a! {7 2 z . $131.... . v a: “‘11 a... a n I. :1 1 u our «‘2’: was: a: t u: u a a t 52)) » 532.5,}.45 «‘Mru‘b‘.‘n!l‘l 7A A; 4 . ,. tn. A ‘ . i 2.. . ‘Q . $~\K{ a. a i . :33. . , i 2 .z . r I... f l ..~«« . . ‘q . s. 1. . x v x .. . . a tin“? .s.v0£2}a. ., A5); $2.50!}: 3.21:}...- lfi.‘ It“; «4" 15......) 1...”. I: v A u . a I: v C 5 ‘33:}. . fi . ifin‘s; g 118 . gifzixgtsuvtlf! TI \. A (gilt £‘.§)\Q1 .. . ¥ :2... 3% 1.13:: ‘13! .1411»... _ ‘1!“ 3: it: ‘2 Kliflisi‘ifixkb? ‘ . f mm“: ‘ .1. , a. I?3Q £04” ginguvzg‘tkzi: .fim...‘ , 1 .1 . 5 .Eu. s 1.1.: a... 2. 38,21,215‘: .331 1% x 2&3 ‘. . 2:11. rixéiigti . 3 Q. ~ .331; 4 , r V 3 ‘ mutizéshyllikn . If??? .33.}... f.§§£-wflmuno 3 a g. .%}sb§. 2 $31.32.: 1 2.1. _.. 1.5.2. .. ... l1}; st as a I» x , )1 K‘6 3 Li. ‘97 \711. J"!!! an. 1‘55 I‘ alas-2‘53? III . a . I! ‘0’: 7 igx)‘ ‘ In izilgz’fizs‘ii 7: 1‘ . iivz‘ 31:733.§§.E.;.5.§3\ {3* . i¢§§i§ . ‘ I109. §4«..sr§\\i|~<. ggéésiéxx E! hfitisusi s: A .1 s vVSi-IK at}! ‘6 323(5))... £3.15? .3.§I{t §‘\«ii\$is§% ,1 :21. 1. . ‘ {z a . 5 iii: Ag?! ! i . Q , wilt-ii. ‘ it! (1.1.31 “£211.: 11‘! 51512....111ihiflr 313...; .zli... .ll! 5.5,)???(n‘t 323.533 t\\(. \ I thkki) v.1 an: . 5.... gig-$5.; \ l ‘ a r .. m V a .1. , )Sxitxigt )‘(xzz\tl.\gt§ a .. .e . . V gégzg‘. I.» iciguuhflg :i‘iiflufifi A. ACQ.‘ :- st.<{x(.\’(< \s It‘silglq\\ . 3,7 . rlxb. 0319:61‘13 ‘ 41‘)‘. 9 L‘S‘ £339 «is £2.31..sz ~ . xx \ a\ p . U ‘ . . \ x . . ,. a...‘ x 3N1); first... 1131;}: §\ ,. i . \ i . ‘3 . \t. « \t 3 . v \ .)\>5¢41§0\4;._‘:\«\~§.111h\)&)-4§7ltf . i|\\\ ‘52 ~\a.l~\\)\b >‘.r1v(\$\ 73(25773Ki5x 3 1293 10132 cowl L L I B R A R Y Michigan State University This is to certify that the thesis entitled CHILDREN'S PERCEPTIONS OF'SEX ROLE STEREOTYPES ON SELECTED TELEVISION PROGRAMS presented by 1 ELIZABETH SMITH HOUSTON has been accepted towards fulfillment of the requirements for Ph.D. degree in Family Ecology "0 , 7 , Major professor Date December 15, 1977 0-7639 PLACE IN RETURN BOX to remove this ch TO AVOID FINES return on or before date d U6. eckout from your record. DATE DUE DATE DUE DATE DUE mag 44 DEC 0 4 1995 W'ML T ‘ xx * \\\ \ \ x \ \ \ \ \ MSU Is An Affirmative Action/Equal Opportunity Institution chMpmG—p. 1 CHILDREN'S PERCEPTIONS OF SEX ROLE STEREOTYPES ON SELECTED TELEVISION PROGRAMS By Elizabeth Smith Houston A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family Ecology 1978 ‘llld'llttllll o ABSTRACT CHILDREN'S PERCEPTIONS OF SEX ROLE STEREOTYPES ON SELECTED TELEVISION PROGRAMS By Elizabeth Smith Houston This study was designed to compare responses of Black and White children who watched selected television tapes of stereo— typed and counter-stereotyped roles performed by black and white characters with the responses of children who did not view any tapes. 1 The three hundred eighteen subjects used in this study were from kindergarten, second and fifth grade from three elemen- tary schools in Lansing, Michigan. The study consisted of two experimental groups and one control group. Experimental group I consisted of 106 subjects, experimental group II had l08, and the control group numbered l04. The design of this study was a post-test-only-control- group design. The data were collected by the use of question- naires. The four-way analysis of variance (ANOVA) fixed effect model was used to test the hypotheses on stereotype, counter- stereotype, and racial identification responses. The significance level was set at .05. Elizabeth Smith Houston The data showed children in fifth grade who viewed counter- stereotyped roles being more influenced by the tape than second grade and kindergarten children. Second grade children in the counter-stereotype group were more stereotyped in their responses than second grade children who watched the stereotype tape. No significant differences were found in the stereotyped responses of boys and girls within this study. A comparison was made of stereotyped responses of older and younger children. The kindergarten children in the stereotype and control groups were more stereotyped than second and fifth graders. The fifth grade children provided more counter-stereotyped responses than younger children in kindergarten and second grade. A comparison of responses by children in the stereotype group with those in the control group showed no difference in the responses of kindergarten children in the stereotype or control groups. The second and fifth grade children in the stereotype group were more stereotyped than those second and fifth grade children in the control group. Both second and fifth grade children in the control group were more counter-stereotyped in their responSes than those in the stereotype group. Data on racial identification showed no difference in the identification scores for White males and females. Black males had the lowest identification score for race and sex. Black females scored lower than White females also. The race by grade interaction showed Black children at the kindergarten and second grade levels with lower identification Elizabeth Smith Houston scores than White children at this level. Black children in kindergarten had the lowest identification score while White children in kindergarten had the highest identification scores. There were no differences in the identification scores for Black and White children at the fifth grade level. The data provided additional information on occupational and household roles which was significant though not related to the hypotheses tested in the study. Results from occupational roles showed no difference in the responses of children in the stereotype and counter-stereotype groups. Children in these groups were more stereotyped than children in the control group. ' Both Black and White kindergarten children were more stereotyped in their responses on occupational roles than second and fifth grade children. A Data from the analysis on household roles showed females more stereotyped in their responses than males. Fifth grade children in the stereotype group were more stereotyped than kindergarten or second grade children. Those second grade children in the counter-stereotype group were more stereotyped than second grade children in the stereotype or control groups. There were no differences in the responses of kindergarten child- ren in all groups. The results from this study provided data to indicate that watching selected television portrayals of males and females per- forming sex stereotyped and counter-stereotyped roles may have Elizabeth Smith Houston some influence on children's responses to sex stereotypes and racial identification. The degree of influence was determined by sex, race, grade, and treatment group of the children. DEDICATION In memory of Mama (my grandmother), who expired shortly before this degree was completed at the honorable age of 80. To Popa (my grandfather) age 86, for continued good health and an active life. PaSt racial and social inequalities in the south prohibited Mama and Popa from achieving the educational level desired, but through this degree a future dream has become a reality for them. To my mother, for understanding and moral support. To my husband for inspiration, patience and en- couragement. To my two lovely children. ii Ill ACKNOWLEDGMENTS Special thanks go to my major advisor, Dr. Eileen Earhart, for the assistance, support, and generosity of her time throughout the duration of this research. I am also appreciative for the support from other committee members, Dr. Norma Bobbitt, Dr. Jane Dyer, and Dr. Louis Romano. Thanks also to Dr. Verna Hildebrand for her assistance during the initial stages of the work. My appreciation is extended to the principals, teachers and students from the three elementary schools in Lansing, Michigan, for their participation in this project. Special thanks to Gordon Earhart, Dr. Alice Whiren, Anne Soderman, Lillian Phenice, Pat Fuerstenau, and Kalina Gonska, who gave generously of their time to assist in collecting the data. Thanks also to Dr. Bradley Greenberg, Dr. Charles Atkin, and the CASTLE staff members in the Department of Communication at Michigan State University. The child care assistance provided by my sister and her husband, Florine and Emmitt Ford, was appreciated. My daughter Angela, eight, and son David III, three,adjusted well to the unusual circumstances of this period, for which I am grateful. My deepest thanks and appreciation are due my wonderful husband David. Without his understanding, support and encourage- ment, this degree would not be completed. iii LIST OF LIST OF LIST OF CHAPTER I. II. III. TABLE OF CONTENTS TABLES ..................... FIGURES .................... APPENDICES ................... INTRODUCTION .................. Statement of the Problem ............ Need for the Study ............... Purpose of the Study .............. Hypotheses ................... Objectives ................... Assumptions .................. Conceptual Definitions ............. Conceptual Framework .............. Overview .................... REVIEW OF LITERATURE .............. Sex Role Development in Children ........ Social Learning Theory ............. Modeling and Observation ........... Imitation .................. Identification ................ Cognitive Development Theory .......... Television as a Source of Learning ....... Occupational Role Portrayals on Television . . . Sex Role Portrayals on Television ...... Racial Portrayals on Television ....... Summary ......... - ........... THE RESEARCH DESIGN .............. Description of Sample and Population ...... Procedures ................... Design of Study ................ Testable Hypotheses .............. Analysis of Data ................ Summary .................... Page vi —-l £0 CDVOSOSU'IU'I-Dw-d IV. ANALYSIS OF RESULTS ................ 55 Hypothesis l ................... 55 Hypothesis 2 ................... 59 Hypothesis 3 ................... 50 Hypothesis 4 ................... 51 Hypotheses 5 and 6 ................ 54 Additional Data .................. 74 Occupational Roles ............... 74 Household Roles ................. 77 Summary ...................... 82 V. SUMMARY, DISCUSSION AND IMPLICATIONS ....... 83 Summary of Findings ................ 33 Discussion of Findings .............. 89 Viewing Stereotyped and Counter-Stereotyped Portrayals ................... 89 Stereotyped Responses of Boys and Girls ...... 90 Stereotyped Responses of Older and Younger Children .................... 91 Comparison of Stereotype Group and Control Group .................. 92 Racial Identification for Black and White Children .................... 93 Occupational Roles ................ 96 Household Roles .................. 97 Limitations of the Study ..... . ....... 98 Implications for Future Research ......... 99 Implications for Practical Use ......... lOO APPENDICES ........................ 10] REFERENCES ........................ 117 Table 0») 0000000000 0501-wa LIST OF TABLES Race and Sex of All Subjects ....... Race and Grade of_Kindergarten Subjects Race and Grade of Second Grade Subjects Race and Grade of Fifth Grade Subjects . . Sex and Race of Stereotype Group (Group I) ..... Sex and Race of Counter-Stereotype Group (Group II) ....... p ......... Sex and Race of Control Group (Group III) Cronbach Coefficient Reliability Analysis for Racial Identification Questions . . Cronbach Coefficient Reliability Analysis for Racial Identification Questions . . Design of Study and Cell Size ...... Analysis of Variance for Main Effects in Comparison of the Stereotype and Counter-Stereotype Groups ....... Analysis of Variance Showing Two-Way Interactions for Stereotype-Counter Stereotype Groups ........... Grade Means for Groups I and II ..... Analysis of Variance for Stereotype and Counter-Stereotype Groups ....... Means for Grade by Group Interaction for All Grade Levels and Groups ............ vi 00000 Page 4T 4T 42 42 44 44 45 47 47 50 56 57 57 59 Table 4.6 Grade Means for Group I (Stereotype) and Group III (Control) ............. Analysis of Variance Showing Main Effect for Race Identification ............. Analysis of Variance Showing Two-Way Interaction for Race Identification ............. Means for Sex by Race Interaction for Race Identification ............. Means for Race by Grade Interaction for Race Identification ............. Means for Grade by Group Interaction for Race Identification ............. Analysis of Variance Showing Three-Way Interaction for Race Identification ....... Means for Sex by Race by Grade Three-Way Interaction for Race Identification ....... Analysis of Variance Showing Main Effects for Occupational Roles ............. Analysis of Variance Showing Two-Way Interaction for Occupational Roles ............. Means for Race by Grade Two-Way Interaction for Occupational Roles ............. Analysis of Variance Showing Main Effects for Household Roles ............... Analysis of Variance Shbwing Grade by Group Two-Way Interactions for Household Roles Means for Grade by Group Two-Way Interaction for Household Roles ............... vii Page 62 65 65 66 68 69 7O 71 75 76 76 78 79 79 Figure h (A) (0'00 boom .7a .7b .10 LIST OF FIGURES Question in the Area of Household Roles ..... Question in Area of Occupational Roles ...... Question in Area of General Roles ........ Post-Test-Only Control Group Design ....... Graph of Group I and II Grade by Group Interaction .................. Graph of Group by Grade Interaction for All Grade Levels and Groups ........ Graph of Means for all Grade Levels in Group I and Group III ............. Graph of Means Showing Sex by Race Interaction . . Graphed Means for Race by Grade Interaction Graphed Means for Grade by Group Interaction for Race Identification ............ Graphed Means for Sex by Race by Grade Three-Way Interaction for Race Identification, Level A, Males ................. Graphed Means for Sex by Race by Grade Three-Way Interaction for Race Identification, Level B, Females ................ Graph of Means for Race by Grade Two-Way Interaction for Occupational Roles ....... Graph for Grade by Group Two-Way Interaction for Household Roles .............. Summary of Hypotheses Tested and Decision Rules ................. viii Page 46 46 49 58 62 63 66 68 69 72 73 77 80 82 LIST OF APPENDICES Appendix A. Role Portrayals on Stimulus Tape for Stereotype Group Role Portrayals on Stimulus Tape for Counter-Stereotype Group ............ Questionnaire ............. ' ....... Stereotype (S) - Counter-Stereotype (CS) Scoring Scale ................... Racial Identification Scoring Scale ......... Questions on Occupational Roles ........... Questions on Household Roles ............. ix Page 101 105 113 116 CHAPTER I INTRODUCTION Statement of the Problem Today, television has pervaded the homes of most American children. Research by DeFleur and DeFleur (1967), Culley, Lazer and Atkin (1976), and Unger (1977) reveals that in most American families there are at least one television set, and that from pre-school through elementary school, one-fourth to one-third of a child's waking hours are spent watching television. DeFleur (1964) and Unger (1970) concluded that children spend more time watching television than doing anything else except sleeping. Evidence by Unger also pointed out that television is not only influencing the life of the average American child and family, it is shaping it. Among all forms of mass media, television is regarded as the leading source of learning for children. However, learning from television in most cases is incidental for small children. According to DeFleur (1964) and DeFleur and DeFleur (1967), children watch television for entertainment, but they absorb incidental information while they are being entertained. Such information is then stored, without the child realizing it, for later use. The large amount of time spent by children observing various characters and identifying with many of them indicates that there is a need for an analysis of the effects of tele- vision role stereotyping on this group of youngsters. This study was developed to examine the various occupational roles and sex stereotypes portrayed on television and their effects on children watching them. Studies indicate that television networks have not kept , up with the changing roles of either females or minorities in the labor force. Dobkin (1977) reported on a survey conducted in March, 1975, which revealed that wives' earnings averaged 26.3 percent of the family's income. He concluded that "the concept of a family with the husband as the only breadwinner, the wife is a homemaker out of the labor force and there are children, may be a useful one for illustrative purposes, but it does not represent the typical family of the mid 1970's." The labor report by Dobkin (1977) reflected the growing number of women in the work force. This report indicated both husband and wife were working in nearly half of the nation's marriages. Television fails to reflect not only the quantity of women in the work force, but also the diversity of their work roles. Females are still shown in traditional, stereotyped roles. Long and Simon (1974) found that women in television were portrayed as being dependent and performing expressive and semi-emotional roles within the family. All of the women in their survey were housewives, secretaries or quasi-secretaries, and none were portrayed as doctors, professionals or executives. DeFleur (1964) concluded that the world of work on television is a man's world and the distribution of the sexes is more unequal on television than in the real world. The DeFleur study of the Indiana labor force indicated 83.9 percent of males and 16.1 percent of females shown on television were workers. However, in the actual Indiana labor force, 68.9 percent of the workers were males and 31.1 percent were