesis 5. Null Hypothesis SD: There is no difference in the mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' post—viewing cooperation. Symbolically: Ho : ”l + u3 = “6 2 Where: Legend is the same as for Null Hypothesis 5. Alternative Hypothesis 5D: The mean scores of post-viewing cooperation will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. Symbolically: H1 : ”l + n3 >116 2 78 Where: Legend is the same as for Null Hypothesis 5. Data Analysis The data were analyzed on a Control Data Corporation 6500 computer at the Michigan State University Computer Center. The integrated system of computer programs in the 6.5 version of the Statistical Package for the Social Sciences (SPSS, 1976) was used to analyze the data. An overall F test (one-way fixed effects model of analysis of vari- ance) was performed on the data for the five dependent variables. The Tukey post hoc technique was used for pair-wise comparisons in hypo- theses one and two. The Scheffe post hoc technique was used for the complex comparisons in hypotheses three, four and five. The dependent measures of recall of pictorial content, consequences, verbal content, and inappropriateness of anti-social behavior were dummy scored. Correct answers received a one while incorrect answers received a zero. Recall of content consisted of twenty-eight questions. If a child answered all the questions correctly he/she would receive a score of twenty-eight, and if all questions were answered incorrectly the child would receive a score of zero. The range for recall of conse- quences and verbal content is zero to twelve, and the range for inappro- priateness of anti-social behavior is zero to twenty. The scores of subjects in each treatment condition were summed and the total was divided by the number of subjects in each condition to obtain a mean for each dependent measure. The dependent measure of post-viewing cooperation consists of a measure of time and frequency. This measure was analyzed in several 79 ways. First, the amount of time the "HELP" and "NO HELP" buttons were depressed was summed for each treatment condition and analyses of vari- ance were performed. Next, the number of times in which the "HELP" and "NO HELP" buttons were depressed was summed for each treatment condition and analyses of variance were performed. Then the difference between the times the "HELP" and "NO HELP" buttons were depressed was obtained for each subject and summed within treatment conditions. An analysis of variance was then performed on this time difference measure. The same procedure was performed on the frequencies the "HELP" and "NO HELP" but- tons were depressed, and an analysis of variance was performed on this difference in frequencies. An average time of "HELP" and "NO HELP" was obtained by dividing the time each button was depressed by the frequency each button was depressed. Separate analyses of variance were performed on average time "HELP" and average time "NO HELP". Finally, the average time "NO HELP" was subtracted from average time "HELP" to determine whether or not each suchect demonstrated cooperating behavior. In summary, the measures of post-viewing cooperation were: (1) Time "HELP": total amount of time "HELP" button was depressed. (2) Frequency "HELP": total number of times "HELP" button was depressed. (3) Time "NO HELP": total amount of time "NO HELP" button was depressed. (4) Frequency "NO HELP": total number of times "NO HELP" button was depressed. (5) Difference in Time: time "HELP" minus time "NO HELP." (6) Difference in Frequency: frequency "HELP" minus frequency "NO HELP." (7) Average Time "HELP": time "HELP" divided by frequency "HELP." (8) Average Time "NO HELP": time "N0 HELP" divided by frequency "NO HELP." 80 (9) Difference in Average Time: average time "HELP" minus average time "NO HELP." CHAPTER IV FINDINGS The findings of this study are divided into the following sections: descriptions of the sample, description of the measures, results of ana- lysis of variance tests and test of hypotheses. The presentation of the data will examine each dependent variable for the total sample and for the control variable, sex. Description of Sample The data were collected at the following schools on these dates: Sycamore Elementary School -- May l6, l7 and 19; Resurrection School -- May 18, 20 and 26; and Midway Elementary School -- May 23, 24 and 25, 1977. The sample consisted of 28 (21%) six year olds, 67 (51%) seven year olds and 37 (28%) eight year olds for a total of 132 children. Sycamore contributed 46 (35%) students to the total sample, Midway contributed 53 (40%) and Resurrection contributed 33 (25%) students. Because the videotaped segments used in the stimulus for the experi- ment contained a school bus and children, subjects were asked if they rode a bus to and/or from school. Students who indicated they rode buses received additional comments by the experimenter when the child was re- turned to the classroom. The nature of these comments centered around the child being afraid to ride a school bus now that he had seen the vid- eotape. None of the children indicated they were afraid of riding a school bus. Of the total sample, 33 (25%) rode buses to and/or from school, while 81 82 99 (75%) did not. Of the total sample of 132 subjects, 66 (50%) were males and 66 (50%) were females. All 132 children are included in analyses for the total sample, while 66 males and 66 females are included in analyses for the control variable, sex. Description of Measures There are a total of five dependent measures: 1) recall of pictorial content; 2) recall of consequences of anti-social behavior; 3) recall of verbal content; 4) perceived inappropriateness of anti-social behavior; and 5) post-viewing cooperation. Post-viewing cooperation consists of the following measures: 1) the amount of time the "HELP" button is de- pressed; 2) the frequency with which the "HELP" button is depressed; 3) the amount of time the "NO HELP" button is depressed; 4) the frequency with which the "NO HELP" button is depressed; 5) the difference in time obtained by subtracting "NO HELP" time from "HELP" time; 6) the difference in frequency obtained by subtracting "NO HELP" frequency from "HELP" fre- quency; 7) an average time of "HELP" obtained by dividing "HELP" time by "HELP" frequency; 8) an average time of "NO HELP" obtained by dividing "NO HELP" time by "NO HELP" frequency; and finally 9) a difference in average time obtained by subtracting average "NO HELP" time from average "HELP" time. Means were derived for all dependent measures for each treatment condition and an analysis of variance was performed on the data. Where F tests were significant Tukey and Scheffe post hoc tech- niques were used to determine the extent to which treatment conditions contributed to the overall significance. The dependent measure of recall of pictorial content contains 28 questions. The subjects receive a "l" for a correct answer and a "0" for 83 an incorrect answer. The possible range of scores is O - 28. The actual range of scores for the total sample is 11 - 20. Controlling for sex, the range for males is 13 - 20, and the range for females is 11 - 20. The dependent measure of recall of consequences contains 12 questions. Responses are dummy coded so that a correct response receives a "1", while an incorrect response receives a "0". The possible range of scores is O - 12. The actual range of scores for the total sample is 3 - 12. When controlling for sex, the actual range of scores for males is 3 - 11 and for females the range is 3 - 12. The dependent measure of recall of verbal content contains 12 questions. This variable is also dummy coded. Correct answers receive "1" and incorrect answers receive a "0". The possible range of scores is O - 12. The actual range of scores for the total sample for recall of verbal content is 3 - 12. The actual range for males is 3 - 11 and for females the actual range is 3 - 12. The dependent measures of perceived inappropriateness of anti-social behavior contains 20 questions. This variable is also dummy coded. Cor- rect answers receive a "1" and incorrect answers receive a "O". The pos- sible range of scores is 0 - 20. The actual range for the total sample for perceived inappropriateness of anti-social behavior is 6 - 20. Con- trolling for sex, the range for males is 9 - 20, and for females the range is 6 - 20. The dependent measure of cooperation contains two separate measures, time and frequency. The time measure has a possible limit of O - 50 seconds, while the frequency has no limit. Time and frequency were tabulated for the "HELP" and "NO HELP" categories. The actual range of "HELP" time is O - 50 seconds and the actual range for frequency "HELP" 84 is O - 34. The actual range of "NO HELP" time is 0 - 50 seconds and the actual range of "NO HELP" frequency is 0 - 20. Table I presents a correlation matrix of all dependent measures. Re- call of pictorial content is significantly correlated with recall of con- sequences of anti-social behavior and recall of verbal content. Recall of verbal content is highly correlated with recall of consequences and frequency of "NO HELP". A possible explanation for the significant corre- lations among the dependent measures of recall is the fact that recall relates to the videotape stimulus. In the stimulus pictorial, verbal and consequences of anti-social behavior content relate to each other to present the message of the program. For example, the verbal content re- lates to what the child is seeing on the screen, which may be pictorially describing consequences of anti-social behavior. Other significant corre- lations are in the measures of cooperation. "HELP" time is correlated with "HELP" frequency, Difference in Time, and Difference in Average Time. An explanation for these correlations is the fact that "HELP" frequency is a good indicator of "HELP" time in the gaming situation. In addition, "HELP" time is part of the measures of time differences. "NO HELP" fre- quency is correlated with "NO HELP" time. An explanation for this is that the amount of "NO HELP" time is dependent on the depressing of the "NO HELP" button which is the measure of frequency. "NO HELP" time is correlated with average "NO HELP" time. The explanation for this is that "NO HELP" time is considered in the measure of "NO HELP" average time. Difference in Time is correlated with Difference in Frequency. Again this shows the relationship between the amount of time and the number of times the "HELP" or "NO HELP" buttons were depressed. Difference in Frequency is correlated with average "HELP" time. "HELP" average time is obtained 85 by dividing "HELP" time by "HELP" frequency, thus explaining the relation- ship between difference in frequencies and "HELP" average time. In summary, the dependent measures of recall are not correlated with perceived inappropriateness of anti-social behavior and the cooperation measures. The one exception to this is the correlation between verbal con- tent and "NO HELP" frequency. Perceived inappropriateness of anti-social behavior is not correlated with the cooperation measures. Table I Correlation Matrix of Dependent Measures é) >8 O- H U H C.