females. These findings indi- cate the difficulty for young children to obtaihfiaccurate and up-to-date information about the distribution of occupations by watching television. Children also learn by imitating models shown on tele- vision programs in occupational and household roles. If this imitative behavior is rewarded, children will repeat such behavior and accept it as appropriate. Need for the Study American children today are exposed to a wider variety of occupations from their television viewing than from any other source. However, television portrays occupational roles for men and women in mostly stereotyped situations (DeFleur, 1967; Long and Simon, 1974; Miller and Reeves, 1976). A The male child in our present society is expected to select an occupation, but the female also has the potential to become an integral part of the labor force. Although television is an important source of information about roles and occupations for men and women, there is limited empirical evidence on its impact (DeFleur, 1967). Data also appear to be limited on how children perceive the roles of men and women on television programs. There is limited evidence on the influence certain television programs will have on children's perceptions of the appropriate and inappropriate sex roles for males and females. The results from this study provide empirical data on the effects of portrayals of stereotyped and counter-stereotyped roles on children watching selected television programs. This study also examines the perceptions of roles by Black and White children viewing black and white characters to determine if a difference exists due to the race of the viewers. Purpose of the Study The purpose of the study was to determine the difference in Black and White children's responses as a result of viewing stereo- typed and counter-stereotyped role portrayals on television programs with Black andldhite male and female characters.1 Responses of two experimental groups viewing television programs were compared with the control group who did not view any tape. Children in kinder- garten, second and fifth grades were used to determine the differ- ence in responses based on sex, race and grade level. The overall broad objective of this study was to provide basic knowledge on the 1This study was a part of a larger, on-going project in the Department of Communications at Michigan State University. Project title, "Parental Mediation of Children's Social Learning from Tele- vision." Grant #9D-C-635, funded by the Office of Child Develop- ment, Washington, D.C. Project co-directors are Dr. Bradley Green- berg and Dr. Charles Atkin. impact of te and househol Hypothesis 1. Hypothesis 2. Hypothesis 3. Hypothesis 4. Hypothesis 5. Hypothesis 6. In a trayed on te tion. There 1. levision on children's attitudes toward occupational d roles for Black and White males and females. Hypotheses Children who view counter-stereotyped portrayals will give more counter-stereotyped responses than children who view stereotyped portrayals. Boys will give more stereotyped responses than girls. (Older children will give more stereotyped responses \than younger children. Children who view stereotyped roles will give stronger (more) stereotyped responses than children in the control group. Black children will give responses denoting identi- fication with black characters, rather than with white characters. White children will give more responses denoting identification with white characters, rather than with black characters. Objectives ddition to the child's perception of roles being por- 1evision, other variables will provide valuable informa- fore, the following objectives have been formulated: To determine if there is a difference in the attitudes of children viewing stereotyped and counter-stereotyped portrayals. To determine if there is a difference between responses of boys and girls. To determine if there is a difference between responses of older children and younger children. To determine if a relationship exists between the two experimental groups and the control group. 5. To determine the degree of identification of Black children with black characters in television programs. 6. To determine the degree of identification of White children with white characters in television programs. Assumptions The assumptions underlying the study are the following: l. All children involved in the study can identify a work role. ‘ 2. There is at least one television set in each child's home. 3. All children involved have acquired a sex role identity based on their biological sex. 4. Identification with characters will vary according to sex and race. Conceptual Definitions The following terms are used throughout the study. Stereotype refers to a conception or image with special meaning and held commonly by members of a group, race or sex. Counter-stereotype refers to the reverse direction, that is, the opposite of stereotype. Imitate means to follow or endeavor to follow in action or manner, in some cases to mimic or copy. Modeling refers to a standard or example for imitation or comparison. Occupational_portrayal refers to the appearance of a person on the television screen performing some kind of recognizable occupational duty (DeFleur, l964). . Incidental learning refers to learning that is unplanned, the retention of any content of a television show in the absence of an induced set or intent to learn specific behaviors or atti- tudes (Bandura, 1963). i: Identification means the viewer puts himself in the place of a character and momentarily feels that what is happening to that character is happening to him (Maccoby and Wilson, l957). , Sex role means a behavior that an individual assumes or is assigned based on an understanding of maleness or femaleness. Conceptual Framework The theoretical foundation for this study is based on the social learning theory by Bandura (197l; 1977). This theory examines observational learning by modeling, imitation and identi- fication. Television programs within this study provide children the opportunity to learn about occupational roles, household roles, sex roles and racial roles. There is evidence that a relationship exists between what children watch on television programs and their perceptions of appropriate roles for males and females. Learning by children is not limited to observing models within the family unit, it is also influenced by observing and imitating and in many cases identifying with television characters, peers, teachers and other adults within the neighborhood. Television programs have been so influential that much of what children learn comes from observing some model on television through incidental learning. From an ecological perspective, television is regarded as part of the near environment. It is also a means for allowing out- side communication into homes through its programming. The family system has been found to be open to this form of programming and com- munication. This is evident by the number of hours children spend watching television. According to Unger (1977), this means of com— . munication is not only influencing children's lives,it is shaping them. The degree to which a family ecosystem is open or closed to television programs depends on parents' perception of what children are learning from television. If parents believe what children are learning is positive, the system will remain open to this form of media. 0n the contrary, when parents have overall negative feelings concerning television programs and their effects on children's learning, the family will either close its system to television viewing or regulate and control viewing for children. Overview Supportive research on social role learning as it applies to children's television and the type of portrayals by males and females will be reviewed in the second chapter of this study. The design of the study, sample and instrument will be discussed in Chapter III, and the analysis of the results in Chapter IV. A discussion of the summary, conclusions and implications of the study will be presented in Chapter V with recommendations for further research. CHAPTER II REVIEW OF LITERATURE The literature pertinent to this study is reviewed in four areas. The development of sex roles is reviewed to lay the foundation for sex role behavior by children throughout child- hood and throughout their adult lives. Social learning theory is reviewed to show some important ways children learn through modeling, imitation and identification. Cognitive development theory indicates learning that is based on abstract learning, which occurs in stages according to the maturation of children. And finally, television portrayal of males and females is reviewed to show how it influences children's thinking about roles for men and women. Sex Role Development in Children The differentiation of sex roles appears to start immediately after birth when it is announced that the baby is a boy or a girl. With this announcement babies are named appro- priately for their sex and from then on are referred to as a boy or a girl. They are usually dressed in pink or blue and placed in a nursery decorated in the same pink or blue colors. Embedded in the decoration will be animals for boys and dainty 10 little dolls with umbrellas for girls to symbolize the femininity and masculinity of the sexes. Thereafter, indoctrination into masculinity or femininity is promulgated by children's clothes and hair styles and by selecting sex appropriate play materials and recreational activities (Bandura, 197l; Mussen, l97l). Researchers agree that the process of sex role develop- ment begins very early in life. They also agree that these years are very crucial in the development of sex roles, and once established, the child's sex role is permanently fixed; that is, it is irreversible. Mussen (l97l) indicated the first two years as the most critical period of sex typing. Brown (1957) and Hartup and Zook (l960) agree that by the age of three or four boys express clear-cut preferences for masculine activities, toys and objects. After this age, sex typed behavior becomes more firmly embedded with age. * 'Mussen(l97l), in his study of sextyping, suggests that boys are more likely to show earlier and sharper awareness of sex appropriate behavior and interest. He further maintains that girls between the ages of three and ten will show rather strong preferences for masculine games, activities, and objects. This latter finding is supported by Brown (l957), Kagan (1964) and Hartup and look (l960). At this age very few boys prefer feminine activities,but in this age range younger boys will have stronger preferences for "masculine" than for "feminine" toys. Kagan (1964) maintains there is a clear-cut difference in sex based on children's preferences in games and toys. II In his study he noted that boys chose objects related to sports, machines, aggression, speed and power roles. Girls, on the other hand, selected games and objects associated with the kitchen and home, babies, personal attractiveness and fantasy roles which have a subordinate relation to-a male, such as nurse and secretary. Kagan also described knives, boats, planes, trucks and cement mixers as being regarded by children as mas- culine, and dolls, cribs, dishes and nurses' equipment as feminine. Brown (l956) devised a scale to measure children's sex role preference, referred to as the It Scale for Children (ITSC). The test was conducted using 78 males and 68 females in kinder- garten from a middle class section of Denver. The scale con- sisted of 36 picture cards depicting various objects, figures and activities commonly associated with masculine or feminine roles. In doing this type of picture preference test using the "It" figure, the author assumed the child would project himself into the it-figure based on his or her sex role preference and also attribute to it the sex role preference of the child. The findings from Brown's study indicated significant differences occurring between boys and girls, which suggests the existence of definite dichotomous sex role patterns in young children. This study also revealed the fact that boys show sig- nificantly greater preferences for masculine roles than girls show for feminine roles. Girls between the ages of three and ten years of age show a strong preference for masculine games, 12 activities, and objects; however, boys did not show an equal preference for feminine activities during this period. Five- year-old boys showed a stronger stereotype preference for mas- culine toys than girls showed for feminine toys. Hartup and Zook (1960) extended Brown's work with the It Scale for Children. According to these authors, preferences of the kind measured by the It Scale are assumed to be acquired before the child is of kindergarten age. This study examined sex role preferences of three- and four-year-old children. The results of the test by Hartup and Zook corroborated the findings by Brown (1956). The three—year-old boys pre- ferred the stronger stereotyped masculine roles more than the girls preferred the stereotyped feminine roles. The four—year- old boys also showed stronger masculine preferences than girls showed feminine preferences. Minuchin (1969) also supported these findings. Maccoby and Jacklin (1974) reported that a study by Sears and his colleagues (1965) attempted to discover whether one sex was more clearly sex typed than the other. The results revealed no sex differences in four-year-old boys and girls in their willingness to adopt an opposite sex role. Maccoby and Jacklin (1974) also reported that the results of a study by Hartup and Moore (1963) did find a sex difference in willingness to engage in cross-sex activities. These authors offered nursery school age children two toys, one an unattractive sex neutral toy and the other an attactive toy suitable for the 13 other sex. The results of this test showed boys were more likely to avoid the sex-inappropriate toy than were girls. The boys avoided the feminine toys more when the experimenter was present; however, the girls showed interest in boys' toys as much when the experimenter was present as in his absence. Mussen (1971) and Kagan (1964) noted social class differences in the selection of games and toys by children. Boys and girls of the lower class became aware of their appro- priate sex role patterns much earlier than middle class children and their preferences appeared to be much more stereotyped than middle class children's. Rabban (1950) noted the difference in sex typing between lower and middle class was greatest among girls. Lower class girls made definite sex-appropriate choices by the age of six, whereas middle class girls had not reached this level at the age of eight. These findings suggest there is a differentiation between the adult masculine and feminine role in the lower class families and that lower class mothers are much more encouraging of sex-appropriate sex typing. Masingale (1971) studied the effects of the father's absence and parental role play behavior of children using forty Black five-year-olds enrolled in a Head Start Center in Louisiana. All subjects involved in this study were from low socioeconomic families. The doll play parental role choice indicated these Black children made the appropriate sex role choices regardless of the father's presence or absence. 14 These findings further pointed out that females initiated almost twice as many doll interactions as the males, that father- present children initiated almost twice as many doll interactions during their doll play, and that females were twice as supportive in their doll play as were the males. The American culture has different standards with regard to sex roles for males and females. One of the standards, as reported by Kagan (1964), involves aggression. This sex role standard discourages verbal and physical aggression among girls and women,but it encourages boys and men to express ag- gression. when attacked, threatened or dominated by another male. This view is supported by Bandura (1962), Bandura, Ross and Ross (1961), and Maccoby and Wilson (1957). The second standard of American culture involving sex typed behavior is dependency, passivity and conformity. Girls are encouraged to express these behaviors while boys and men are discouraged and pressured to suppress such behaviors. Kagan (1964) investigated sex typing and found nurturant behaviors generally more appropriate for females than for males. Bennett and Cohn (1959) revealed children view women and females as more nurturant than men and that adult women see themselves as more nurturant than men or males. These findings suggest a complete circle with both children and adults expect- ing and receiving more dependence, passivity and nurturing from females and more aggression from males. 15 Social Learning_Theory The social learning theory is somewhat traditional, however, it is widely accepted by researchers. This theory is based on reward for appropriate sex role and punishment for inappropriate sex role choices. Mussen (1971) has indicated that sex-appropriate behaviors are reinforced or rewarded by parents and are, therefore, repeated by the child. On the other hand, a child displaying sex-inappropriate behavior increases chances of punishment which will diminish the strength and frequency of such behavior, and through punishment the behavior should be extinguished. The aspect of social learning theory within this study is based on the premise,as indicated by Bandura (1971, 1977), that children learn by identification, imitation and modeling. Bandura (l97l) maintained that in the social learning system new patterns of behavior can be acquired through either direct experience or by observing the behavior of others. Modeling and Observation Bandura (1977) has pointed out that learning could be hazardous and difficult if people always had to rely on their own actions and instinct. According to Bandura (1977), most behaviors are learned through the observation of some modeling. The social learning theory maintains that through modeling, learning takes place in the form of informative functions. When one observes a model, symbolic representations 16 of the activities performed by the model serve as guides for the appropriate performances by the observer. Mischel (1974) inves- tigated the acquisition of sex typed behavior and found observational learning, either live or in films, television or books, to be the first step in the acquisition of sex typed behavior. Bandura (1977) suggested four processes involved in observational learning: (1) attentional processes; (2) reten— tion processes; (3) motor reproduction processes; and (4) moti- vational processes. During the attention process, Bandura (l97l) maintained, learning by observation is minimal unless people attend to and accurately perceive the significant features of the modeled behavior. The amount of attention placed on modeled behavior determines what is observed and the influence the modeling has on the observer. Observational experiences in amount and type ware generally regulated by the observer's characteristics, the feature of the modeled activities and the structural arrangement of human interactions. The effectiveness of modeled conduct also varies within individuals and social groups. Bandura suggested that attention to modeled behavior is regulated by the model's interpersonal attraction. Attention is sought out for models with pleasing qualities. On the other hand, those models who lack these pleasing qualities are often ignored or rejected. It has been demonstrated by watching television that -modeling which is rewarding can hold children's and adults' l7 attention for extended periods of time. Bandura, Gruser and Menlove (1966) found that the models presented on television were so effective in getting observers' attention that viewers learned a large part of what they saw without requiring any special incentives. This process is similar to what DeFleur (1964) terms incidental learning. The second process in observational learning described by Bandura is the retention process. Here the emphasis is on remembering what is observed by a model. The memory of the modeled behavior must be retained in symbolic form for permanent memory. Bandura (1977) maintained that one way of retaining memory is by repeated exposure of modeled performances. Repeated exposure by visual means is very important in observa- tional learning during early childhood when verbal skills are not fully developed. The motor reproduction process is the third process of observational learning examined by Bandura. This process allows the observer to convert symbolic representations into actions that are appropriate. In this process the observer's actions are based on informative feedback. The fourth process in observational learning in the social learning theory is motivational process. This process focuses on those observed behaviors that seem to be effective for others and are favored over behaviors that might have nega- tive consequences. Hicks (1971) investigated girls' attitudes 18 toward modeled behaviors and imitative performance using girls between seven and nine years of age. The results of this study indicated significantly more imitative performance was produced if the child's attitude toward the modeled behavior was positive rather than negative. Imitation Another aspect of the social learning theory is imitation. Mussen, et al..(l97l) refer to imitation as observational learn- ing or modeling in human development. There is not enough empir- ical evidence to determine how much of sex typed responses develop in children as a result of imitating models. However, researchers agree that learning by imitating models starts very early in childhood. When a child's imitative behavior is rewarded frequently, the reward becomes an incentive to continue the imita- tive process. Research by Bandura and others indicated that sex typed responses in children develop through imitation of like sex parents' behavior. Bandura (1971) concluded that powerful and nurturant models were more likely to be imitated than models that did not possess these qualities. Maccoby and Jacklin (1974) investigated sex typing and role modeling and found since parents were highly available, nurturant and powerful, they are the models children would most likely copy in acquiring sex typed behavior, especially in the early preschool years. 19 Bandura (1971) pointed out that children are exposed to many models (teachers, prestigious television characters, other adults in the neighborhood, and peers) and they do not rely solely on parents as models during periods of development. The findings by Lane (1968) refute Bandura's (l97l) notion as previously indicated. Lane interviewed fifteen boys and fifteen girls in the first and second grades of an elemen- tary parochial school to assess the influence of parents' covert and overt sex role identification on the sex typing of the child. The results of Lane's study showed boys and girls more appro- priately sex typed as the same sex parent. These findings suggest that parents' attitudes about appropriate sex behavior play an important role in developing sex appropriate behaviors by the child. Research conducted by Bandura, Ross and Ross (1963) demonstrates that cross-sex imitation does occur when one model has more control than the Opposite sex. Cook and Smothergill (1973) also examined the relationship of race and sex of model and observer in imitative performance and knowledge using 154 black and white preschoolers with an average age of 55.9 months attending a Head Start program. The results of this investiga- tion indicated white models were imitated more than black models regardless of race or sex of the observer. Imitating white models was due to the fact that lflfites appear to have more power, prestige and influence in the society than Blacks; therefore, children prefer to imitate Whites as persons they would want to be like. 20 Many-Blacks model their behavior after Whites because they view Whites as being more competent (Walters, 1977). Walters further pointed out that Whites on television, in particular young White males, always get the competent roles, thus increasing young Blacks' desire to imitate White models. Maccoby and Jacklin (1974) found that same sex models would be imitated more than the opposite sex because children tend to imitate models that are perceived similar to themselves. Lane (1968) found some indication that the same sex parent appeared to be more important for the child's identification than the opposite sex parent. This indication did not prove to be significant; however, the concept of modeling as a means of learning sex appropriate behavior can be applied in this instance. Maccoby and Jacklin (1974) offer the following statement by Mischel (1970) in support of their argument for same sex modeling and imitation. From the viewpoint of social learning theory, the greater attentiveness to same sex models, especially when they are displaying appropriate sex typed behavior, probably reflects that people generally are reinforced throughout their histories more for learning the sex typed behavior of same sex models than those of cross- sex models. It certainly seems likely that children are much more frequently rewarded for watching and imitating same sex models rather than crossesex models, especially when the models display sex typed behaviors. (p. 286) Kohlberg (l97l) maintains there was a tendency to imi- tate self like objects as a result of the child's development and liking for the same sex parent. 21 Identification It is difficult to differentiate between the concepts of imitation and identification. Bandura and Walters (1963) argue that there is no difference in the two concepts. They further point out that in experimental psychology observational learning is referred to as imitation, whereas in theories of personality it is labeled identification. Within the context of the social learning theory identi- fication denotes a special kind of imitation. This means the observer will duplicate the model's pattern of behavior without any special training or reward because of the relationship between the observer and the model. According to Kagan (1958), a child's identification is strengthened by believing that he/she possesses some of the model's characteristics and feel- ings. The following paragraph serves to support the view held by Kagan. If a six year old boy is identified with his father, he necessarily regards himself as possessing some of his father's characteristics, one of which is maleness or masculinity. Moreover, if a child is identified with a model, he will behave, to some extent, as if events that occur to the model are occurring to him. If a child is identified with his father, he shares vicariously in the latter's vic- tories and defeats; in his happiness and in his sor- row; in his strengths and in his weaknesses. (Kagan, 1964, p. 146) Maccoby and Wilson (1957) investigated whether viewers identify themselves with movie characters most like 22 themselves when they are given the opportunity to select the protagonist. The study utilized two characters, a lower class boy and an upper class boy, in a movie with both portraying leading characters. The results of this aspect of the investi- gation showed regardless of the social class level of boys, they tended to identify with the middle class boy in the movie if they had high level of occupational aspirations. These findings suggest that identification with screen characters by boys is more related to the social class level that he aspires to rather than the current level of his family. The second part of this investigation also predicted that boy and girl viewers would choose to identify with like sex characters. The results indicated boys remembered more of the "boy alone" and "boy stimulus" content. Boys also remembered the aggressive acts better than girls if the boy was the aggres- sive hero. Girls, on the other hand, remembered more "girl alone" and "girl stimulus" materials than boys. Girls' remem— brances of aggressive acts were not significant even when the girl was the aggressive hero. Both of these findings by Maccoby and Wilson suggest the memory for movie materials was influenced by identification with the character within the movie. Feilitzen and Linne (1975) used Scandinavian subjects to investigate identification with television characters. These re- searchers found that individual characteristics of the child are crucial factors in determining with whom the child will identify. 23 In this study boys identified with figures like themselves; that is, boys identified with boys, and lower class children identified with the media figures from the lower social class. These latter findings by the Scandinavian researchers are in opposition to the findings by Maccoby and Wilson which showed that lower class children identified with characters for which they had occupational aspirations. Cognitive Development Theory According to Kohlberg (1966) a child's sex typing is determined by the child's own perception of the physical and social world. This perception usually includes the child's sense of self and his own sex role and comes with changes in age and mode of thinking. He maintains that 1earning,and particularly observational learning, does play an important role in the acquisition of sex role, but the most important factor is the child's cognitive activities, that is, the child's selecting and structuring perceptions with knowledge and understanding. Kohlberg further maintains that a child's sex role is not a' biological instinct, but is rooted in his own concept of physi- cal surroundings. According to Kohlberg's theory, the child's sex role concept is established at about five or six years of age when there is an understanding of conservation. During this stage the child learns physical properties of mass, numbers, and weight. The child's cognitive development is then 24 increased and he learns stereotypes of masculine and feminine behaviors. The following paragraph illustrates Kohlberg's theory of cognitive development as it relates to sex role development. A boy's preference for same-sex peers is established before his preference for same-sex parent figures. The cognitive developmental theory suggests two reasons for this discrepancy. The first is that the boy's classifi- cation of adult males in the common category "we males" is a more cognitively advanced achievement, and therefore, comes later than his classification of other boys in that category. It is not until about age five or six, when the child begins to sort objects predominantly on the basis of similar attributes, that he forms groupings which include same-sex figures of diverse ages. The second consideration is that the boy's affectional tie to his mother is deep, and it takes time before the boy's self- conceptual or sex-role identity consideration can lead him to subordinate it to the development of a tie to the father. (Maccoby and Jacklin, 1974, p. 291) The stage between five and six is the time a child learns that males are attive, dominant, powerful and aggressive and that females are more nurturant. Kohlberg points out that this beha- vior is not derived from parental behavior but is a result of universally perceived sex differences in male and female body structure and capacities. Kohlberg's theory of sex role development is based in part on Piaget's theory of cognitive development. Piaget's theory proposes that children learn in stages and that age alone does not determine when a child is ready to learn new concepts. It seems apparent that Kohlberg's theory of cognitive development in the acquisition of sex roles is in conflict with the social learning theory proposed by Bandura. Both theories 25 indicate learning by modeling, imitation and identification, but Bandura and Kohlberg maintain these are acquired at different stages of a child's life, and for different reasons. Television as a Source of Learning Television is teaching children at all ages and develop- mental levels. This is illustrated by hearing very young child- ren, even toddlers, repeat what they have seen on the screen. Cully, Lazer and Atkin (1976) maintain that preschoolers are able to recognize certain cues to indicate when a favorite pro- gram is about to come on television long before they learn to tell time. If one follows Kohlberg's theory of cognitive develop- ment, it might be correct to assume that a child aged three or four will not understand what he sees portrayed on television. On the other hand, Bandura's theory of social learning suggests repetition as a means for learning, which means if a child views certain roles several times, the child will accept these roles as appropriate and store that information for later use. According to Bandura, learning has taken place when this happens. Children are exposed to many subjects today by means of television, films, and movies, and in the process incidental learning takes place (Hale, Miller, Stevenson, 1968; DeFleur, 1964; DeFleur and DeFleur, 1967). An investigation by Hale, Miller and Stevenson (1968) showed a curvilinear relation- ship would be found in a situation in which there was no 26 obvious reason for learning to take place. These investigators used 444 middle class children in grades three through seven from Minneapolis. The subjects were shown a film in guise of a reward for participating in a larger study. The film consisted of an eight minute dramatic skit in sound and color with four characters. After subjects completed the viewing of the film they were handed a booklet to answer questions. The retention of any part of the film content constituted incidental learning according to the investigators. T The results of the above study provided evidence of a curvilinear relationship between age and incidental learning. There was an increase in the amount of incidental learning between grades three and six and a drop in learning at grade seven. These findings seem to suggest that between grades three through six children tend to focus their attention on a wide variety of environmental features and are eager to learn about all of them. On the contrary, for children in grade seven, the decline in incidental learning appears to suggest that by this age children are able to disregard anything that is not essential and focus their attention on specific features and situations which enhance their learning. Occppational Role Portrayal on Television Television appears to be lacking in its portrayals of women and minorities in occupational roles. Men are dominant in 27 every aspect of occupational portrayal. Statistics reveal that 51 percent of the United States population are women and out of this number, 46.1 percent of the women with children under age eighteen are employed outside the home (Dobkin, 1977). It can be concluded from these figures that men are over-represented and women and minorities under-represented in television pro- gramming. A survey by Simmons, Greenberg and Atkin (1977) found more than three-fourths of all characters on Saturday morning television were males. In cartoon programs, five out of every six human characters were male. Stein and Friedrich (1975) indicated 70 to 75 percent of the leading characters were males on prime time shows which they surveyed. In this same report, males were engaged in many prestigious occupations, while females had no occupations outside the home. The investigation by Simmons, Greenberg