‘ OJ OJ U) 9- OJ Q4 LT: 00 (D H :3 H ‘3 OJ OJ OJ 0“ OJ C H c: On :1: 5 OJ :1: 2 OJ OJ H H > g g éi if :2 Es ha m 0 ‘< u H cu m m 5 5 O H C: H 0-: >5 3: >- U U 5-! ['4 O CO OJ D- H U U C" C“ a 'H :3 0 OJ C: O C: OJ OJ OJ OJ OJ H U‘ H H :1: OJ 2 OJ H H CO DO H 0 OJ CO Q- :J :3 OJ OJ CO “I OJ U [D .0 On OJ 0" OJ C" '44 ‘H H H ‘H c; a u m .E o .5 m ru ‘H m m in -.-4 o 0 c: u u -.-4 -.-4 :> :> H a. U > H E—I In _ [-¢ Fm Q C: <1 <3 G ** ** Pietorial 1000 .4? 03¢ .03 -004 -o 09 009 001 -004 -003 -005 0002 009 Consequences 1.00 .43***.13 -.02 .003 .03 -.10 -.12 -.10 -.17 -.003 .12 Verbal 1.00 .13 -.11 .13 .13 .20‘*-.01 .05 -.21**-.01 -.05 Inappropriateness 1.00 -.08 .004 .13 .09 .01 .05 -.10 .09 -.08 Time Help 1.00 .16‘-.09 -.13 .23**-*.02 .5 *-*.11 .23** Frequency Help 1.00-.05 .21“: .OS .14 -.21**-.12 -.04 Time No Help 1.00 .4 *—*.1f-.05 -.07 .8 *5.11 Frequency No Help 1.00 -.13 -.07 -.14 .07 -»06 Difference Time 1.00 .93***.6f*-*.15 .07 Difference Frequency 1.00 .43 -.05 -.04 Average Time Help 1.00 -.06 .14 Average Time No Help 1.00 -u11 Difference Average Time 1.00 *** = .001 level of significance ** = .01 level of significance * = .05 level of significance 86 ANOVA Results One way analysis of variance was performed on the dependent measures of: a) recall of pictorial content, b) recall of consequences of anti- social behavior, c) recall of verbal content, d) perceived inappropriate- ness of anti-social behavior, and e) post-viewing cooperation. The results for each dependent variable are presented for the total sample, females and males. Table 11 contains the group means, standard deviations and analysis of variance tests for the total sample for the dependent measures of re- call of pictorial content, recall of consequences of anti-social be- havior, recall of verbal content, perceived inappropriateness of anti- social behavior and post-viewing cooperation. The F tests for the dependent measures of recall and perceived inappropriateness of anti- social behavior are significant beyond the .05 level. For the dependent measure of post-viewing cooperation, time "HELP", Difference in Time, and Average Difference in Time are all significant at the .10 level or beyond. Comment conditions are responsible for significant F tests for the dependent measures of recall and perceived inappropriateness of anti- social behavior. 87 Table II Analysis of Variance Tests on Dependent Measures -- Total Sample Recall of Pictorial Content Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 18.1818 15.3182 18.2727 14.8182 15.4286 15.5217 (1.6224) (1.4601) (1.1205) (2.2390) (2.2265) (2.2537) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 262.6786 52.5357 14.9570 .0001 WITHIN GROUPS 126 442.5638 3.5124 TOTAL 131 705.2424 Recall of Consequences Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 9.3636 7.5455 9.8182 7.1364 7.7619 7.8696 (1.3644) (1.1843) (1.5004) (1.7538) (1.0443) (1.6041) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 127.9227 25.5842 12.4550 .0001 WITHIN GROUPS 126 258.8273 2.0542 TOTAL 131 386.7500 Recall of Verbal Content Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 8.6818 6.5455 8.1818 6.0000 6.9524 6.7826 (1.6729) (1.1843) (1.7630) (1.6330) (1.3593) (1.5361) SOURCE D.F. SUM OR SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 115.8997 23.1799 9.7890 .0001 WITHIN GROUPS 126 298.3654 2.3680 TOTAL 131 414.2652 88 Table II (Continued) Inappropriateness of Anti-Social Behavior Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 16.6818 15.3636 15.8182 14.2273 15.0476 15.5217 (1.5549) (1.8910) (2.4424) (3.1310) (1.8835) (2.1078) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 72.9373 14.5875 2.9380 .015 WITHIN GROUPS 126 625.6915 4.9658 TOTAL 131 698.6288 Analysis of Variance for Post-Viewing Cooperation -- Help Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 15.4091 13.3636 18.2273 23.9091 11.9545 23.2273 (12.5230) (15.2176) (13.9657) (18.8728) (14.1908) (19.8948) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 2781.7273 556.3455 2.1720 .061 WITHIN GROUPS 126 32272.9091 256.1342 TOTAL 131 35054.6364 Post-Viewing Cooperation -- Help Frequency Group Means (Standard Deviations) Mother Mather No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 6.9091 6.2727 5.6364 5.6364 6.5909 7.2727 (6.6254) (7.0521) (3.4023) (6.1221) (10.2291) (8.5867) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 49.2500 9.8500 .1840 .968 WITHIN GROUPS 126 6742.0455 53.5083 TOTAL 131 6791.2955 89 Table II (Continued) Post-Viewing Cooperation -- No Help Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 4.3182 .8182 2.7273 .7273 4.1364 2.6818 (8.4650) (1.5625) (6.8046) (1.5486) (11.3821) (5.6853) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 264.2500 52.8500 1.1140 .350 WITHIN GROUPS 126 5978.1364 47.4455 TOTAL 131 6242.3864 Post-Viewing Cooperation -- No Help Frequency Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.8636 .9545 1.0000 1.3636 .7273 1.5455 (4.2460) (2.0811) (1.7457) (5.1041) (1.0771) (3.2179) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 19.7879 3.9576 .3770 .864 WITHIN GROUPS 126 1322.4545 10.4957 TOTAL 131 1342.2424 Post-Viewing Cooperation -- Difference In Time Group Means (Standard Deviations) Mother Mather No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 11.0909 12.5455 15.5000 23.1818 7.8182 20.5455 (16.6099) (15.8737) (15.9157) (19.4101) (18.9476) (22.1830) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 3724.4015 744.8803 2.3330 .046 WITHIN GROUPS 126 40221.3182 319.2168 TOTAL 131 43945.7197 90 Table II (Continued) Post-Viewing Cooperation -- Difference in Frequency Group Means (Standard Deviations) Mbther MOther No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 5.0455 5.3182 4.6364 4.2727 5.8636 5.2727 (4.6545) (7.2532) (4.1696) (5.6752) (10.5257) (9.8231) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 248.4924 49.6985 .6140 .689 WITHIN GROUPS 126 10193.2273 80.8986 TOTAL 131 10441.7197 Post-Viewing Cooperation -- Average Time Help Group Means (Standard Deviations) MOther Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 3.0273 4.5464 4.3214 11.1936 4.2318 5.5364 (3.2356) (10.5324) (5.2896) (16.4172) (10.5539) (10.0253) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 932.6890 186.5378 1.7790 .123 WITHIN GROUPS 126 13244.2198 105.1129 TOTAL 131 14176.9088 Post-Viewing Cooperation -- Average Time No Help Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.2778 .3095 .8032 .3750 3.1745 .8053 (2.2395) (.5009) (2.1596) (.8443) (10.7444) (1.7253) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 124.4595 24.8919 1.1570 .334 WITHIN GROUPS 126 2710.2827 21.5102 TOTAL 131 2834.7422 91 Table II (Continued) Post-Viewing Cooperation -- Difference in Average Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.7500 4.2423 3.5182 10.8186 1.0573 4.7311 (3.4634) (10.6596) (5.3811) (16.6334) (15.6735) (10.3374) SOURCE D.F. SUMS OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 1300.7135 260.1427 1.9970 .084 WITHIN GROUPS 126 16416.6762 130.2911 TOTAL 131 17717.3897 92 Table III contains the group means, standard deviations and analysis of variance tests for males for the dependent measure of recall of pictorial content, recall of consequences of anti-social behavior, recall of verbal content, perceived inappropriateness of anti-social behavior and post- viewing cooperation. The F tests for recall of pictorial content, conse— quences of anti-social behavior and verbal content are significant beyond the .01 level. Perceived inappropriateness of anti-social behavior and post—viewing cooperation are not significant. Comment conditions are re— sponsible for the significant F tests. Table III Analysis of Variance Tests on Dependent Variables (Males) Recall of Pictorial Content Group Means (Standard Deviations) Mbther Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 18.4545 15.8182 18.3636 15.7273 16.3636 16.5455 (1.4397) (1.5374) (.9244) (1.4206) (2.1574) (1.8091) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 82.6667 16.5333 5.5110 .0001 WITHIN GROUPS 6O 152.3636 2.5394 TOTAL 65 235.0303 93 Table III (Continued) Recall of Consequences Group Means (Standard Deviations) Mbther MOther No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 9.8182 7.8182 9.3636 7.2727 7.9091 8.6364 (1.2505) (1.2505) (1.6895) (1.5551) (1.0445) (1.6895) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 52.9848 10.5970 5.1500 .001 WITHIN GROUPS 60 123.4545 2.0576 TOTAL 65 176.4394 Recall of Verbal Content Group Means (Standard Deviations) Mbther Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 9.3636 6.8182 7.7273 6.6364 7.0909 7.7273 (1.3618) (.7508) (1.7939) (1.5667) (1.7581) (1.1037) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 54.2576 10.8515 5.2510 .0001 WITHIN GROUPS 60 124.0000 2.0667 TOTAL 65 178.2476 Inappropriateness of