and Atkin (1977) found women having less status and lower income jobs on television. When women were shown working, they were usually in a stereotyped occupation such as nursing. Miller and Reeves (1976) analyzed all prime time tele- vision characters with speaking parts for one week in November, 1974. Their conclusion supports the distorted view of occupa- tions on television. Their data also showed males dominating prestigious professional occupations and highly active jobs such as law officers. Men dominated such occupations as doctors, lawyers and policemen, while women were shown in occupations such as nurse, secretary and waitress. Data by Long and Simon (1974) 28 also support these findings. These authors added that men were shown as always facing the challenges and triumphs of the outside world. In most instances they were shown as strong, capable and highly intelligent. Young children today are exposed to many different types. of work as they watch television. This media provides them the opportunity to learn about a wider variety of occupations than the usual traditional educational system. Television is an important source of communication because children spend more time watching than doing any other single activity except sleep. The incidental learning from television can contribute greatly toward attitudes about future occupations and those persons who may perform certain work roles. Data by DeFleur (1964) showed that out of 436 portrayals on prime time television and on weekends, twenty-four categories could be developed. Out of these twenty-four categories or groups, the most frequent form of work was related to the en- forcement and administration of the law with about one-third of the televised labor force involved. These data will give children a biased view of occupations leaning heavily in the direction of the law. Another important point about the data by DeFleur is that men and women professional workers were over-represented. Those professional persons with high social prestige in the labor force constituted about one-third of the televised labor force. Over half the men in this survey held jobs that rated 29 high in occupational prestige. DeFleur (1964) concluded from this survey that according to television portrayals, the world of work was a man's world. The same trend was leaning toward a man's world in the actual labor force, according to a comparison with the Indiana labor force. However, the distribution of the sexes appears to be more unequal for television than in the actual labor force. The comparison showed 83.9 percent males in televised work roles and only 16.1 percent females engaged in work roles. At that time Indiana labor force consisted of 68.9 percent male and 31.1 percent female employment. After reviewing these data one can easily conclude that television portrayed a distorted picture of occupational roles. Young viewers would have difficulty obtaining accurate informa- tion about the distribution of occupations and sexes by watching television. A survey of children's educational television by Dohr- mann (1975) indicated that on these shows jobs were held almost exclusively by males. One such example is Sesame Street and its occupational portrayal. Mr. Hooper owns the store, David works in his soda fountain and Louis has a fixit shop. Susan, the female character, is often referred to as a nurse, but these skills are not shown in the neighborhood. This report shows 15 percent of the work force on television are females, which is a distortion of reality since women comprised about 40 percent of the real world employment as of 1975. 30 In summary, one can conclude from previous data presented throughout this section that occupational portrayals.presented on television are very selective, unreal, biased,.stereotyped and misleading. Television is the leading source pf learning from among the media for children and yet it remains forlfrom reality. Sex Role Portrayals on Television Traditional sex role stereotypes are present in most aspects of television programming. Data show heavy viewers of television are generally more traditionally sex stereotyped than light viewers (McGhee, 1975; Frueh and McGhee, 1974; Stein and Friedrich, 1975; Waters, 1977). When children see stereotyped roles portrayed consistently, theytend to believe thatwhat they see is reality. An examination of a survey by Long and Simon (1974) provides a clear profile of how women are portrayed on televi- sion. This survey was based on twenty-two shows containing thirty-four female characters during a one-week period. It investigated the roles women perform and the status they occupy on children and family-oriented programs. The results of this survey reveal none of the married women worked outside the home and of the single women who did, only two held jobs with prestige in the society. Women in this survey did not occupy jobs with authority either at home or at work. Women were portrayed as silly, over—emotional, 31 passive, unauthoritative, and dependent on husbands or boy- friends. Women were always concerned about their appearance, their families and their homes. Verna (1975) added that women were shown as less career-oriented and less knowledgeable than men. Long and Simon concluded from their survey that the image of women on television was very traditional. Showing ‘15, “1». women as dependent and performing expressive and social/emotional roles within a family can be characterized as a stereotyped portrayal. All of the women in the survey were housewives, secretaries or quasi-secretaries; there were none portrayed in positions of authority such as doctors, professors, or executives. This survey ascertains that young children are not provided, via television, with a true image of the roles that women are capable of performing and are performing outside the television screen. 1 Another significant study related to sex portrayals was conducted by Sternglanz and Serbin (1974) These researchers observed and analyzed the sex role models that were shown on children's television shows. A comparison was also made of the ways males and females were treated on these shows. The pro- grams chosen for this study included the highest rated children's shows based on the Nielsen report in 1971 and 1972. Four of the ten most popular shows had to be eliminated because they did not have any female characters. 32 The results of this study showed males as being signi- ficantly more aggressive and constructive. Males were also shown emitting more total behavior and were more likely rewarded for such actions. 0n the other hand, females were shown as different and being punished for high levels of activity. The conclusions of this study indicate that commercially produced television programs shown throughout America are carrying differ- ent messages about the appropriate behavior for males and females. Another way of saying this is that commercially produced tele- vision is an important source of learning and from it, children are learning stereotyped sex roles. Just as children may learn sex role stereotypes from television, television can also serve as an important media and a leading source for teaching children counter-stereotyped behaviors and occupations. Miller and Reeves (1976) investigated children's exposure to the portrayal of women in counter- stereotyped occupations to determineif children would approve these occupations as real life occupations for women. The 200 subjects included third through sixth grade children from an elementary school in Eaton Rapids, Michigan. The findings of this study showed that children were more likely to say, after exposure to the television portrayals, it was appropriate for girls to aspire to various occupations. In this study no differences were found in exposure to counter- stereotyped portrayals and perception of real life occupations based on age and family socioeconomic status. These findings 33 also indicated that frequent exposure to counter-stereotyped sex roles and occupational portrayals of females on television will indeed cause some modification and changes in real life sex-role perceptions by children. Much of the evidence presented on sex roles has indicated that television reinforces sex stereotypes through its program- ming. However, this need not be the case. If children continue to watch television at the current pace of about six hours per day, television can portray a more positive image of females and can take the lead in changing stereotyped values that have been reinforced in the past. Findings from the Miller and Reeves study suggest that sex role images can bechanged by portraying more counter-stereotyped roles of males and females in all phases of television programming, including children's shows. Racial Portrayals on Television As a result of watching television, viewers obtain information about the world, people, and social relationships. In the past television portrayals have been dominated by Anglo- Americans. Churchill Roberts (1970-71) described it as being ”explicitly and glaringly white.” In the past five years there has been a trend toward integrating the television screen to include other ethnic groups. However, Stein and Friedrich (1975) reported that in both children's programs and prime time television most ethnic minority groups were virtually ignored and not included. Black Americans were the only group given some recognition and representation. This representation of 34 Blacks included very stereotyped portrayals which have been highly criticized. The portrayals of Blacks changed from demeaning characterization to include roles of higher status as a result of pressure from black militancy and protest groups in the late 1960's (Dominich and Greenberg, 1970). However, a survey by Simmons, Greenberg and Atkin (1977) revealed Blacks have been consistently shown performing in lower socioeconomic status roles than Whites. This study also showed Blacks being over-represented in lower occupa- tions and under-represented in higher occupations. Blacks were represented in 7 percent of professional roles and 6 percent managerial roles. About one-fourth of private household, farm managing and laborer types of occupations were held by Blacks in this survey. A survey of children's programs on Saturday morning in the San Francisco area by Ormiston and Williams (1973), where two-thirds of all school children are not Anglo, revealed that 17 of 27 programs surveyed contained no minority represen- tation at all. Those shows containing minorities portrayed them in highly stereotyped roles. For example, a majority of Blacks .were portrayed as villains and three-fourths as followers. The Chinese portrayal, in one program, depicted a father with nine children. McGee-Banks (1977) made a comparison of those television programs with integrated casts and those with all black casts to determine the types of knowledge television viewers are gaining 35 about Blacks and if the image of Blacks in segregated and integrated settings was consistent with that held by society based on social theory. According to Gordon's theory of cultural assimilation on which Ms. Banks based her research, Blacks who are highly assimilated culturally are usually those with high economic and social mobility. Those persons are usually judged more favorably by society than less assimilated’Blacks. This theory also proposed that Blacks who are highly assimilated will par- ticipate in racially integrated settings more so than less assimilated Blacks, and finally, those Blacks who are least assimilated will be judged negatively by the larger society. The programs selected for this study represented the three major networks--ABC, CBS and NBC. Those programs included "Mannix," "Good Times," "That's My Mama," "Get Christie Love," "Ironsides," and "Sanford and Son." The data indicated black characters in all black television casts performed a greater number of stereotyped characterizations and had more personal and family problems, and tended to have a lower social status. Those black characters portrayed in integrated casts had greater numbers of socially valued characteristics, commun- ity problems, and displayed higher social status symbols. These findings appear to be consistent with society's view of Blacks in segregated and integrated settings. While the above data might reflect accuracy in certain areas, it is difficult to accept as a universal pattern for 36 Blacks based ontheselinfited data. More research is needed to gain total support for this concept. From the data presented, one can conclude that television can reinforce and instill negative values for Blacks. When Blacks are portrayed negatively on television children may con- clude that to be Black is to be poor, carefree and uneducated. Those Blacks who are less assimilated are more likely to accept this type of image. When Blacks are portrayed in more integrated programs or in a more positive image on all black shows, the results could mean a more positive self-concept will be accepted by blacks. It must be noted that both positive and negative stereo- types about race can be learned from exposure to television. Television portrayals can have a great influence on children, particularly when there is little personal contact and other factual information about the group or its activities (Stein and Friedrich, 1975). Research has shown that Blacks are heavy viewers of tele- vision (Waters, 1977). Evidence also points out that heavy viewers are usually more stereotyped and traditional in their behavior patterns. Blacks are heavy viewers and heavy believers that what they see on the screen is real, according to the above researcher. 37 Summar There is evidence to indicate the differentiation of sex roles begins very early in life. The early years of a child's life appear to be the most crucial years in the development of sex roles. Research by Brown (1957), Hartup and Zook (1960), Mussen (1971) and Kagan (1964) indicated children have clear-cut sex differences asearly as three or four years ofage. At this age boys tended to have stronger preferences for masculine toys and activities than girls have for feminine toys, games and activities. The social learning theory described by Bandura (1971, 1977) indicated children learn appropriate and inappropriate sex roles by observing the behavior of others. Another aspect of the social learning theory is learning by imitation. Powerful and nurturant models are found to be imitated more by children than those models who do not possess these qualities. Results of an investigation by Maccoby and Jacklin (1974) found child- ren were more likely to imitate parents in acquiring sex type behavior during the early preschool years. Findings from a study by Bandura (1971) indicated children do not have to rely solely on parents as models since they are exposed to other models such as peers, teachers, other adults in the neighborhood and television characters. The results of an investigation by Cook and Smothergill (1973) indicated white models would be imitated more than black models regardless of the race or sex of the child observing the 38 behaviors. These findings pointed out that white models with power, control and prestige would be imitated more than black models who do not possess these qualities. The third aspect of the social learning theory pointed out that children tended to identify with models. They take on some of the models' characteristics and feelings and believe what is happening to the model is actually happening to them. Evidence pointed out that regardless of the social class level, boys tended to identify with the middle class boys if they had high levels of aspirations. The literature on the cognitive development theory by Kohlberg (1966) suggests a child's sex role development is not a biological instinct but is rooted in his/her concept of the physical surroundings. According to Kohlberg, a child's sex role concept is established by age five or six when there is an understanding of conservation. It is at this stage a child learns feminine and masculine stereotyped behaviors. In reviewing the literature on television as a source of learning, the evidence showed that children spend from two to six hours each day watching television. While children are being entertained by television, incidental learning takes place without any special incentives. Data from studies by Stein and Friedrich (1974), Simmons, Greenberg and Atkin (1977), Miller and Reeves (1976) and Long and Simon (1974) all indicate that men dominate occupational portrayals on prime time television. Men are also shown 39 over-represented on children's programs and on Saturday morning cartoon programs. When women are shown in occupational roles on television they are usually portrayed in stereotyped roles such as nurses and secretaries. 