Anti-Social Behavior Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 15.9091 14.6364 15.5455 15.3636 14.9091 15.0000 (1.2210) (1.4334) (2.7336) (2.4606) (2.0226) (2.6077) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 11.9545 2.3909 .5130 .765 WITHIN GROUPS 60 279.6364 4.6606 TOTAL 65 291.5909 94 Table III (Continued) Post-Viewing C00peration -- Help Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 16.1818 17.5455 19.7273 24.3636 18.4545 20.4545 (11.5570) (18.5707) (14.5951) (17.0960) (17.6826) (18.3323) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 465.7576 93.1515 .3420 .885 WITHIN GROUPS 60 16324.7273 272.0788 TOTAL 65 16790.4848 Post-Viewing Cooperation -- Help Frequency Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 6.2727 5.8182 5.1818 5.4545 8.7273 9.6364 (3.5239) (4.0452) (2.4421) (5.2797) (14.2555) (11.4566) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 191.5758 38.3152 .5790 .716 WITHIN GROUPS 60 3970.9091 66.1818 TOTAL 65 4162.4848 Post-Viewing Cooperation -- No Help Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 4.3636 .9091 4.4545 1.0000 2.3636 3.5455 (9.8415) (1.7003) (9.3420) (1.9494) (6.8742) (6.5782) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 148.1364 28.0273 .5980 .702 WITHIN GROUPS 60 2813.4545 46.8909 TOTAL 65 2953.5909 95 Table III (Continued) Post-Viewing Cooperation -- No Help Frequency Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone .9091 1.1818 1.0909 2.5455 .4545 2.6364 (1.3751) (2.7136) (1.9212) (7.1744) (.8202) (4.2959) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 44.9848 8.9970 .6460 .665 WITHIN GROUPS 60 835.4545 13.9242 TOTAL 65 880.4394 Post-Viewing Cooperation -- Difference in Time Group Means (Standard Deviations) Mbther Mbther No Comment Comment Female Adult Female Adult Peer Viewing Comment No Comment Viewing Alone 11.8182 16.6364 15.2727 23.3636 16.0909 16.9091 (3.6051) (5.8392) (5.2882) (5.4760) (5.3436) (6.4208) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 777.5909 155.5182 .4850 .786 WITHIN GROUPS 6O 19230.7273 320.5121 TOTAL 65 20008.3182 Post-Viewing Cooperation -- Difference in Frequency Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 5.3636 4.6364 4.0909 2.9091 8.2727 7.0000 (4.3422) (4.1297) (3.0151) (3.4192) (14.4713) (13.2061) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 679.7121 135.8424 1.0970 .372 WITHIN GROUPS 60 7437.8182 123.9636 TOTAL 65 8117.5303 96 Table III (Continued) Post-Viewing Cooperation -- Average Time Help Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 2.9700 6.9336 5.5273 11.8709 7.4955 1.9755 (2.2288) (14.5558) (6.8071) (17.5467) (14.4969) (2.2870) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 693.8520 138.7704 1.0590 .392 WITHIN GROUPS 60 7864.5209 131.0753 - TOTAL 65 8558.3729 Post-Viewing C00peration -- Average Time No Help Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.4545 .2782 1.3109 .2955 1.2273 .6555 (2.8147) (.5208) (2.9945) (.6405) (3.4305) (.6920) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 15.2913 3.0583 .6150 .688 WITHIN GROUPS 6O 298.1805 4.9697 TOTAL 65 313.4718 Post Viewing Cooperation -- Difference in Average Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.5155 6.6645 4.2164 11.5755 6.6245 1.3200 (2.3094) (14.6853) (7.0056) (17.7485) (14.6848) (2.5348) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 817.6622 163.5324 1.2160 .313 WITHIN GROUPS 60 8071.4685 134.5245 TOTAL 65 8889.1308 97 Table IV contains the group means, standard deviations and analysis of variance tests for females for the dependent measures of recall of pictorial content, recall of consequences of anti-social behavior, recall of verbal content, perceived inappropriateness of anti-social behavior and post-viewing cooperation. The F tests for all dependent measures except post-viewing cooperation are significant beyond the .01 level of significance. For the dependent measure of cooperation, Time "HELP", Difference in Time, Average Time "HELP" and Difference in Average Time are significant at the .10 level or beyond. Comment conditions are re- sponsible for significant F tests for the dependent measures of recall. Table IV Analysis of Variance Tests on Dependent Variables (Females) Recall of Pictorial Content Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 17.9091 14.8182 18.1818 13.9091 14.4000 14.5833 (1.8141) (1.2505) (1.3280) (2.5867) (1.8974) (2.2747) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 196.8652 39.3730 10.6220 .0001 WITHIN GROUPS 60 222.4076 3.7068 TOTAL 65 419.2727 98 Table IV (Continued) Recall of Consequences Group Means (Standard Deviations) Mether Mether No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 8.9091 7.2727 10.2727 7.0000 7.6000 7.1667 (1.3751) (1.1037) (1.1909) (2.0000) (1.0750) (1.1934) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 92.6000 18.5200 9.9800 .0001 WITHIN GROUPS 60 111.3394 1.8557 TOTAL 65 203.9394 Recall of Verbal Content Group Means (Standard Deviations) Mbther Mbther No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 8.0000 6.2727 8.6364 5.3636 6.8000 5.9167 (1.7321) (1.4894) (1.6895) (1.5015) (.7888) (1.3790) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 88.0288 17.6058 8.1390 .0001 WITHIN GROUPS 60 129.7894 2.1632 TOTAL 65 217.8182 Inappropriateness of Anti-Social Behavior Group Means (Standard Deviations) Mother Mether No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 17.4545 16.0909 16.0909 13.0909 15.2000 16.0000 (1.5076) (2.0715) (3.2115) (3.4192) (1.8135) (1.4771) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 115.6121 23.1224 4.8670 .001 WITHIN GROUPS 60 285.0545 4.7509 TOTAL 65 400.6667 99 Table IV (Continued) Post-Viewing Cooperation -- Help Time Group Means (Standard Deviations) Mether Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 14.6364 9.1818 16.7273 23.2727 5.4545 26.0000 (13.9447) (10.1471) (13.8426) (21.3265) (4.1560) (21.8678) SOURCE D.F. SUM OF SQUARE MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 3441.7576 688.3515 2.8690 .022 WITHIN GROUPS 60 14393.2727 239.8879 TOTAL 65 17835.0303 Post—Viewing Cooperation -- Help Frequency Group Means (Standard Deviations) Mbther Mbther No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 7.5455 6.7273 6.0909 5.8182 4.4545 4.9091 (8.8810) (9.3605) (4.2298) (7.1249) (2.5442) (3.3602) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 71.5305 14.3061 .3390 .887 WITHIN GROUPS 60 2529.0909 42.1515 TOTAL 65 2600.6212 Post-Viewing Cooperation -- No Help Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 4.2727 .7273 1.0000 .4545 5.9091 1.8182 (7.3224) (1.4894) (1.8439) (1.0357) (5.9091) (1.8182) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 271.6364 54.3273 1.0820 .379 WITHIN GROUPS 60 3011.6364 50.1939 TOTAL 65 3283.2727 100 Table IV (Continued) Post-Viewing Cooperation -- No Help Frequency Group Means (Standard Deviations) Mother Mether No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 2.8182 .7273 .9091 .1818 1.0000 .4545 (5.8278) (1.2721) (1.6406) (.4045) (1.2649) (.8202) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 47.8939 9.5788 1.4120 .233 WITHIN GROUPS 60 407.0909 6.7848 TOTAL . 65 454.9848 Post-Viewing Cooperation -- Difference in Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 10.3636 8.4545 15.7273 22.8182 -.4545 24.1818 (15.6478) (10.8293) (14.9739) (21.4747) (17.0080) (23.3303) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 4795.0303 959.0061 3.0590 .016 WITHIN GROUPS 60 18811.4545 313.5242 TOTAL 65 23606.4848 Post-Viewing Cooperation -- Difference in Frequency Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 4.7273 6.0000 5.1818 5.6364 3.4545 4.4545 (5.1399) (9.6125) (5.1734) (7.2010) (3.2362) (3.6977) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 45.6364 9.1273 .2470 .940 WITHIN GROUPS 60 2215.8182 36.9303 TOTAL 65 2261.4545 101 Table IV (Continued) Post-Viewing Cooperation -- Average Time Help Group Means (Standard Deviations) Mother Mbther No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 3.0845 2.1591 3.1155 10.5164 .9755 9.0973 (4.1244) (2.9231) (3.0369) (16.0345) (.6742) (13.3394) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 863.7391 172.7478 2.2040 .066 WITHIN GROUPS 60 4702.7636 78.3794 TOTAL 65 5566.5026 Post-Viewing Cooperation -— Average Time No Help Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.1000 .3409 .2955 .4545 5.1218 .9545 (1.5937) (.5034) (.5101) (1.0357) (14.9103) (2.3922) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 191.1030 38.2206 .9880 .433 WITHIN GROUPS 60 2321.6655 38.6944 TOTAL 65 2512.7685 Post-Viewing Cooperation -- Difference in Average Time Group Means (Standard Deviations) Mother Mother No Female Adult Female Adult Peer Viewing Comment Comment Comment No Comment Viewing Alone 1.9845 1.8200 2.8200 10.0618 —4.1464 8.1427 (4.4426) (3.1700) (3.2647) (16.2707) (15.2378) (13.8702) SOURCE D.F. SUM OF SQUARES MEAN SQUARE F RATIO F PROBABILITY BETWEEN GROUPS 5 1415.0826 283.0165 2.3270 .054 WITHIN GROUPS 60 7297.4979 121.6250 TOTAL 65 8712.5806 102 Table V presents a summary of the F tests for each dependent measure. Table V Summary Table Overall F Tests for Dependent Variables Total Sample Dependent Variable F Ratio Probability Recall of Pictorial Content 14.9570 .0001 Recall of Consequences 12.4550 .0001 Recall of Verbal 9.7890 .0001 Inappropriateness of Anti-Social Behavior 2.9380 .015 Cooperation Help Time 2.1720 .061 Help Frequency .1840 .968 No Help Time 1.1140 .350 No Help Frequency .3770 .864 Difference in Time 2.3330 -046 Difference in Frequency .6140 ~689 Average Time Help 1.7750 .123 Average Time No Help 1.1570 .334 Difference in Average Time 1.9970 .084 103 Table V (Continued) Total Sample -- Males Dependent Variable F Ratio Probability Recall of Content 5.5110 .0001 Recall of Consequences 5.1500 .001 Recall of Verbal 5.251 .0001 Inappropriateness of Anti-Social Behavior .5130 .765 Cooperation Help Time .3420 .885 Help Frequency .5790 .716 No Help Time .5980 .702 No Help Frequency .6460 .665 Difference in Time .4850 .786 Difference in Frequency 1.0970 .372 Average Time Help 1.0590 .392 Average Time No Help .6150 .688 Difference in Average Time 1.2160 .313 Total Sample —- Females Dependent Variable F Ratio Probability Recall of Content 10.6220 .0001 Recall of Consequences 9.9800 .0001 Recall of Verbal 8.1390 .0001 Inappropriateness of Anti-Social Behavior 4.8670 .001 Cooperation Help Time 2.8690 .022 Help Frequency .3390 .887 No Help Time 1.0820 .379 No Help Frequency 1.4120 .233 Difference in Time 3°0590 ~016 Difference in Frequency .2470 .940 Average Time Help 2.2040 .066 Average Time No Help .9880 .433 Difference in Average Time 2.3270 .054 104 Hypotheses Test The above analysis of variance tests were performed to determine if the independent variable of viewing condition has an effect on children's ability to recall pictorial content, consequences of anti-social behavior and verbal content, to perceive as inappropriate anti-social behavior and to demonstrate post-viewing cooperation. Where initial analysis of vari- ance tests are significant the appropriate post hoc analysis is made to test specific hypotheses. Test of Hypothesis 1 Null Hypothesis 1: There is no difference in mean scores be- tween "mother comment" and "mother no comment" conditions on subjects' ability to recall pictorial content. Alternative Hypothesis 1: The mean scores of recall pictorial content will be higher in the "mother comment" condition than in the "mother no comment" condition. The hypothesis was tested by the Tukey post hoc technique to determine whether the contrast between scores on the "mother comment" and "mother no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participating in the condition where mothers provide comments are able to recall significantly more content than those subjects in the condition where mothers provide no comments (mean "mother comment" - 18.1818, mean "mother no comment" = 15.3181). Table VI shows the contrast of means for "mother comment" and "mother no comment" conditions as well as the level of significance for the contrast. 105 Hypothesis 1A Null Hypothesis 1A: There is no difference in mean scores be- "mother comment" and "mother no comment" conditions on tween subjects' ability to recall consequences of anti-social behavior. Alternative Hypothesis 1A: The mean scores of recall of con- sequences of anti-social behavior will be higher in the "mother comment" condition than in the "mother no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "mother comment" and "mother no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects partici- pating in the condition where mothers recite comments are able to recall significantly more consequences of anti-social behavior than those sub- jects in the condition where mothers provide no comments. (Mean "mother comment" = 9.3636, mean "mother no comment" = 7.5455). Table VI shows the contrast of means for "mother comment" and "mother no comment" conditions as well as the level of significance for the contrast. Hypothesis 1B Null Hypothesis 1B: There is no difference in mean scores between "mother comment" and "mother no comment" conditions on subjects' ability to recall verbal content. Alternative Hypothesis 1B: The mean scores of subjects' ability to recall verbal content will be higher in the "mother comment" condition than in the "mother no comment" condition. 106 This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "mother comment" and "mother no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participating in the condition where mothers are providing comments are able to recall more verbal content than those subjects in the condition where mothers provide no comments (mean "mother comment" = 8.6818, mean "mother no comment" = 6.5455). Table VI shows the contrast between means as well as the level of significance for the contrast. Hypothesis lC Null Hypothesis 1C: There is no difference in mean scores between "mother comment" and "mother no comment" conditions on subjects' perceived inappropriateness of anti-social behavior. Alternative Hypothesis 1C: The mean scores of perceived in- appropriateness of anti-social behavior will be higher in the "mother comment" condition than in the "mother no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "mother comment" and "mother no comment" conditions is significant. The null hypothesis was tested and accepted. There is no statistically significant difference in scores on perceived inappropriateness of anti-social behavior for subjects participating in the condition where mothers are providing comments and the condition where mothers provided no comments (mean "mother comment" = 16.6811, mean "mother no comment" = 15.3636). Table VI shows the contrast between means; 1C is accepted. 107 Hypothesis 1D Null Hypothesis 1D: There is no difference in mean scores be- tween "mother comment" and "mother no comment" conditions on subjects' post-viewing cooperation. Alternative Hypothesis 1D: The mean scores of postdviewing cooperation will be higher in the "mother comment" condition than in the "mother no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "mother comment" and "mother no comment" condition is significant. The Tukey post hoc technique is used when initial analysis of variance tests produce significant F ratios. For the total sample post hoc analyses on post-viewing coopera- tion were performed on Time Help, Difference in Time, and Difference in Average Time. The null hypotheses were tested and accepted. There are no statistically significant differences in scores on post-viewing coopera- tion (Time Help, Difference in Time, and Difference in Average Time) for - subjects participating in the condition where mothers are providing com- ments and the condition where mothers provide no comments (Time Help -— mean "mother comment" = 15.4091, mean "mother no comment" = 13.3636; Difference in Time -- mean "mother comment" = 11.0909, mean "mother no comment" 8 12.5455; Difference in Average Time -- mean "mother comment" = 1.6773, mean "mother no comment" = -.3014). Table VI shows the contrast of means. 108 Table VI Hypothesis 1, 1A, 1B, 1C, 1D: Tukey contrasts between "mother comment" and "mother no comment" conditions and levels of significance. Dependent Variables Contrast Significance Ho Recall of Content 2.8636 .01 Reject Recall of Consequences 1.8181 .01 Reject Recall of Verbal 2.1363 .01 Reject Perceived Inappropriateness 1.3182 N.S. Accept Cooperation Help Time 2.0455 N.S. Accept Help Frequency .6364 N.S. Accept No Help Time 3.5000 N.S. Accept No Help Frequency .9091 N.S. Accept Difference in Time , 1.4546 N.S. Accept Difference in Frequency .2727 N.S. Accept Average Time Help 1.5191 N.S. Accept Average Time No Help .9678 N.S. Accept Difference in Average Time 2.4926 N.S. Accept Hypothesis 2 Null Hypothesis 2: There is no difference in mean scores between "female adult comment" and "female adult no comment" conditions on subjects' ability to recall content. Alternative Hypothesis 2: The mean score of recall of pic- torial content will be higher in the "female adult comment" condition than in the "female adult no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "female adult comment" and "female adult no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects parti- cipating in the condition where the female adult is providing comments are 109 able to recall more verbal content than those subjects in the condition where the female adult provides no comments (mean "female adult comment" = 18.2727; mean "female adult no comment" = 14.8182). Table VII shows the contrast between means as well as the level of significance for the con- trast. Hypothesis 2A Null Hypothesis 2A: There is no difference in mean scores between "female adult comment" and "female adult no comment" conditions on subjects' ability to recall consequences of anti-social behavior. Alternative Hypothesis 2A: The mean scores of recall conse- quence of anti-social behavior will be higher in the "female adult comment" condition than in the "female adult no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "female adult comment" and "female adult no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects parti- cipating in the condition where the female adult is providing comments are able to recall more consequences of anti-social behavior than those sub- jects in the condition where the female adult provides no comments (mean "female adult comment" = 9.8182, mean "female adult no comment" = 7.1364). Table VII shows the contrast between means as well as the level of signi— ficance for the contrast. 