0n the other hand men are shown as being more powerful, intelligent, and challenging, and are found in such roles as doctors, lawyers, professors and administrators. Occupational portrayals of males and females on television pro- grams are found to be selective, unreal, biased, stereotyped and misleading. Ethnic representation on television programs is very limited. Black Americans are more represented than other groups; however, this representation is found to be very stereotyped. Blacks are over-represented in lower occupations and under- represented in higher occupations. Other ethnic groups such as Spanish Americans and Chinese are also shown in very stereotyped roles on television programs, when they are shown. CHAPTER III THE RESEARCH DESIGN This chapter is comprised of a description of the sample of the study and a discussion of the procedures for collecting the data. The testable hypotheses are also included in this sec- tion, followed by discussion of the means of analysis of the data. Description of Sample and Population The subjects for the experimental and control groups were enrolled in three elementary schools in Lansing, Michigan. The Lansing School District was chosen because of its close proximity to Michigan State University and the willingness of its personnel to participate in the research project. The three schools agreeing to participate in this project were a part of a desegregation plan in which students were bused from one neighborhood school to another. This allowed for a repre- sentation of students from middle and low socioeconomic back- grounds as well as a racial mixture of Black and white children. Three hundred and eighteen subjects, which included the entire kindergarten, second and fifth grades from each of the three schools participating, were used in the study. The children ranged from five through ten years old. The race and sex of the 40 41 total sample are shown in Table 3.1; the race and sex of the children in kindergarten are shown in Table 3.2; the race and sex of the second grade children are shown in Table 3.3; and the race and sex of the fifth grade children are shown in Table 3.4. TABLE 3.1 Race and Sex of All Subjects Males Females (50.9%) (49.1%) Total N N N Black (27.7%) 42 46 88 White (72.3%) _l_2_0_ _l__1_g _2__3_Q Total 162 156 318 TABLE 3.2 Race and Sex of Kindergarten Subjects. Males Females (52.5%) (47.5%) Total N N N Black (25%) 15 15 30 White (75%) _€L§ _4_2 .29 Total 63 57 120 42 TABLE 3.3 Race and Sex of Second Grade Subjects Males Females (44.9%) (55.1%) Total N N ‘ N Black (32.7%) 13 22 35 White (67.3%) §§_ §Z_ _12_ Total 48 59 107 TABLE 3.4 Race and Sex of Fifth Grade Subjects Males Females (56%) (44%) Total N N N Black (25.3%) 14 9 23 White (74.7%) .31 .31 .68 Total 51 4O 91 Procedures Two stimulus tapes were developed for use in the exper— imental conditions. The episodes were located using data from a content analysis of television programs conducted by the staff of Project CASTLE (Children and Social Television Learning) in the Department of Communications at Michigan State University. Each tape was approximately 20 to 25 minutes long. The first 43 tape consisted of all stereotyped roles portrayed by both Black and White males and females. The second tape consisted of all counter-stereotyped role behaviors being portrayed by Black and White males and White females. Black females were not included in the_second tape because there were none of these roles on the television shows from which the content analysis was taken. Seventeen males performing sex-stereotyped roles and nine males performing counter-stereotyped behaviors were included on the tapes. There were nine females performing sex- stereotyped roles and six performing counter-stereotyped roles. The goal was to have an even distribution of males and females and Blacks and Whites included on the tapes. However, males outnumbered females, Blacks were scarcely represented in any other role beside the stereotyped ones, and Black females were not represented at all in the counter-stereotyped roles. Subjects in each grade level were randomly assigned to one of two experimental groups and the control group. Each examiner explained to the group that they were going to watch a show with parts of several different television programs included. No other information was given. Group I viewed the tape with all stereotyped roles and group 11 viewed the tape with all counter-stereotyped roles. A list of all roles in both stimulus tapes is found in Appendix A. Table 3.5 shows the race and sex of children in the stereotype group (Group I), and the race and sex of children in the counter-stereotype group (Group II) are shown in Table 3.6. Groups I and II viewed .1 1.1.1.1. 44 the tapes in separate rooms simultaneously. TABLE 3.5 Sex and Race of Stereotype Group (Group I) Males _ Females (48.3%) (54.7%) Total N N N Black (27.4%) 16 13 29 White (72.6% 32_ 4§_ ‘_ZZ Total 48 58 106 TABLE 3.6 Sex and Race of Counter-Stereotype Group (Group II) Males Females Tot (54.6%) (45.4%) Total N N N Black (26.9%) 13 16 29 White (73.1%) 46. .33 _gpg Total 59 49 108 The third group (control) did not view a stimulus tape. This group was allowed to watch a cartoon, "Josey and the Pussy Cats" later for entertainment. Table 3.7 shows race and sex of children in the control group (Group III). 45 TABLE 3.7 Sex and Race of Control Group (Group III) Males Females (52.9%) (47.1%) Total N N N Black (28.8%) l3 17 30 White (71.2%) .42 32. __14 Total 55 49 104 Data were collected using a questionnaire. The question- naire consisted of 21 items representing three areas--household, occupational and general roles--that could be performed by either males or females. Data from the content analysis from the CASTLE study were used as partial basis for developing questions. Each question was followed by a picture of four figures--a White male, White female, Black male, and Black female. Children were allowed to circle a maximum of four choices for each question. Examples of the questions asked are shown in Figures 3.1, 3.2 and 3.3. The complete questiénnaire with directions appears in Appendix 8. Who shops for food for the family? FIGURE 3.1 Question in the Area of Household Roles 46 Who can be a police officer? FIGURE 3.2 Question in Area of Occupational Role Who talks a lot on the telephone? FIGURE 3.3 Question in Area of General Roles The questionnaire was pretested using twenty kindergarten and second grade children from one of the East Lansing elementary schools. This pretest was conducted to obtain clarity and under- standing of wording and directions by the children. This school was not used in the final sample. 47 The Cronbach coefficient reliability analysis was conducted for each of the two categories on the questionnaire to determine if the instrument was reliable. The Cronbach Coefficient analysis was used because of its uniqueness for testing questions where the subject's score on each question could take on a range of values (Mehrens and Lehmann, 1973). Table 3.8 shows a reliability coefficient of .855 for stereotype- counter-stereotype (S-CS) questions. As shown in Table 3.9, the reliability coefficient alpha was .934 for questions on race identification. The reliability analysis for both the S-CS scale and the race identification scale indicated the entire instrument used for this study has a high reliability score. TABLE 3.8 Cronbach Coefficient Reliability'Analysis for Stereotype/Counter-Stereotype Questions. Reliability Standard Coefficient Mean Variance Deviation Alpha S-CS 67.500 457.670 21.393 .855 TABLE 3.9 Cronbach Coefficient Reliability Analysis for Racial Identification Questions. Reliability * Standard Coefficient Mean Variance Deviation Alpha Tota1 Race 104.93391 475.9798 21.816 .934 48 The Pearson correlation coefficient for total race and total counter-stereotype (CS) shows a negative correlation of -.327. This correlation is significant at the .001 level and indicated students with high CS scores also had high racial identification scores. Each child in the study was given a questionnaire book- let and a pencil and instructed to circle the answer(s) of their choice. It was pointed out that there were no right or wrong answers to any of the questions, so children could make their choices more freely. The directions were read to all groups by an examiner. The questions were also read to the kindergarten and second grade children; however, the fifth graders were allowed to read their own questions based on their reading ability recommended by their teachers. There were two adults with each of the three groups at all times. Onethite and one Black examiner rotated in all grades between groups one and two, to control for this extraneous variable. There was also one adult male monitor present in each of these two experimental groups at all times. The control group maintained a White female examiner and monitor in all three grades throughout the testing. Experimental groups I and II were shown video tapes first, then responded to the questionnaire. The control group answered the questionnaire first, then watched a cartoon. The cartoon was not a part of the research project. 49 Design of the Study The design of this study is based on the post-test-only- control-group design (Campbell and Stanley, 1966). This design is appropriate for this study because of its applicability to research without a pretest and it is especially useful for lower elementary grades. The symbolic form for this study is as follows. Experimental Group I R X] 01 n = 106 Experimental Group II R X2 01 n = 108 Control Group R 01 n = 104 Total N = 318 FIGURE 3.4 Post-test-only-control-group design The control group was compared with groups I and II to determine if there were any differences in the responses as a result of viewing the stimulus tapes. The four independent variables or factors within this study were sex, race, grade level and groups. The design for this study is a factorial design with the following levels of each factor: (2 X 2 X 3 X 3) two levels of sex (male and female), two levels of race (Black 50 and White), three levels of grades (kindergarten, second, fifth), and three levels of treatment groups (Group I, stereotype; Group II, counter-stereotype; Group III, control). The following table represents the above-mentioned design. TABLE 3.10 Design of Study and Cell Size Group I Group II Group III Counter- Stereotype Stereotype Control Totals BLACK Female K 5 7 3 15 2 5 6 ll 22 5 3 3 3 9 Male K 5 4 6 15 2 6 ‘ 5 2 13 5 5 4 5 14 WHITE Female K 18 11 13 42 2 16 12 9 37 5 ll 10 10 31 Male K 9 23 16 48 2 12 10 13 35 5 ll 13 13 37 TOTALS 106 108 104 318 51 Testable Hypotheses The following six hypotheses were tested in this study. Null hypothesis. There will be no difference in the responses on the questionnaire of children viewing counter- stereotyped roles and those viewing stereotyped portrayals on television. Alternative hypothesis. Children who view counter-stereo- typed portrayals will give more counter—stereotyped responses on the questionnaire than children who view stereotyped portrayals. Null hypothesis. There will be no difference in the stereo- typed responses on the questionnaires of boys and girls. Alternative hypothesis. Boys will give more stereotyped responses on the questionnaire than girls. Null hypothesis. There will be no difference in the number of stereotyped responses on the questionnaire by older children and younger children. Alternative hypothesis. Older children will give more stereotyped responses on the questionnaire than younger children. Null hypothesis. There will be no difference in the responses of the children who view stereotyped roles and those in the control group. Alternative hypothesis. Children who view stereotyped roles will give stronger (more) stereotyped responses than children in the control group. Null hypothesis. There will be no difference in the responses of Black children viewing programs with black participants denoting identification with black characters. Alternative hypothesis. Black children who view programs with black participants will give more responses on the questionnaire denoting identification with black characters rather than white characters. Null hypothesis. There will be no difference in the responses of White children viewing programs with white participants denoting identification with white characters. 52 Alternative hypothesis. White children who view the pro- grams with white participants will give more responses on the questionnaire denoting identification with white characters rather than black characters. Analysis of Data The data obtained from the questionnaires were coded, key punched on computer cards, and verified. It was determined a_pripri which items on the questionnaire constituted stereo- typed (S) responses. A scale was set up to determine the counter-stereotype (CS) score with a range of O to 8. The higher number represents a more counter-stereotype score and the lower number represents a sex-stereotype score. Each item was scored according to this scale and the total score for each subject was recorded with a range from 0 to 168 as the overall CS score. (Appendix C shows the CS scoring scale.) The racial identification scale was also detenmined prior to the collection of the data. This scale ranged from 0 to 8, with eight meaning the subject had selected two characters of the same race as himself/herself and achieved the highest possible score on that item. Thus, the higher the score the more identification with the subject's own race. The total racial identification scores ranged from 0 to 168. The racial identification scoring is found in Appendix 0. Three independent judges at Michigan State University, Dr. E. Earhart, Chairman of Family and Child Sciences Department; Dr. B. Greenberg, Professor and Chairman of the Department of 53 Communications; and Dr. C. Atkin, Associate Professor of Communi- cations, reviewed the questionnaire and rating scales for counter- stereotype and racial identification scores and agreed this procedure was feasible. These three judges have been involved in research related to television for several years. The statistical test used in this study was the four-way analysis of variance (ANOVA) fixed effect model. The computer program used for analyzing the data was the Northwestern Univer- sity Statistical Package for the Social Sciences (SPSS). The analysis of variance (ANOVA)was used because of its unique method of splitting the total variations of the data into different com- ponents that measure the different sources of each variation (Walpole, 1974, p. 267). It allows testing and comparing the means of two or more groups simultaneously. Within this study the independent fixed variables are group, sex, race and grade levels. The dependent variable is the students' responses to the questions about sex roles and racial identification. The analysis of variance (ANOVA) allows the comparison of the means of the independent and the dependent variables as described above. The.assumptions of normality, equality of variance and independence have all been met in order to do this analysis. The alpha level of significance was set at .05 for decisions about rejection of all the hypotheses within the study. 54 mm The post-test-only-control-group design was used to test the hypotheses stated within this study. This design allows for testing for main effects and interaction among groups. All subjects were randomly assigned to one of the two experimental groups or the control group. The population for this study was the Lansing, Michigan elementary school district. The sample consisted of 318 subjects from the kindergarten, second, and fifth grades. Of the entire sample, 27.7 percent (n - 88) were Black students and 72.3 per- cent (n - 230) were White students. Subjects were randomly assigned to three groups (stereo- type, counter-stereotype, or control). The first two groups viewed a stimulus tape of television portrayals of males and females performing household, occupational and general roles. The third group (control) answered the questionnaire, then watched a cartoon for entertainment. The data from the three groups were analyzed to compare responses by the three groups and to determine if the stimulus tapes had any influence on subjects' responses. The four-way analysis of variance (ANOVA) fixed effect model was used and the alpha level of significance was set at .05. CHAPTER IV ANALYSIS OF RESULTS The results of the data analyses are presented within this chapter. Each hypothesis will be stated separately, followed by the analysis. Hypothesis 1 HO: Null hypothesis: There will be no difference in the responses of children viewing counter-stereotyped roles and those viewing stereotyped portrayals on television. H1: Alternative hypothesis: Children who view counter- stereotyped portrayals will give more c0unter-stereotyped responses on the questionnaire than children who view stereotyped portrayals. The analysis of variance (ANOVA) revealed an overall main effect for counter-stereotype/stereotype groups. Data show a main effect for grade at the .001 level and for group at .025 level of significance. Therefore, the stimulus tape shown to Group II of all counter-stereotyped (CS) roles indeed had a significant effect on the responses of subjects according to group and grade levels. A main effect was also found for race at .038 significance level. This treatment made a difference in the response of subjects according to race, although the difference was not as high. Blacks were more stereotyped in 55 56 their responses than Whites. Differences by race were not a part of this hypothesis. The main effects for sex, race, grade and group for stereotype and counter-stereotype groups are shown in Table 4.1. TABLE 4.1 Analysis of Variance for Main Effects in Comparison of the Stereotype and Counter-Stereotype Groups Source of Sum of Mean Signif. Variation Squares DF Squares F of F Main Effects 27147.672 6 4524.612 11.411 .001 ' Sex 723.002 1 723.002 1.823 .178 NS Race 1731.719 1 1731.719 4.367 .038 * Grade 21493.587 2 10746.793 27.104 .001 * Group 2979.124 2 1489.562 3.757 .025 * *Significant at the .05 level. NS = Not Significant. A two-way interaction of grade by group was significant at the .003 level for those subjects in Group II viewing the counter- stereotyped portrayals. Therefore, the treatment (stimulus tape) made a difference in subjects'responses but the differences depended upon the grade level and group subjects were in. There were no three or four-way significant interactions. Data for the two-way interactions are shown in Table 4.2. A comparison of the grade by group means is shown in Table 4.3 for subjects in all grade levels in the stereotype 57 (Group I) and counter-stereotype (Group II). TABLE 4.2 Analysis of Variance Showing Two-Way Interaction for Stereotype and Counter-Stereotype Groups Source of Sum of Mean Signif. Variation Squares DF Squares F of F Two-Way Interaction 11178.051 13 859.850 2.169 .011 Sex-Race 146.775 1 146.775 .370 .543 NS Sex-Grade 54.958 2 27.479 .069 .933 Sex-Group 1466.291 2 733.146 1.849 .159 Race-Grade 1772.400 2 886.200 2.235 .109 Race-Group 877.122 2 438.561 1.106 .332 Grade-Group 6569.098 4 1642.274 4.142 .003 * *Significant at the .05 level. NS = Not Significant TABLE 4.3 Grade Means for Group I and Group II Group I Group II Grades Stereotype Counter-Stereotype K 56.621 62.911 2 68.897 66.545 5 69.500 75.433 The means show those subjects in the kindergarten and fifth grades were affected most as a result of viewing the tape with all counter-stereotyped roles. The null hypothesis for Hypothesis I was rejected; however, this decision was based on 58 the interactions occurring in Groups I and II for the three grade levels. The means for the two-way interaction are shown in the graph of Figure 4.1. 