110 Hypothesis 2B Null Hypothesis 2B: There is no difference in mean scores between "female adult comment" and "female adult no comment" conditions on subjects' ability to recall verbal content. Alternative Hypothesis 2B: The mean scores of recall of verbal content will be higher in the "female adult comment" condition than in the "female adult no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "female adult comment" and "female adult no comment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects parti- cipating in the condition where the female adult is providing comments are able to recall more verbal content than those subjects in the condition where the female adult provides no comments (mean "female adult comment" = 8.1818, mean "female adult no comment" = 6.000). Table VII shows the con- trast between means as well as the level of significance for the contrast. Hypothesis 2C Null Hypothesis 2C: There is no difference in mean scores between "female adult comment" and "female adult no comment" conditions on subjects' perceived inappropriateness of anti- social behavior. Alternative Hypothesis 2C: The mean scores of perceived inappropriateness of anti-social behavior will be higher in the "adult comment" condition than in the "adult no comment" condition. This hypothesis was tested by the Tukey post hoc technique to determine 111 whether the contrast between scores in the "female adult comment" and "fe- male adult no comment" conditions is significant. The null hypothesis was tested and accepted. There is no statistically significant difference between the conditions where the female adult provides comments and where the female adult does not provide comments (mean "female adult comment" = 15.8182, mean "female adult no comment" = 14.2273). Table VII shows the contrast between means. Hypothesis 2D Null Hypothesis 2D: There is no difference in mean scores between "female adult comment" and "female adult no comment" conditions on subjects' post-viewing cooperation. Alternative Hypothesis 2D: The mean scores of post-viewing cooperation will be higher in the "female adult comment" condition than in the "female adult no comment" condition. This hypothesis was tested by the Tukey post hoc technique to deter- mine whether the contrast between scores in the "female adult comment" and "female adult no comment" condition is significant. Post hoc analyses were performed on Time Help, Difference in Time, and Difference in Average Time. There are no statistically significant differences in scores on post-viewing cooperation (Time Help, Difference in Time, and Difference in Average Time) for subjects participating in the condition where an adult is providing comments and the condition where an adult provides no com- ments (Time Help -- mean "female adult comment" = 18.2273, mean "female adult no comment" = 23.9091; Difference in Time -— mean "female adult com- ment" = 15.5000, mean "female adult no comment" = 23.1818; Difference in Average Time -- mean "female adult comment" = 3.5182, mean "female adult no comment" = 10.8186. 112 Table VII Hypotheses 2, 2A, 2B, 2C, 2D: Tukey contrasts between "female adult com- ment" and "female adult no comment" conditions and levels of significance. Dependent Variables Contrast Significance Ho Recall of Content 3.4545 .01 Reject Recall of Consequences 2.6818 .01 Reject Recall of Verbal 2.1818 .01 Reject Perceived Inappropriateness 1.5909 N.S. Accept Cooperation Help Time 5.6818 N.S. Accept Help Frequency 0.00 N.S. Accept No Help Time 2.0000 N.S. Accept No Help Frequency .3636 N.S. Accept Difference in Time 8.6364 N.S. Accept Difference in Frequency .2273 N.S. Accept Average Time Help 6.8722 N.S. Accept Average Time No Help .4282 N.S. Accept Difference in Average Time 7.3004 N.S. Accept Hypothesis 3 Null Hypothesis 3: There is no difference in mean scores between the mother and female adult "comment" conditions and the mother and female adult "no comment" conditions on sub- jects' ability to recall pictorial content. \ Alternative Hypothesis 3: The mean scores of recall of pic- torial content will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between scores in the "comment" and "no comment" conditions was significant. The Scheffe post hoc technique is more powerful than the Tukey post hoc technique when comparisons among means 113 are complex as in Hypotheses 3, 4 and 5. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participating in the conditions where "comments" are made are able to recall more content than those subjects in conditions where "no comments" are made (mean "comments" = 18.1818 + 18.2727, mean "no comments" = 15.3182 + 14.8182). Table VIII shows the contrast between means as well as the level of sig- nificance for the contrast. Hypothesis 3A Null Hypothesis 3A: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and the mother and female adult "no comment" conditions on subjects' ability to recall consequences of anti-social behavior. Alternative Hypothesis 3A: The mean scores of recall of conse— quences of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. This hypothesis was tested by the Scheffe post hoc technique to de- termine whether the contrast between scores in the "comment" and "no comment" conditions was significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participating in the conditions of anti-social behavior where comments are made are able to recall more consequences than subjects participating in conditions where no comments are made (mean "comments" = 9.3636 + 9.8182, mean "no comments" = 7.5455 + 7.1364). Table VIII shows the contrast between means as well as the level of significance for the contrast. 114 Hypothesis BB Null Hypothesis 3B: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and mother and female adult "no comment" conditions on subjects' ability to recall verbal content. Alternative Hypothesis 3B: The mean scores of recall of ver- bal content will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. This hypothesis was tested by the Scheffe post hoc technique to de- termine whether the contrast between scores in the "comment" and "no come ment" conditions is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participatiqgin the conditions where comments are made are able to recall more verbal content than those subjects in conditions where no comments are made (mean "com- ments" = 8.6818 + 8.1818, mean "no comments" = 6.5455 + 6.000). Table VIII shows the contrast between means as well as the level of significance for the contrast. Hypothesis 3C Null Hypothesis 3C: There is no difference in mean scores be- tween mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' perceived inappro- priateness of anti-social behavior. Alternative Hypothesis 3C: The mean score of perceived in- appropriateness of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. 115 This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between scores in the "comment" and "no comment" conditions is significant. The null hypothesis was tested and rejected at the .10 significance level. Subjects participating in the conditions where comments are made perceive more anti-social behavior as inappropri- ate than those subjects in conditions where no comments are made (mean "commenu§'- 16.6818 + 15.8182, mean "no comments" - 15.3636 + 14.2273). Table VIII shows the contrast between means as well as the level of signi- ficance for the contrast. Hypothesis 3D Null Hypothesis 3D: There is no difference in mean scores between mother and female adult "comment" and mother and female adult "no comment" conditions on subjects' post- viewing cooperation. Alternative Hypothesis 3D: The mean score of post-viewing cooperation will be higher in the mother and female adult "comment" conditions than in the mother and female adult "no comment" conditions. This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between scores in the "comment" and "no comment" conditions is significant. Post hoc analyses are performed on Time Help, Difference in Time, and Difference in Average Time. There are no statis- tically significant differences in scores on post-viewing cooperation (Time Help, Difference in Time, and Difference in Average Time) for sub- jects participating in the conditions where comments are made and the conditions where comments are not made (Time Help -- mean "comments" = 116 15.4091 + 18.2273, mean "no comments" = 13.3636 + 23.9091; Difference in Time -- mean "comments" = 11.0909 + 15.5000, mean "no comments" 8 12.5455 + 23.1818; Difference in Average Time -- mean "comments" = 1.7500 + 3.5182, mean "no comments" = 4.2423 + 10.8186). Table VIII Hypotheses 3, 3A, BB, 30, 3D: Scheffé contrasts between "comment" and "no comment" conditions and levels of significance. Dependent Variables Contrast Significance Ho Recall of Content 6.3181 .01 Reject Recall of Consequences 4.4999 .01 Reject Recall of Verbal 4.3181 .01 Reject Perceived Inappropriateness 2.9091 .01 Reject Cooperation Help Time 3.6363 N.S. Accept Help Frequency .6364 N.S. Accept No Help Time 5.5000 N.S. Accept No Help Frequency .5455 N.S. Accept Difference in Time 9.1364 .05 Accept Difference in Frequency .0910 N.S. Accept Average Time Help 8.3913 N.S. Accept Average Time No Help 1.3960 N.S. Accept Difference in Average Time -9.7927 N.S. Accept Hypothesis 4 Null Hypothesis 4: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and "peer viewing" condition on subjects' ability to recall pic- torial content. Alternative Hypothesis 4: The mean scores on subjects' abil- ity to recall pictorial content will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. 117 This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between scores in the "comment" conditions and "peer viewing" condition is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects participating in the conditions where comments are made are able to recall more content than those subjects viewing with a peer (mean "comment" = 18.1818 + 18.2727, mean "peer" = 15.4286). Table IX shows the contrast between means as well as the level of significance for the contrast. Hypothesis 4A Null Hypothesis 4A: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and "peer viewing" condition on subjects' ability to recall con- sequences of anti-social behavior. Alternative Hypothesis 4A: The mean scores of recall of con- sequences of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "peer view- ing" condition. This hypothesis was tested by the Scheffé post hoc technique to deter- mine whether the contrast between scores in the "comment" conditions and the "peer viewing" condition is significant. The null hypothesis is re- jected at the .01 level of signifiance. Subjects participating in the conditions where comments are made are able to recall more consequences of anti-social behavior than those subjects viewing with a peer (mean "comments" = 9.3636 + 9.8182, mean "peer viewing" = 7.7619). Table IX shows the con- trast between means as well as the level of significance for the contrast. 118 Hypothesis 4B Null Hypothesis 4B: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and "peer viewing" condition on subjects' ability to recall ver- bal content. Alternative Hypothesis 4B: The mean scores of recall of ver- bal content will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. This hypothesis was tested by the Scheffé post hoc technique to deter- mine whether the contrast between scores in the "comment" and "peer view- ing" conditions was significant. The null hypothesis was rejected at the .05 level of significance. Subjects participating in the conditions where comments are made are able to recall more verbal content than those subjects viewing with a peer (mean "comment" = 8.6818 + 8.1818, mean "peer viewing" = 6.9524). Table IX shows the contrast between means as well as the level of significance for the contrast. Hypothesis 4C Null Hypothesis 4C: There is no difference in the mean scores between the mother and female adult "comment" conditions and "peer viewing" condition on subjects' perceived inappropriate- ness of anti-social behavior. Alternative Hypothesis 4C: The mean scores of perceived in- appropriateness of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. 119 This hypothesis was tested by the Scheffe post hoc technique to de- termine whether the contrast between scores in the "comment" and "peer viewing" conditions is significant. The null hypothesis was tested and accepted. There is no statistically significant difference in mean scores between subjects in "comment" conditions and those subjects viewing with a peer (mean "comment" = 16.6818 + 15.0182, mean "peer viewing" = 15.0476). Table IX shows the contrast between means. Hypothesis 4D Null Hypothesis 4D: There is no difference in mean scores be- tween the mother and female adult "comment" conditions and "peer viewing" condition on subjects' post-viewing cooperation. Alternative Hypothesis 4D: The mean scores of post-viewing cooperation will be higher in the mother and female adult "comment" conditions than in the "peer viewing" condition. This hypothesis was tested by the Scheffé post hoc technique to de- termine whether the contrast between scores in the "comment" conditions and "peer viewing" condition is significant. Post hoc analyses were per- formed on Time Help, Difference in Time, and Difference in Average Time. There are no statistically significant differences in scores on post— viewing cooperation (Time Help, Difference in Time, Difference in Average Time) for subjects participating in the conditions where comments are made and the peer viewing condition (Time Help -- mean "comment" = 15.4091 + 18.2273, mean "peer viewing" = 11.9545; Difference in Time -- mean "com- ment" = 11.0909 + 15.5000, mean "peer viewing" = 7.8182; Difference in Average Time == mean "comment" = 1.7500 + 3.5182, mean "peer viewing" = 1.0573). 120 Table IX Hypotheses 4, 4A, 4B, 4C, 4D: Scheffe contrasts between "comment" condi- tions and "peer viewing" conditions and levels of significance. Dependent Variables Contrast Significance Ho Recall of Content 2.7987 .01 Reject Recall of Consequences 1.8290 .01 Reject Recall of Verbal 1.4794 .05 Reject Perceived Inappropriateness 1.2040 N.S. Accept Cooperation Help Time 4.8673 N.S. Accept Help Frequency .3182 N.S. Accept No Help Time .6137 N.S. Accept No Help Frequency .7046 N.S. Accept Difference in Time 5.4773 N.S. Accept Difference in Frequency 1.0227 N.S. Accept Average Time Help -.5575 N.S. Accept Average Time No Help 2.1343 N.S. Accept Difference in Average Time 1.4001 N.S. Accept Hypothesis 5 Null Hypothesis 5: There is no difference in mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' ability to recall pictorial content. Alternative Hypothesis 5: The mean scores of recall of pictorial content will be higher in the mother and female adult "comment" conditions than in the "viewing alone" con- dition. This hypothesis was tested by the Scheffe post hoc technique to de- termine whether the contrast between scores in the "comment" conditions 121 and "viewing alone" condition is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects parti- cipating in conditions where comments are made are able to recall more content than those subjects viewing alone (mean "comments" = 18.1818 + 18.2727, mean "alone" = 15.5217). Table X shows the contrast between means as well as the level of significance for the contrast. Hypothesis 5A Null Hypothesis 5A: There is no difference in mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' ability to re- call consequences of anti-social behavior. Alternative Hypothesis 5A: The mean scores of recall of consequences of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. This hypothesis was tested by the Scheffe post hoc technique to determine whether the contrast between scores in the "comment" condi- tions and "viewing alone" condition is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects in conditions where comments are made are able to recall more consequences of anti-social behavior than those subjects viewing alone (mean "com- ments" = 9.3636 + 9.8182, mean "viewing alone" = 7.8696). Table X shows the contrast between means as well as the level of significance for the contrast. 122 Hypothesis 5B Null Hypothesis 5B: There is no difference in mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' ability to recall verbal content. Alternative Hypothesis 5B: The mean scores of recall of verbal content will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. This hypothesis was tested by the Scheffe post hoc technique to de- 0 "comment" conditions termine whether the contrast between scores in the and "viewing alone" condition is significant. The null hypothesis was tested and rejected at the .01 level of significance. Subjects partici- pating in conditions where comments are made are able to recall verbal content more than those subjects viewing alone (mean "comments" = 8.6818 + 8.1818, mean "viewing alone" = 6.7826). Table X shows the contrast between means as well as the level of significance for the contrast. Hypothesis 5C Null Hypothesis 5C: There is no difference in mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' perceived inappro- priateness of anti-social behavior. 123 Alternative Hypothesis 5C: The mean scores of perceived in- appropriateness of anti-social behavior will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. O This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between scores in the "comment" conditions and "viewing alone" condition is significant. The null hypothesis was tested and accepted. There is no statistically significant difference in mean scores between subjects participating in "comment" conditions and those subjects viewing alone (mean "comments" 8 16.6818 + 15.8182, mean "viewing alone" = 15.5217). Table X shows the contrast between means. Hypothesis 5D Null Hypothesis 5D: There is no difference in mean scores between the mother and female adult "comment" conditions and "viewing alone" condition on subjects' post-viewing cooper- ation. Alternative Hypothesis 5D: The mean scores of post-viewing cooperation will be higher in the mother and female adult "comment" conditions than in the "viewing alone" condition. This hypothesis was tested by the Scheffe post hoc technique to deter- mine whether the contrast between'bommenfl'conditions and"viewing alone" condition is significant. Post hoc analyses were performed on Time Help, Difference in Time, and Difference in Average Time. There are no statisti- cally significant differences in scores on post-viewing cooperation (Time Help, Difference in Time, Difference in Average Time) for subjects partici- pating in the conditions where comments are made and the viewing alone 124 condition (Time Help -- mean "comments" = 15.4091 + 18.2273, mean "viewing alone" = 23.2273; Difference in Time -- mean "comments" = 11.0909 + 15.5000, mean "viewing alone" = 20.5455; Difference in Average Time -- mean "comments" = 1.7500 + 3.5182, mean "viewing alone" = 4.7314). Table X Hypotheses 5, 5A, 5B, 5C, 5D: Scheffe contrasts between "comment" conditions and "viewing alone" condition and levels of significance. Dependent Variables Contrast Signifiance Ho Recall of Content 2.9545 .01 Reject Recall of Consequences 1.7213 .01 Reject Recall of Verbal 1.6492 .01 Reject Perceived Inappropriateness .7283 N.S. Accept Cooperation Help Time 6.4091 N.S. Accept Help Frequency .9995 N.S. Accept No Help Time .8409 N.S. Accept No Help Frequency .1137 N.S. Accept Difference in Time 7.2501 N.S. Accept Difference in Frequency .8863 N.S. Accept Average Time Help 1.8621 N.S. Accept Average Time No Help .2353 N.S. Accept Difference in Average Time 2.0973 N.S. Accept Tests of Hypotheses, Controlling for Sex Each hypothesis was tested controlling for the sex of the subject. The sample consisted of 66 (50%) males and 66 (50%) females. Hypotheses 1, 1A, 1B, 1C, 1D Hypotheses 1, 1A, 1B, 1C, and 1D examined the contrast between mean scores between "mother comment" and "mother no comment" conditions on sub- jects' ability to (1) recall pictorial content, (1A) recall consequences of anti-social behavior, (1B) recall verbal content, (1C) perceive 125 anti-social behavior as inappropriate, and (1D) demonstrate post-viewing cooperation. When controlling for sex of subjects these hypotheses were tested by the Tukey post hoc technique to determine whether contrasts be- tween means of "mother comment" and "mother no comment" conditions were significant. Table XI shows the contrasts of means for males and females for all dependent measures between "mother comment" and "mother no comment" conditions, as well as the levels of significance for the contrasts. The findings for each dependent variable are as follows: Recall of Pictorial Content (Hypothesis 1) The null hypothesis for recall of pictorial content for both males and females were tested and rejected at the .01 level of significance. Males and females participating in conditions where mothers are providing com- ments are able to recall more content than those subjects who are in the condition where mothers are not providing comments. Recall of Consequences of Anti-Social Behavior (Hypothesis 1A) The null hypothesis for recall of consequences of anti-social behavior was tested and rejected at the .05 level of significance for males and was rejected at the .10 level of significance for females. Males and