80 - 75 — 7?.4 5th Grade 70 — ca/ 65 r— 68.8 W395 2nd Grade 60 - ./6‘2.9 Kindergarten 55 — 56.6 I l Group 1 Group 2 Stereotype (Counter Stereotype) FIGURE 4.1 Graph of Groups I and II Grade by Group Interaction The graph shows the children in kindergarten and fifth grade who watched the counter-stereotyped tape in Group II gave more counter-stereotyped responses than the kindergarten and fifth grade children in Group I. Those children in the second grade who viewed the tape in Group II were not affected by viewing the counter-stereotyped roles. 59 Hypothesis 2 HO: Null hypothesis: There will be no difference in the stereo- typed responses on the questionnaire of boys and girls. H1: Alternative hypothesis: Boys will give more stereotyped responses on the questionnaire than girls. The analysis of variance (ANOVA) test showed no main effects for sex in the stereotyped responses. The alpha level was set at .05 for significance. The F test for sex showed a significance level of .178. The treatment made no difference in the way boys and girls responded on the questionnaire in the stereotype group. No significant main effects for sex are shown in Table 4.4. The null hypothesis was not rejected for Hypothesis 2. TABLE 4.4 Analysis of Variance for Stereotype and Counter-Stereotype Groups Source of Sum of Mean Signif. Variation Squares DF Squares F of F Main Effect Sex 723.002 1 723.002 1.823 .178 NS NS = Not Significant. 60 Hypothesis 3 HO: Null hypothesis: There will be no difference in the number of stereotyped responses by older children and younger children. H1: Alternative hypothesis: Older children will give more stereotyped responses on the questionnaire than younger children. The data showed a significant group by grade interaction for stereotyped and counter-stereotyped responses. This inter- action was significant at the .003 level. There was a difference in the stereotyped responses of children; however, this differ- ence in response depended on the grade of subjects and the group or treatment that was viewed. The means for all three grade levels and all three groups are reported in Table 4.5. TABLE 4.5 Means for Grade by Group Interaction for All Grade Levels and Groups Group I Group II Group III Grade Stereotype Counter-Stereotype Control K 56.621 62.911 56.000 2 68.897 66.545 73.057 5 69.500 75.433 91.771 Grand Mean = 68.1981 The group by grade interaction for all groups and grade levels showed the younger children in kindergarten were more 61 stereotyped in their responses in Group I than both second and fifth grades. The stimulus tape in Group II made a difference in the responses of the kindergarten children; however, their responses were still lower than both second and fifth grade sub- jects in that same group. The children in second grade scored slightly lower (more stereotyped) than those second graders in Group I. The kindergarten children in Group III (control group) were more stereotyped in their responses than kindergarten children in Group II. They were also more stereotyped in their responses than second and fifth grade children in Groups I and II. Data also show the older children (fifth grade) were much more counter-stereotyped in their responses than younger children. The younger children (kindergarten) were more stereo- typed in their responses than older children. Therefore, the null hypothesis for hypothesis three was rejected. Figure 4.2 shows the graphed means for grade by group interactions for all grade levels and all groups. Hypothesis 4 HO: Null hypothesis: There will be no difference in the responses of the children who view stereotyped roles (Group I) and those in the control group (Group III). H1: Alternative hypothesis: Children who view stereotyped roles will give more stereotyped responses than children in the control group. A significant interaction was found for the three grade levels and Groups I (stereotype) and Group III (control), based 62 on the analysis of variance. The raw score means for all grade levels in Groups I and III are shown in Table 4.6. 95 '- 90 F- 9177 5th Grade 85 - 80 - 75 " 5.4 70 - 69//;0 2nd Grade 65 - 68.8 66.5 60 T" 62.9 55 - 56.6 56.0 Kindergarten 50 F- l l 1 Group 1 Group 2 Group 3 FIGURE 4.2 Graph of Group by Grade Interaction for All Grade Levels and Groups TABLE 4.6 Grade Means for Group I (Stereotype) and Group III (Control) Grade Group I (Stereotype) Group 111 (Control) K 56.621 56.000 2 68.897 73.057 5 69.500 91.774 Grand Mean = 68.1981 63 The data show a difference in the stereotyped responses of subjects is based on the grade of the subjects and the group they were in. Those children in Group I who viewed the stimulus tape with all stereotyped roles portrayed were more stereotyped in their responses in both second and fifth grades than those children in the control group. The stereotyped scores for children in kindergarten were about the same for Group I and for the control group (Group III). The null hypothesis was rejected for Hypothesis 4, based on the significant interactions. The graph in Figure 4.3 shows the differences in means for groups I and III. 95 - 9L7 90 — 5th Grade 85 - 80 - 75 P- 2nd Grade 70 - 695"—”,E—a””77336 65 - 688 60 h- 55 — 56.6— Ttio Kindergarten I I Group 1 Group 3 FIGURE 4.3 Graph of Means for All Grade Levels in Group I and Group III 64 The results of Hypotheses 5 and 6 will be reported together. Hypothesis 5 H0: Null hypothesis: There will be no difference in the responses of Black children viewing programs with black and white participatns denoting identification with black characters. H1: Alternative hypothesis: Black children who view programs with black and white participants will give more responses denoting identification with black characters rather than white characters. Hypothesis 6 H0: Null hypothesis: There will be no difference in the responses of White children viewing programs with white participants denoting identification with white characters. H1: Alternative hypotheses: White children who view the pro- grams with white participants will give more responses denoting identification with white characters rather than black characters. The data from the ANOVA test showed a main effect for race at .001 level and a main effect for grade at .001 level. The treatments had an impact on the race of subjects within some grade level. However, this impact was dependent upon other variables such as grade, sex, race, and group. The ANOVA table for race identification showing the significant main effects is shown in Table 4.7. Data from the ANOVA test show three significant two-way interactions. The sex and race interaction was significant at .033, race and grade was significant at .001 and grade and group showed a significant interaction at .006. The two-way 65 interactions are shown in Table 4.8. TABLE 4.7 Analysis of Variance Showing Main Effects for Racial Identification Source of _ Sum of Mean Signif. Variation Squares DF Square F of F Main Effects 38865.182 6 6477.530 26.386 .001 Sex 524.734 1 524.734 2.137 .145 NS Race 20323.056 1 20323.056 82.785 .001 * Grade 16816.061 2 8408.031 34.250 .001 * Group 380.909 2 190.454 .776 .461 NS *Significant at the .05 level. NS = Not Significant. TABLE 4.8 Analysis of Variance Showing Two—Way Interaction for Race Source of Sum of Mean Signif. Variation Squares DF Square F of F Two-Way Interaction 37973.217 13 2921.017 11.899 .001 Sex-Race 1124.872 1 1124.872 4.582 .033 * Sex-Grade 40.159 2 20.080 .082 .921 Sex-Group 35.209 2 17.605 .072 .931 Race-Grade 31112.935 2 15556.468 63.368 .001 * Race-Group 574.764 2 287.382 1.171 .312 Grade-Group 3605.512 4 901.378 3.672 .006 * *Significant at the .05 level. 66 The means for sex by race interaction for racial identi- fication are shown in Table 4.9, followed by a graph Of these means in Figure 4.4. TABLE 4.9 Means for Sex by Race Interaction for Race Identification Race Black White Male 86.738 109.991 Female 96.173 109.854 Grand Mean = 104.8742 110 105 100 95 90 85 \Nhhe ¢—* 4—= 1099 1098 96.1 Black 867 l L, Mam ande FIGURE 4.4 Graph of Means Showing Sex by Race Interaction 67 The graphed means of race by sex interaction for race identifications show Black males with lower identification scores than Black females. Black males and females had lower identification scores than both White males and females. White males and females' identification scores were about the same. There was a significant difference in the identification scores for Blacks and Whites as shown in the graph. The second significant two-way interaction for race identification was race by grade. This interaction was signi- ficant at the .001 alpha level. The raw score means are shown in Table 4.10, followed by a graph of these means in Figure 4.5. As shown in Figure 4.5, Blacks in kindergarten scored lower than Blacks in second and fifth grades. Those Blacks in second grade scored higher than the fifth graders. The graph means also show Whites in kindergarten with the highest overall identification scores. Whites in second grade scored slightly higher than Blacks in second grade and slightly higher than fifth grade Blacks and Whites. There was no difference in the identification scores for Blacks and Whites in the fifth grade. The third and last significant two-way interaction for race identification was grade by group. This interaction was significant at the .006 level. The means for this interaction are shown in Table 4.11. In order to get a visual picture of the total interaction and the location of the differences in means, a graph of the means from the raw score of racial identi- fication is shown in Figure 4.6. 68 TABLE 4.10 Means for Race by Grade Interaction for Race Identification Race Grade Black White K 78.033 126.477 2 99.771 101.152 5 97.130 97.308 Grand Mean = 104.8742 130 - 125 — ‘126A 120 h- 115 - 110 - 105 - 1on1 100 — 7.3 White 95 - 99.7 97.1. Black 90 - 85 - 80 F- 75 — 78.0 I J 1 Kindergarten 2nd Grade 5th Grade FIGURE 4.5 Graphed Means for Race by Grade Interaction 69 TABLE 4.11 Means of Grade by Group Interaction for Race Identification Group I Group II Group III Grade Stereotype Counter-Stereotype Control K 114.135 111.044 118.526 2 95.769 103.393 103.657 5 102.000 96.700 93.225 Grand Mean = 104.8742 120 3 Kindergarten 115 - 1185 110 - 114.1 111.0 105 — 2nd Grade 100 - 95 *- 5th Grade 90 (- l I I Group 1 Group 2 Group 3 FIGURE 4.6 Graphed Means for Grade by Group Interaction for Race Identification 70 The graphed means indicate that students identified with their own race differently, based on the group they were in and their grade level. The kindergarten children in Group II (counter- stereotype) were less affected by the treatment than those kinder- - garten children in Group I (stereotype). However, the kinder- garten children in the control group (Group III) had higher identification scores than both second and fifth grades. The fifth grade children had lower identification scores in Group I (stereotype group) than those in Group II (counter- stereotype group). However, there were no differences in the identification scores for the second grade children in Group II and those in Group III (control group). Data from the analysis of variance also revealed a significant three-way interaction for sex, race and grade at the .046 level. Even though this significance level is not very high, the data could be useful. The ANOVA table with three-way interaction appears in Table 4.12. TABLE 4.12 Analysis of Variance Showing Three-Way Interaction for Race Identification Source of Sum of Mean Signif. Variation Squares DF Square F of F 3-Way Interactions 4295.116 12 357.926 1.458 .140 Sex Race Grade 1533.644 2 766.822 3.124 .046 * Sex Race Group 1026.259 2 513.130 2.090 .126 Sex Grade Group 239.429 4 59.857 .244 .913 Race Grade Group 1820.543 4 455.136 1.854 .119 *Significant at the .05 level. 71 The means for the sex by race by grade interaction for race are shown in Table 4.13 to determine where the differences are. TABLE 4.13 Means for Sex by Race by Grade Three-Way Interaction for Race Identification M A L E F E M A L E Grade Black White Black White K 71.200 127.875 84.866 124.881 2 95.769 100.028 102.136 102.216 5 95.000 96.216 100.444 98.612 Grand Mean = 104.8742 Since this is a three-way interaction, the graph will have to be split into two levels for easier interpretation. Level one shows means for sex by race by grade for males and level two shows sex by race by grade for females. Both graphs are shown in Figure 4.7a and 4.7b. As seen in Figure 4.7a, Black males in kindergarten scored lower than Black males in second and fifth grades. The scores for Black males in kindergarten were also lower than White males in kindergarten. There was no difference in the identification score for Black males in second grade and Black males in fifth grade. 72 130 - 127.8 125 '- 120 — 115 '- 110 '- 105 "' 100 '- 95 "" 90 -' 85 F 80 " 75 '- 71.2 White Males Black Males Kindergarten 2nd Grade 5th Grade FIGURE 4.7a Graphed Means for Sex by Race by Grade Three-Way Interaction for Race Identification Level A, Males White males in kindergarten had the highest overall identification score for all grade levels. However,ldhite males had a lower identification score at the second grade level. White males at this level scored slightly higher than Black males in second grade. White males in fifth grade showed the lowest identification score for alllWhites. The score forlthite males at the fifth grade level were only slightly higher than 73 Black males at this grade level. The mean scores were 96.2 and 95.0 respectively. Figure 4.7b shows White females in kindergarten with a higher overall identification score. The identification scores for second grade were the same for both Black and White females. White females in fifth grade scored slightly lower than Black females in this grade level. The scores for White females in fifth grade were lower than White females in kindergarten and second grade. 125 1'- 124.8 120 '- 115 '- 110 P- 105 - 100 - 95 - 90 - 85 - 80 - 84.8 75 - Black Females White Females L 1 l Kindergarten 2nd Grade 5th Grade FIGURE 4.7b Graphed Means for Sex by Race by Grade Three-Way Interaction for Race Identification Level 8, Females 74 Black females in kindergarten scored lower than all other grade levels. The identification score for those Black females in fifth grade was slightly higher than for White females in the same grade level. Based on the significant interactions for race and grade in determining racial identification, the null hypotheses for Hypothesis 5 and Hypothesis 6 will be rejected. A differ- ence in the responses of both Black and White children existed; however, the difference was due to the dependent variables of sex, grade and group of children. Additional Data Some additional information on occupational roles and household roles not included in the stated hypotheses is of interest. Occupational Roles Those questions related to occupational roles showed significant main effects for race at .015, for grade at .001 and group at the .028 level. The main effect for group showed children in Group III with higher scores than Groups I and II. There were no differences in the responses of children in Groups I and II. The main effect for race and grade also shows significant interaction. The main effects for occupational roles are shown in Table 4.14. A list of all questions related to occupational roles is found in Appendix E. 75 TABLE 4.14 Analysis of Variance Showing Main Effects for Occupational Roles Source of Sum of Mean Signif. Variation Squares DF Square F of F Main Effects 20091.453 6 3348.576 21.481 .001 Sex 148.664 1 148.664 .954 .330 Race 925.464 1 925.464 5.937 .015 * Grade 18002.066 2 9001.033 57.741 .001 * Group 1124.228 2 562.114 3.606 .028 * *Significant at the .05 level. Along with the main effects for race and grade the ANOVA test showed a significant two-way interaction at the .002 level. The data revealed the difference in response was due to the race of the children and the grade level for the children. There were no three-way or four-way interactions present for occupa- tional roles. The significant two-way interaction for occupa- tional roles is shown in Table 4.15. The raw score means for the race by grade interactions are shown in Table 4.16 and a graph of the means appears in Figure 4.8. The graphed mean for the two-way interaction of race by grade for occupational roles shows kindergarten Whites with a lower occupational score than Blacks in kindergarten, and a lower score than Whites in second or fifth grade. The second grade Blacks recorded a lower occupational score than the second 76 TABLE 4.15 Analysis of Variance Showing Two-Way Interactions for Occupational Roles Source of Signif. Variation Squares DF Square F of F 2-Way Interactions 4801.163 13 369.320 2.369 .005 Sex-Race 43.860 1 43.860 .281 .596 Sex-Grade 316.450 2 158.225 1.015 .364 Sex-Group 654.981 2 327.490 2.101 .124 Race-Grade 1919.320 2 959.660 6.159 .002 * Race-Group 431.029 2 215.515 1.383 .253 Grade-Group 1440.842 4 360.210 2.311 .058 *Significant at the .05 level. TABLE 4.16 Means for Race by Grade Two-Way Interaction for Occupational Roles Race Grade Black White K 35.966 32.977 2 38.257 47.208 5 48.173 53.441 Grand Mean = 42.537 grade Whites, Those subjects in fifth grade showed a higher score than kindergarten or second grade for both races. However, there was little difference in the occupational responses of those Black children in fifth grade and the second grade Whites. The higher scores for fifth grade Blacks anlehites indicated more 77 liberal responses for occupational roles for males and females. All the fifth graders mean scores were above the overall grand mean of 42.537 for this two-way interaction. 