females participating in the condition where mothers are providing comments are able to recall more consequences of anti-social behavior than those subjects in the condition where mothers are not providing comments. Recall of Verbal Content (Hypothesis 1H) The null hypothesis for recall of verbal content was tested and re- jected at the .01 level of significance for males and was rejected at the .10 level of significance for females. Males and females participating in the condition where mothers are providing comments are able to recall more 126 verbal content than those subjects in the conditions where mothers are not providing comments. Perceived Inappropriateness of Anti—Social Behavior (Hypothesis 1C) The null hypotheses for perceived inappropriateness of anti-social behavior were tested and accepted for both males and females. There are no statistically significant differences in scores on perceived inappropriate- ness of anti-social behavior for males and females participating in the condition where mothers are providing comments and in the conditions where mothers are not providing comments. Post-Viewing Cooperation (Hypothesis ID) The null hypotheses for post-viewing cooperation were tested and accepted for females. The analysis of variance tests for post-viewing cooperation for males are not significant at the .10 level or beyond; therefore, post hoc analyses would be inappropriate and the null hypotheses are accepted for males. There are no statistically significant differences between scores on post-viewing cooperation for females participating in the condition where mothers are providing comments and in the condition where mothers are not providing comments. Table XI Hypotheses 1, 1A, 1B, 1C, 1D. Controlling for Sex: Tukey contrasts between "mother comment" and "mother no comment" conditions and levels of significance. Dependent Variables Contrast Significance Ho Recall of Pictorial Content Males 2.6363 .01 Reject Females 3.0909 .01 Reject Recall of Consequences Males 2.0000 .05 Reject Females 1.6364 .10 Reject Table XI (Continued) 127 Dependent Variables Contrast Significance Ho Recall of Verbal Males 2.5454 .01 Reject Females 1.7273 .10 Reject Perceived Inappropriateness Males 1.2727 N.S. Accept Females 1.3636 N.S. Accept Cooperation Help Time Males 1.3637 N.S. Accept Females 5.4546 N.S. Accept Help Frequency Males .4545 [N.S. Accept Females .8182 N.S. Accept No Help Time Males 3.4545 N.S. Accept Females 3.5454 N.S. Accept No Help Frequency Males .2727 N.S. Accept Females 2.0909 N.S. Accept Difference in Time Males 4.8182 N.S. Accept Females 1.9091 N.S. Accept Difference in Frequency Males .7272 N.S. Accept Females 1.2727 N.S. Accept Average Time Help Males 3.9636 N.S. Accept Females .9254 N.S. Accept Average Time No Help Males 1.1763 N.S Accept Females .7591 N.S. Accept Difference in Average Time Males 5.1490 N.S. Accept Females .1618 N.S. Accept 128 Hypotheses 2, 2A, ZB, 2C, 2D Hypotheses 2, 2A, 2B, 2C and 2D examined the contrast between mean scores for "female adult comment" and "female adult no comment" conditions on subjects' ability to (2) recall pictorial content, (2A) recall conse- quences of anti-social behavior, (2B) recall verbal content, (20) perceive anti-social behavior as inappropriate and (2D) demonstrate post-viewing cooperation. When controlling for sex of subjects these hypotheses were tested by the Tukey post hoc technique to determine whether contrasts be- tween means of "female adult comment" and "female adult no comment" condi- tions were significant. Table XII shows the contrasts of means for males and females for all dependent measures between "female adult comment" and "female adult no comment" conditions, as well as the levels of significance for the contrasts. The findings for each dependent variable are as fol— lows: Recall of Pictorial Content (Hypothesis 2) The null hypotheses for recall of pictorial content for both males and females were tested and rejected at the .01 level of significance. Males and females participating in the condition where adult provides come ments are able to recall more content than those subjects in the conditions where the female adult does not provide comments. Recall of Consequences of Anti-Social Behavior (Hypothesis 2A) The null hypothesis for recall of consequences of anti—social behavior was tested and rejected at the .05 level of significance for males and was rejected at the .01 level for females. Males and females participating in the conditions where the female adult provides comments are able to recall more consequences of anti-social behavior than those males and females in the conditions where the female adult does not provide comments. 129 Recall of Verbal Content (Hypothesis 2B) The null hypothesis for recall of verbal content was tested and accepted for males and was rejected at the .01 level of significance for females. Females participating in the condition where the female adult provides comments are able to recall more verbal content than those females in the condition where the female adult does not provide comments. There is no statistically significant difference in means for males in "female adult comment" and "female adult no comment" conditions. Perceived Inapproppiateness of Anti—Social Behavior (Hypothesis 20) The null hypotheses for perceived inappropriateness of anti-social behavior were tested and accepted for both males and females. There are no statistically significant differences in scores on perceived inappropri— ateness of anti-social behavior for males and females participating in the condition where the female adult provides comments and in the condition where the female adult does not provide comments. Post—Viewinngooperation (Hypothesis 2D) The null hypotheses for post-viewing cooperation were tested and accepted for females. The analysis of variance tests for post-viewing cooperation for males were not significant at the .10 level or beyond; therefore, post hoc analyses would be inappropriate and the null hypotheses are accepted for males. There are no statistically significant differences between scores on post-viewing cooperation for females participating in the condition where a female adult provides comments and the condition where a female adult does not provide comments. 130 Table XII Hypotheses 2, 2A, 2B, 20, 2D. Controlling for Sex: Tukey contrasts between "female adult comment" and "female adult no comment" conditions and levels of significance. Dependent Variables Contrasts Significance Ho Recall of Pictorial Content Males 2.6363 .01 Reject Females 4.2727 .01 Reject Recall of Consequences Males 2.0909 .05 Reject Females 3.2727 .01 Reject Recall of Verbal Males 1.0909 N.S. Accept Females 3.2728 .01 Reject Perceived Inappropriateness Males .1819 N.S. Accept Females ,3.0000 .05 Reject Cooperation Help Time Males 4.8183 N.S. Accept Females 6.5454 N.S. Accept Help Frequency Males .2727 N.S. Accept Females .2727 N.S. Accept No Help Time Males 3.4545 N.S. Accept Females .5455 N.S. Accept No Help Frequency Males 1.4546 N.S. Accept Females .7273 N.S. Accept Difference in Time Males 8.0909 N.S. Accept Females -7.0909 N.S. Accept 131 Table XII (Continued) Dependent Variables Contrasts Significance Ho Difference in Frequency Males 1.1818 N.S. Accept Females .4546 N.S. Accept Average Time Help Males 6.3436 N.S. Accept Females 7.4009 N.S. Accept Average Time No Help ' Males 1.0154 N.S. Accept Females .1590 N.S. Accept Difference in Average Time Males 7.3591 N.S. Accept Females 7.2418 N.S. Accept Hypotheses 3, 3A, 3B, 3C, 3D Hypotheses 3, 3A, 3B, 3C and 3D examined the contrasts between mean scores for "comment" and "no comment" conditions on subjects' ability to (3) recall pictorial content, (3A) recall consequences of anti—social behavior, (3B) recall verbal content, (3C) perceive anti-social behavior as inappropriate, and (3D) demonstrate post-viewing cooperation. When control— ling for sex of the subjects these hypotheses were tested by the Scheffe post hoc technique to determine whether contrasts between means of "comment" and "no comment" conditions were significant. Table XIII shows the contrasts of means for males and females for all dependent measures between "comment" and "no comment" conditions, as well as the levels of significance for the contrasts. The findings for each dependent variable are as follows: 132 Recall of Pictorial Content (Hypothesis 3) The null hypotheses for recall of pictorial content for males and females were tested and rejected at the .01 level of significance. Males and females participating in conditions where comments are made are able to recall more pictorial content than those males and females in conditions where comments are not made. Recall of Consequences of Anti—Social Behavior (Hypothesis 3A) The null hypotheses for recall of consequences of anti-social behavior for males and females were tested and rejected at the .01 level of signi— ficance. Subjects participating in conditions where comments are made are able to recall more consequences of anti-social behavior than those sub- jects in conditions where comments are not made. Recall of Verbal Content (Hypothesis 3B) The null hypothesis for recall of verbal content was tested and re- jected at the .01 level of significance for females. Males and females participating in the conditions where comments are made are able to recall more verbal content than those males and females in the conditions where comments are not made. Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 3C) The null hypothesis for perceived inappropriateness of anti-social behavior was tested and accepted for males and was rejected at the .10 level of significance for females. Females participating in conditions where comments are made are able to perceive as inappropriate more anti- social behavior than those females in conditions where comments are not made. For males, however, there is no statistically significant difference 133 in scores on perceived inappropriateness of anti—social behavior between those who are in "comment" and "no comment" conditions. Post-Viewing Cooperation (Hypothesis 3D) The null hypothesis for post-viewing cooperation was tested and accepted for females. The analysis of variance tests for post-viewing co— operation for males were not significant at the .10 level or beyond; therefore, post hoc analysis would be inappropriate and the null hypotheses were accepted for males. There are no statistically significant differ- ences between scores on post-viewing cooperation for females participating in conditions where comments are made and are not made. Table XIII Hypotheses 3, 3A, 3B, 3C, 3D. Controlling for Sex: Scheffé contrasts be- tween "comment" and "no comment" conditions and levels of significance. Dependent Variables Contrasts Significance Ho Recall of Pictorial Content Males 5.2726 .01 Reject Females 7.3636 .01 Reject Recall of Consequences Males 4.0909 .01 Reject Females 4.9091 .01 Reject Recall of Verbal Males 3.1817 .05 Reject Females 5.0001 .01 Reject Perceived Inappropriateness Males 1.4546 N.S. Accept Females 4.3636 .10 Reject 134 Table XIII (Continued) Dependent Variables Contrasts Significance Ho Cooperation Help Time Males 6.1819 N.S. Accept Females 1.0908 N.S. Accept Help Frequency Males .1818 N.S. Accept Females 1.0909 N.S. Accept No Help Time Males 6.9090 N.S. Accept Females 4.0908 N.S. Accept No Help Frequency Males 1.7273 N.S. Accept Females 2.8182 N.S. Accept Difference in Time Males 13.0909 N.S. Accept Females 5.1818 N.S. Accept Difference in Frequency Males 1.9090 N.S. Accept Females 1.7273 N.S. Accept Average Time Help Males 10.3072 N.S. Accept Females 6.4755 N.S. Accept Average Time No Help Males 2.1917 N.S. Accept Females .6001 N.S. Accept Difference in Average Time Males 12.5081 N.S. Accept Females 7.0773 N.S. Accept 135 Hypotheses 4, 4A, 4B, 4C, 4D Hypotheses 4, 4A, 4B, 4C and 4D examined the contrasts between mean scores for "comment" conditions and "peer viewing" condition on subjects' ability to (4) recall pictorial content, (4A) recall consequences of anti- social behavior, (4B) recall verbal content, (4C) perceive anti-social behavior as inappropriate, and (4D) demonstrate post-viewing cooperation. When controlling for sex of the subjects these hypotheses were tested by the Scheffe post hoc technique to determine whether contrasts between means of "comment" conditions and "peer viewing" condition were significant. Table XIV shows the contrasts of means for males and females for all depend- ent measures between "comment" and "peer viewing" conditions, as well as the levels of significance for the contrasts. The findings for each depend- ent variable are as follows. Recall of Pictorial Content (Hypothesis 4) The null hypothesis for recall of pictorial content was tested and rejected at the .05 level of significance for males and was rejected at the .01 level of significance for females. Males and females participating in conditions where comments are made are able to recall more content than those subjects viewing with a peer. Recall of Consequences of Anti-Social Behavior (Hypothesis 4A) The null hypothesis for recall of consequences of anti-social behavior was tested and rejected at the .10 level of significance for males and was rejected at the .05 level of significance for females. Males and females participating in conditions where comments are made are able to recall more consequences of anti-social behavior than those males and females viewing with a peer. 136 Recall of Verbal Content (Hypothesis 4B) The null hypotheses for recall of verbal content were tested and accepted for both males and females. There are no statistically significant differences between those males and females participating in the conditions where comments are made and those in the "peer viewing" condition. Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 4C) The null hypotheses for perceived inappropriateness of anti-social behavior were tested and accepted for both males and females. There are no statistically significant differences between those males and females par— ticipating in the conditions where comments are made and those in the "peer viewing" condition. Post-Viewing Cooperation (Hypothesis 4D) The null hypothesis for post-viewing cooperation was tested and accepted for females. The analysis of variance tests for post-viewing co- operation for males were not significant at the .10 level or beyond; there- fore, post hoc analyses would be inappropriate and the null hypotheses are accepted for males. There are no statistically significant differences between scores on post-viewing cooperation for females participating in the' condition where comments are made and those viewing with a peer. Table XIV Hypotheses 4, 4A, 4B, 4C, 4D. Controlling for Sex: Scheffe contrasts be- tween "comment" and "peer viewing" conditions and levels of significance. Dependent Variables Contrasts Significance Ho Recall of Pictorial Content Males 2.0455 .05 Reject Females 3.6455 .01 Reject Table XIV (Continued) 137 Dependant Variables Contrasts Significance Ho Recall of Consequences Males 1.6818 .10 Reject Females 1.9909 .05 Reject Recall of Verbal Males 1.4546 N.S. Accept Females 1.5182 N.S. Accept Perceived Inappropriateness Males 1.4546 N.S. Accept Females 1.5727 N.S. Accept Cooperation Help Time Males .4995 N.S Accept Females 10.2274 N.S Accept Help Frequency Males 3.0001 N.S. Accept Females 2.3637 N.S Accept No Help Time Males 2.0455 N.S Accept Females 3.2728 N.S Accept No Help Frequency Males .5455 N.S. Accept Females .8637 N.S. Accept Difference in Time Males 2.5454 N.S. Accept Females 13.4999 N.S. Accept Difference in Frequency Males -3.5455 N.S. Accept Females 1.5001 N.S. Accept Average Time Help Males 3.2469 N.S. Accept Females 2.1245 N.S. Accept Average Time No Help Males .1554 N.S Accept Females 4.4201 N.S. Accept Difference in Average Time Males 3.7586 N.S. Accept Females 6.5487 N.S. Accept 138 Hypotheses 5, 5A, 5B, 5C, 5D Hypotheses 5, 5A, 5B, 5C and 5D examined the contrasts between mean scores for "comment" conditions and "viewing alone" conditions on subjects' ability to (5) recall pictorial content, (5A) recall consequences of anti- social behavior, (SB) recall verbal content, (SC) perceive anti-social behavior as inappropriate, and (5D) demonstrate post-viewing cooperation. When controlling for sex of the subjects these hypotheses were tested by the Scheffe post hoc technique to determine whether contrasts between means of "comment" conditions and "viewing alone" condition were significant. Table XV shows the contrasts of means for males and females for all depend- ent measures between "comment" and "viewing alone" conditions, as well as the levels of significance for the contrasts. The findings for each dependent variable are as follows. Recall of Pictorial Content (Hypothesis 5) The null hypothesis for recall of pictorial content was tested and accepted for males and was rejected at the .01 level of significance for females. There is no statistically significant difference between males in the conditions where comments are made and those viewing alone. Females participating in the conditions where comments are made are able to recall more content than those females viewing alone. Recall of Consequences of Anti-Social Behavior (Hypothesis 5A) The null hypothesis for recall of consequences of anti-social behavior was tested and accepted for males and was rejected at the .01 level of significance for females. There is no statistically significant difference between those males participating in the conditions where comments are made and those viewing alone. Females participating in the conditions 139 where comments are made are able to recall more content than those females viewing alone. Recall of Verbal Content (Hypothesis 5B) The null hypothesis for recall of verbal content was tested and accepted for males and was rejected at the .01 level of significance for females. There is no statistically significant difference between those males in conditions where comments are made and those viewing alone. Fe- males participating in the conditions where comments are made are able to recall more verbal content than those females viewing alone. Perceived Inappropriateness of Anti-Social Behavior (Hypothesis 5C) The null hypotheses for perceived inappropriateness of anti-social behavior were tested and accepted for both males and females. There are no statistically significant differences between males and females in condi- tions where comments are made and those viewing alone. Post—Viewing Cooperation (Hypothesis 5D) The null hypotheses for post-viewing cooperation were tested and accepted for females. The analysis of variance tests for post-viewing cooperation for males were not significant at the .10 level or beyond; therefore, post hoc analyses would be inappropriate and the null hypotheses are accepted for males. There are no statistically significant differences between scores on post-viewing cooperation for females participating in the conditions where comments are made and those viewing alone. 140 Table XV Hypotheses 5, 5A, 5B, SC, SD. Controlling for Sex: Sheffé contrasts be- tween "comment" and "viewing alone" conditions and levels of significance. Dependent Variables Contrasts Significance Ho Recall of Pictorial Content Males 1.8646 .10 Reject Females 3.4622 .01 Reject Recall of Consequences Males .9545 N.S. Accept Females 2.4242 .01 Reject Recall of Verbal Males .8182 N.S. Accept Females 2.4015 .01 Reject Perceived Inappropriateness Males .7273 N.S. Accept Females .7727 N.S. Accept Cooperation Help Time Males 2.4999 N.S. Accept Females 10.3182 N.S. Accept Help Frequency Males 3.9092 N.S. Accept Females 1.9091 N.S. Accept No Help Time Males .8636 N.S. Accept Females .8182 N.S. Accept No Help Frequency Males 1.6364 N.S. Accept Females 1.4092 N.S. Accept Difference in Time Males 3.3636 N.S. Accept Females .1364 N.S. Accept Difference in Frequency Males -2.2728 N.S. Accept Females .5001 N.S. Accept ‘x‘llilnl‘il 1". lil