55 — White 50 F- Black 45 '- 40 - 35 r- 30 '- 329 I I l Kindergarten 2nd Grade 5th Grade FIGURE 4.8 Graph of Means for Race by Grade Two-Way Interaction for Occupational Roles Household Roles The results from questions in the area of household roles indicated a main effect for sex at .017, and group at the .050 level. The main effect for sex indicated the treatments did indeed have a significant effect on the way males and females responded to the questions on household roles. Females were more stereotyped in their responses to household questions than males. The main effect for group indicated a significant effect .. 3.1.2.? 1141.: .5. 4mm. .... r... 78 existed within the groups which affected the way the children responded. Table 4.17 shows the significant main effects for household roles. Questions on household roles are found in Appendix F. TABLE 4.17 Analysis of Variance Showing Main Effects for Household Roles Source of Sum of Mean Signif. Variation Squares DF Square F of F Main Effects 1039.616 6 1073.269 2.636 .017 Sex 381.863 1 381.863 5.810 .017 * Race 73.216 1 73.216 1.114 .292 Grade 119.786 2 59.893 .911 .403 Group 397.065 2 198.533 3.021 .050 * *Significant at the .05 level. The data from the analysis of variance also show one significant two-way interaction of grade by group in household roles. This interaction indicated the main effect that existed for group is dependent upon which group and grade students were in. The two-way interaction of grade by group was significant at the .012 level. The grade by group two-way interaction is shown in Table 4.18. There were no significant three- of four- way interactions. The means for the two-way interaction on household roles are found in Table 4.19 and a graph of the means appears in Figure 4.9. 79 TABLE 4.18 Analysis of Variance Showing Grade by Group Two-Way Interaction for Household Roles Source of Signif. Variation Squares DF Square F of F 2-Way Interactions 1031.279 13 79.329 1.207 .273 Sex-Race 38.779 1 38.779 .590 .443 Sex-Grade 26.380 2 13.190 .201 .818 Sex-Group 30.029 2 15.014 .228 .796 Race-Grade 45.956 2 22.978 .350 .705 Race-Group 6.200 2 3.100 .047 .954 Grade-Group 859.693 4 214.923 3.270 .012 * *Significant at the .05 level. TABLE 4.19 Means for Grade by Group Two-Way Interaction for Household Roles Group I Group II Group III Grade Stereotype Counter-Stereotype Control K 16.351 16.733 16.391 2 17.435 14.606 17.971 5 13.933 18.000 22.354 80 24 -' 5th Grade 22 - £3 20 " 180 18 - 17K /17.9: 2nd Grade 16 '- 16.5 16.7 . R Kindergarten 14 - 145 12 - 139 10 - I I I Group 1 Group 2 Group 3 FIGURE 4.9 Graph for Grade by Group Two-Way Interaction for Household Roles The means for the two-way interaction for grade by group in household roles indicated fifth graders in Group I were more stereotyped in their responses to questions on household roles. Those fifth grade children in Group II scored higher than those in Group I. The increase may be due to viewing the tape with all counter-stereotyped roles. The fifth grade children in Group III (control group) scored higher than children in kinder- garten and second grade in all groups. 81 The data show second grade children in Group I with a higher score than the kindergarten and fifth grade children. Those second grade children viewing the tape in Group II had a decline in scores. This group was more stereotyped in their responses to household questions than second graders in Group I and Group III. The second grade children in Group II were also more stereotyped in their responses than the kindergarten child- ren in all groups. There were no differences in the responses of kinder- garten children in all three groups. The children in kindergar- ten had a higher score than fifth grade children in Group I and and second grade children in Group II. However, the kindergarten children in Group III scored lower than second and fifth grade children in that same group. The summary for all the null hypotheses tested, and the decision rules, are found in Figure 4.10 on the following page. H1: H2: H3: H4: H5: H6: 82 Hypotheses Decision Rule The Null Hypothesis Was: There will be no difference in the responses on the question- naire of children viewing counter-sterotyped roles and Rejected those viewing stereotyped por- trayals on television. There will be no difference in the stereotyped responses on the Not Rejected questionnaire of boys and girls. There will be no difference in the number of stereotyped re- sponses on the questionnaire by Rejected older children and younger children. There will be no difference in the responses of the children who view stereotype roles and those in the control group. Rejected There will be no difference in the responses of Black children viewing programs with black par- Rejected ticipants denoting identification with black characters. There will be no difference in the responses of White children viewing programs with white par- Rejected ticipants denoting identification with white characters. FIGURE 4.10 Summary of Hypotheses Tested and Decision Rules CHAPTER V SUMMARY, DISCUSSION AND IMPLICATIONS Summary The evidence in this study indicated watching a selected number of television portrayals of males and females performing sex-stereotyped and counter-stereotyped roles did influence some children's responses to sex stereotypes and racial identification. The degree of influence was determined by sex, race, grade and treatment group of the children. The sample consisted of three hundred and eighteen child- ren from three elementary schools in the Lansing, Michigan school district. From these three schools all children from the kinder- garten, second and fifth grade classrooms were used in this study. Of the entire sample, 27.7 percent (n = 88) were Black students and 72.3 percent (n = 230) were White students. Two experimental groups were compared with one control group for this study. Experimental group I consisted of 106 sub- jects and experimental group II consisted of 108 subjects. The control group (group III) had a total of 104 subjects. Subjects were randomly assigned to one of the three groups. The post test-only-control-group design was the method used for the hypotheses to be tested. Data were obtained through the 83 84 use of questionnaires. The Cronbach coefficient reliability analysis was conducted for the CS-S questions and for racial identification questions to determine the reliability of the instrument. The results from the coefficient alpha showed a .85 reliability level for the stereotype, counter—stereotype questions. The analysis for racial questions showed a .93 reliability level. The results from the Cronbach coefficient reliability analysis showed the instrument used was reliable in measuring the variables of stereotype and counter- stereotype responses and for measuring racial identification. The experimental groups watched stimulus tapes, and answered the questionnaire afterward. The S group (stereotype) watched a tape consisting of all stereotyped roles being performed by both Black and White males and females. The counter-stereotype group (CS group) watched a tape with all counter-stereotyped roles being performed by Black and White males and White females. The control group (C group) answered the questionnaire and watched a cartoon for entertainment. The cartoon was not a part of the research design. The viewing of stimulus tapes and completion of the ques- ' tionnaire were conducted within each of the three schools. The four-way analysis of variance (ANOVA) fixed effect model was used to analyze the data. This method of analysis allowed test- ing for significant main effects and interactions among groups. The two experimental groups were compared with the control group to determine if the stimulus tapes had any influence on children's responses on the questionnaire. The alpha level of significance 85 was set at .05 as a basis for the decision to reject the null hypothesis within the study. In comparing responses of children who viewed counter- stereotyped roles, the data show watching the tape with counter— stereotyped roles had the most influence on the fifth grade children, followed by the children in kindergarten. The mean scores for these children showed an increase over those children in S group after watching the stimulus tape with all stereotyped roles. The second grade children in CS group were more stereotyped in their responses than those second grade children in the S group who watched all stereotyped roles. The treatment produced a negative effect on the second grade children in the CS group. Data from the analysis of variance show no significant difference existed between the stereotyped responses of boys and those of girls. The treatment produced the same effect for both sexes in the stereotype responses. In comparing the stereotype responses of older children with those of younger children, the data show kindergarten child- ren in S and C groups were more stereotyped in their responses than second and fifth graders. The older children (fifth graders) provided more counter-stereotyped responses than younger children. Responses of children in the stereotype group (group I) were also compared with those by children in the control group (group III). The results show no differences in the responses on kindergarten children in S group or C group. However, the 86 second and fifth grade children in S group were more stereotyped in their responses than those second and fifth grade children in C group. The treatment in S group influenced the children in second and fifth grades to respond in a more stereotyped manner. However, the stereotyped responses for group I showed second grade children slightly more stereotyped than fifth graders. The fifth grade children in group III showed the highest counter-stereotyped score of all grade levels, followed by the second grade. Both of these grade levels showed children having higher counter-stereotyped scores than those children who watched the stereotyped stimulus tape in S group. A comparison of responses denoting racial identification showed Black males with the lowest identification score for both race and sex. Black females also had lower identification scores than White males and females. There were no differences in the identification scores for White males and females. Data from race by grade interaction showed Black children at the kindergarten and second grade levels with lower identifica- tion scores than White children. Black children in kindergarten showed the lowest identification scores, while White children in kindergarten showed the highest identification scores. Whites in second grade scored slightly higher than Black children at the same grade level. There were no differences in the identification scores for Black and White children in the fifth grade. The group by grade interaction for racial identification showed kindergarten children in all three groups with higher 87 identification scores than second or fifth grade children. The second grade children in group I (stereotype group) had lower identification scores than in the CS and C groups. There were no differences in the identification level of second grade children in counter-stereotype and control groups. The fifth graders in C group had a lower identification score than those in CS and S groups. Those fifth graders in S group had a higher identification score than the children in CS and C groups in the same grade. The treatment in CS group did not influence the racial identification responses of children in the fifth grade. Data showed a three-way interaction of race by sex by grade for racial identification. Black males in kindergarten had lower identification scores than Black females and lower scores than White males or females in kindergarten. White males had higher identification scores than White females and also higher scores than Black females and males in kindergarten. There were no differences in the identification score for Black males in second and fifth grades; however, their identifica- tion scores were lower than Black females, White females and White males in both second and fifth grades. White males in kindergarten showed the highest identifica- tion scores of all grade levels. The identification scores for White males in second grade were slightly lower than those scores for White females and Black females. However, Whites in second 88 grade scored higher than Black males in this grade level. White females in kindergarten scored higher than Black females in kindergarten. There were no differences in the identi- fication scores for Black females and White females in second grade. White females in fifth grade scored slightly lower than White females in second grade. Of the White males and females in all grade levels, those females in fifth grade had the lowest racial identification score for Whites. Black females in kindergarten showed a higher identifica- tion score than Black males in kindergarten. However, the scores for Black females in kindergarten were lower than White males and females at the kindergarten level. The data also showed Black females in second grade with higher identification scores than Black males in second grade. Black females in fifth grade showed a higher identification score than Black males in fifth grade. The identification scores for fifth grade Black females were slightly higher than for White females in the fifth grade. The analysis provided additional data on occupational and household roles that was significant, although not related to the hypotheses being tested within this study. Results from occupational roles showed a significant main effect for race, grade, and for group. The main effect for group showed children in C group with higher scores than for S and CS groups. There were no differences in the responses of children in S and CS groups. The children in S and CS groups were more 89 stereotyped in their responses than children in the control group. The main effect for race and grade also showed a signifi- cant two-way interaction for occupational roles. The data show both Black and White kindergarten children being more stereotyped in their responses to occupational roles than second and fifth graders. The fifth grade children, however, were more counter- stereotyped in their responses than both second grade and kinder- garten children. The results from analysis on household roles showed females were more stereotyped in their responses to questions on household roles than were males. Fifth grade children in S group were more stereotyped in their responses than kindergarten or second grade children. Second grade children in group II (counter-stereotyped group) were more stereotyped in their responses to household roles than those second graders in S or C groups. There were no differences in the responses of kindergarten children in all groups. The kindergarten children were less stereotyped than second graders in CS group and fifth graders in S group on household roles. Discussion of Findings Viewing Stereotyped and Counter-Steregtyped Portrayals The children who viewed the stimulus tape with all counter- stereotyped roles portrayed had higher counter-stereotyped scores 90 in all grades except second grade. yThe treatment did have an impact on children in kindergarten and fifth grades in the desired direc- tion. This impact lends support to findings by Miller and Reeves (1976) for the fifth grade. This study indicated viewing counter- stereotyped roles would cause some modification in children's perceptions of the appropriateness of sex roles. The second grade children in the S group were unaffected by viewing all counter-stereotyped roles. This group showed lower scores, thus being more stereotyped than the children who watched all stereotyped roles. It is unknown why this group scored lower. This pattern persisted for the CS group also in responses to ques- tions on household roles. More data are needed to give a logical explanation for the second grade stereotyped responses. Stereotyped Responses of prs and Girls It was assumed that boys would be more stereotyped in their responses than girls based on previous findings by Brown (1956), Hartup and Zook (1960), and Maccoby and Jacklin (1974). The results of this study, however, showed no significant differences in the responses of boys and girls. It can be assumed that viewing programs with males and females performing counter-stereotyped roles has caused a change in the attitudes and perceptions of boys about what is appropriate for each sex. Boys' responses can also be linked to more working mothers today than in the past. According to a survey by Dobkin 91 (1977), 46.1 percent of women with children under the age of eighteen are employed outside the home. While more women are working outside the home, many fathers are becoming more coopera- tive in sharing household tasks and responsibilities which might contribute to a change in stereotype attitude by boys. Other outside influences may have some effect on boys' changes in stereotype responses. 'Besides family attitudes, influences such as television portrayals of males and females, teachers, peers and other adults in the neighborhood could possibly affect changes in stereotyped attitudes of boys. Stereotyped Responses of Older and—Younger Children The evidence from this study shows younger children, those in kindergarten, to be more stereotyped than older children. These findings are supportive of the results of Mussen (1971) and Kagan (1964). The young child uses his/her immediate family as models to copy or imitate. When family members are very stereotyped in their attitudes and roles, children will imitate such behavior. If such behavior is rewarded, the reward serves as an incentive to continue the stereotyped imitative behavior. Parents have the greatest influence on very young children. Children behave in a manner they think is accepted by parents to avoid punishment, according to findings by Mussen (l97l). 92 Young children believe what they see is the way things should be because of their inability to rationalize. When stereo- typed roles are shown on television, children believe that this is the way males and females should be portrayed and those roles are acceptable to a young child. Kindergarten children have not reached the stage where they are able to make rational decisions regarding sex roles. At this stage children do not think beyond themselves and their families. On the other hand, older children are able to rationalize and make decisions better than younger children. They can begin to think in the abstract. Their cognitive skills are far superior to younger children's. Therefore, older children are more influenced by viewing stereotyped roles on television than younger children are. Older children have also been exposed to wider varieties of models than younger children. In this study older children were more influenced by viewing counter-stereotyped roles. The treatment in the counter-stereotype group produced the hypo- thesized results. Comparison of Stereotype Group and the Control Group It was expected that children in the S group would be more stereotyped in their responses than children in the control group. This proved to be the case in all grade levels except for the kindergarten children. There were no differences in the responses of children in kindergarten in the S group and C group. The treat- ment made no difference to kindergarten children in the S group. 93 The scores for the second grade children moved in the direction expected. These children were more stereotyped in the S group after watching the tape with stereotype roles than second grade children in the C group. The treatment was effective for this group. The fifth grade children in the S group were much more stereotyped in their responses compared to fifth grade children in the C group. The fifth grade children had higher counter- stereotype scores than second grade children or children in kinder— garten. The differences in responses for the two groups were not expected to be so great. No data were collected as-a part of this study about home television viewing. Research results by Waters (1977) and Freuch and McGhee (1974) indicated heavy viewers of television were more sex stereotyped than light viewers. One would question whether fifth grade children in group C could have been exposed to stronger stereotype influence at home or if they were light viewers of tele- vision. Racial Identification for Black and White Children The data showed Black males and females identifying less with Black characters than White children identify with White characters. Black males had lower identification scores than Black females. 94 These findings support other research results by Waters (1977), Cook and Smothergill (1973) and Maccoby and Wilson (1957). These researchers found Blacks would imitate and identify with white models more than black models. This was due to Blacks' beliefs that white models and particularly White males have more power, control and prestige in society. They identify with those characters based on their level of aspirations for achieving power and control. There were no differences in the identification levels of White males and females. Both White males and females obtained a higher racial identification score than Black males and females. Black children at the kindergarten level had lower indenti- fication scores than Black children in second and fifth grades. Their scores were also much lower than White kindergarten children who showed the highest identification score for Whites. One reason Black children in kindergarten showed a low identification score might be because television programs with black characters portray Blacks in a lower social class status than Whites (Simmons, Greenberg and Atkin, 1977). Also, Blacks are shown as followers rather than leaders on many programs, in mana- gerial roles rather than administrative roles, and as villains rather than heroes. Blacks in kindergarten are not able to make rational decisions about these programs. They therefore believe what they see is real and do not wish to identify with such charac- ters . 95 White children, on the other hand, see Whites portrayed in professional roles, with power, in control of resources and in prestigious occupations. White kindergarten children would prefer to be like the successful models they see. The second grade children are at a stage where they can begin to sort out facts; therefore, the difference in identifica- tion scores for this group is not as great as for children at the kindergarten level. When children reach the fifth grade level there are no differences in the identification responses of Black and White children. These children do not rely solely on television for their racial influences. Many other outside forces interact to assist them in the decision to identify with their own race. Black children at this stage are exposed to many more cul- tural activities and outside forces to enable them to identify with their own race. Black history has been incorporated into many school curriculums, and children are exposed to more movies with leading black characters. At the fifth grade level parents have had more time to instill children with pride in being Black. Television programming is just one means of influence, but it is by no means the only influence for racial identification for'Black children. Television is found to be a strong influence, however, and if Blacks are portrayed in a more positive manner on all black shows and on integrated shows, Blacks would probably gain a more positive racial identification score. 96 In looking at racial identification by grade and group, the results showed that treatment in the CS group had no effect on the children at the kindergarten level. There were lower identification scores for children viewing counter-stereotyped roles in the CS group. The second grade children in the CS group were influenced by the treatment they viewed. The fifth grade children were least affected by the treatment in the CS group and the C group. The lower identification scores may be due in part to their age, experience with others, social class level, or family attitudes aboUt differences in race. Occgpational Roles The additional data also indicated kindergarten children were more stereotyped in their responses to occupational roles than were second and fifth grade children. There was an increase in CS scores for the second grade. The fifth grade children were more counter-stereotyped in their responses than second grade children. Whites in second and fifth grades showed higher CS scores than Blacks for these grade levels. The higher occupational scores for Whites may be partially attributed to television portrayal of Whites in higher level occu- pations than Blacks. Whites appear on television in such occupa- tions as doctors, lawyers, professionals associated with law enforcement and administration. Blacks are shown as comedians on variety shows or on all black shows depicting low grade occupations. '1 “(E—m 11* 97 The responses of children to occupational roles in second and fifth grades indicated the portrayal of both black and white characters can affect a change in perception about occupational roles for males and females for both Black and White children at this level. A change in occupational portrayals on television could help change the double standard that television programming has created with regard to certain roles being performed by males. Household Roles A main effect for sex showed girls being more stereotyped in their responses to household roles than boys. Females are most often shown on television as housewives, or homemakers, taking care of family and in social-emotional roles. Females are often shown being dependent on husbands or boyfriends and being less knowledge- able and less career-oriented than men. Men are seldom shown performing household roles on tele- vision. However, showing men in more household roles could possibly produce a change in attitude about household roles. This could provide children who do not get this exposure from their own families an opportunity to see that it is acceptable for both sexes to perform household roles. A grade by group interaction showed kindergarten children were not affected by the treatment in the S group or the CS group. Their scores were the same across all groups. These results might be explained by previous data which show younger children more stereotyped than older children and these children have not 98 experienced the exposure to outside influences to the same degree as older children. The second grade children in the CS group produced results in the opposite direction from what was expected. These children were more stereotyped in their responses to household roles than children in the S group viewing a tape with all stereotyped roles and those in the control group. There was only a slight difference in the responses of second grade children in the S group and those in the C group. Children in the fifth grade CS group showed a higher counter-stereotyped score while fifth graders in the S group had the lowest stereotype score on household roles. It is evident that positive portrayals of males and females in household roles might cause some change in children's attitudes over time. Second grade children appear to be more influenced to change than children in kindergarten, according to the results of this study. When children see counter-stereotyped roles portrayed by males and females they will begin to think in a more counter- stereotyped manner. On the other hand, if stereotyped roles con— tinue to be shown on television, children with little outside influence will become more stereotyped in their thinking and atti- tudes. Limitations of the Study 1. The sample was based on three schools that agreed to participate in the study and may not be representative of the f: 31”.“. ‘u r n q. 99 general population.« Generalizations are made only to the three schools and grade levels used in the study. 2. The stimulus tape for counter-stereotype roles did not include Black females. Implications for Future Research The data from this research could be extended through the following suggestions. 1. A more even distribution of Black and White children would perhaps produce different results in a similar study. 2. Conduct a similar study in a different geographical region, such as south, east or west, and compare responses of Blacks and Whites in rural and urban areas. 3. Include Black females in the counter-stereotype stim- ulus tape to determine if different results would be obtained. 4. Use two groups instead of three, one experimental group and one control group. The experimental group will be shown the stereotype tape with a designated time lapse, at which time they would be shown the counter-stereotype tape and questioned to determine if there were any differences in responses over time. These responses would be compared with the responses of children in the control group. 5. Determine the relationship of children's responses to television portrayals to their home life and to their parents' attitudes on sex roles and occupational portrayals. Also, determine the relationship of teachers' attitudes on sex roles to 100 children's responses. 6. Replicate this study with second graders to determine what influenced their responses on household roles. Implications for Practical Use The results from this study could be useful to parents and other interested groups as a basis for requesting a change in portrayals of males and females on television. Teachers in classrooms could use the data in classroom settings to help foster counter-stereotyped attitudes about males and females in occupational and household roles through school activities. Television networks could become more informed through the use of these data about the effects that stereotyped portrayals on television programs have on young children. Data from this study will also be provided to networks to show the relationship of por- traying Blacks in low status occupations and non-heroic roles to low levels of identification in Black children from this sample. APPENDICES APPENDIX A Appendix A ROLE PORTRAYALS 0N STIMULUS TAPE FOR STEREOTYPE GROUP Advertisement - Female making chocolate chip cookies Marcus Welby - Dr. Welby discussing surgical procedures with patient's parents Medical Center - Doctor being monitored while conferencing with male intern - Doctor meets with male attorney about patient LaVerne and Shirley - Male doctor examines LaVerne in her home, she adores him Bob Newhart Show - Bob talks with Carol, female secretary receptionist - Jerry, male dentist, examines child's teeth, talks about accomplishments and satisfaction from work CPO Sharkey - Sharkey discusses negative views about women joining the Army with Black male Sirota's Court - The Press; White male interviewing for paper White male defense attorney Black male attorney White female attorney getting instructions about case Judge Sirota, White male Three for the Road — White female and White male plotting to steal van and camera equipment Emergency One - White-male firefighters fight fire and rescue occupant 101 102 That's My Mama - Earl, Black male shining shoes - Clifton, Black male barber Happy Days - Mother making breakfast while dad sits at table eating - Rich demands oatmeal for breakfast What's Happening - Black male ordering female friend to study - The mother, Black female head of household with 2 children - Mother reprimands son for helping tutor friend and getting low grades in school Little House on the Prairie - Charles, White male father, asks children questions about 1 Bible ‘ - Mother is shown with basket of washed clothes Daytime Drama (Soap Opera) - Female talks on phone to priest (White male) who gives her directions about life 10. 11. 12. 103 ROLE PORTRAYALS 0N STIMULUS TAPE FOR COUNTER-STEREOTYPE GROUP Advertisement - Male cooking food with new microwave oven Welcome Back Kotter - Male school teacher M.A.S.H. - Radar, male, cleaning and dusting office Three for the Road - Pete, male, doing laundry in commercial laundromat Barney Miller - Detective Harris, Black male Kate McShane - Kate, female attorney CPO Sharkey - Female commanding officer - Two segments included Emergency 2 - Female doctor Swiss Family Robinson - Lottie cutting down tree with axe Marcus Welby — Maggie, female, operates medical machinery Sirota's Court - Black male district attorney - Female lawyer That's My Mama - Earl, Black male, making beds 104 13. Joe and Sons - Joe, male, fixing sandwiches for son's lunch 14. The Cop and the Kid - Male head, single family adopting Black child (male) APPENDIX B APPENDIX B QUESTIONNAIRE CHIIDREN'S PHICEPTIOIIB OF SEX ROLE SW 0N SELECTED TELEVISION PROM 106 DIRECTIONS THIS SET OF QUESTIONS WAS MADE TO FIND OUT HOW YOU FEEL ABOUT THEWAYIENANDWOMENARESHOWNONTEIEVISIONANDW—IAT YOU THINK THEY CAN DO IN REAL LIFE. THERE ARE FOUR PICTURES APTER EACH QUESTION. (HOLD UP PICTURES AND IDENTIFY EACH: BLACK MAN, BLACK LADY: WHITE MAN, 'vHITE LADY). I WILL READ THE QUESTIONS FIRST AND THEN YOU CAN CIRCLE ONE OR MORE PICTURES THAT YOU THINK WOULD ANSWER THE QUESTION. REMEMBER YOU MAI CIRCLE MORE THAN ONE PICTURE IF YOU LIKE. THERE IS NO RIGHT OR WRONG ANSWER TO ANY OF THESE QUESTIOES. IF YOU DO NOT UNDERSTAND THE QUESTIONPIEASERAISEIOURHANDANDOIE OPTHEHEIPERS INTHE ROOMWILLHELPIOU. WEWIILDO'IHBEEEXAMPIES TOHELPYOU UNDERSTAND. ......... ANSWER QUESTIONS, IF ANY. YOUMAITURN'II-IEPAGENOWAMDIEWIILEEGIN. 107 A. Who is most like me? B. C. 108 l. ‘Who goes to work every day? ‘Who can be a police Officer? 2. 1y? ami Who cooks the food for the f 3 ‘Who can be a lawyer? h. ‘Who is home more Often than at work? 5. 109 6. Who takes care of little children at home? 7. 8. 9. IO. 110 11. Who cleans the house? 12. 13. 15. 111 16. Who can be a teacher? l7. 19. 20. 21' 112 INFORMATION NAME ACE GRADE IEVEL: KINDERGARTEN 2nd 5th RACE: BLACK WHITE OTHER ARE YOU: A BOY OR A GIRL SCHOOL APPENDIX C Appendix C STEREOTYPE (S) -- COUNTER-STEREOTYPE (CS) SCORING SCALE‘ 0 = 0 counter-stereotype; O stereotype 1 = O counter-stereotype; 2 stereotypes 2 = 0 counter-stereotype; l stereotype 3 = 1 counter-stereotype; 2 stereotypes 4 = l counter-stereotype; l stereotype 5 = l counter-stereotype; O stereotype counter-stereotypes; 2 stereotypes counter-stereotypes; l stereotype 00 \I 01 II N N N counter-stereotypes; 0 stereotype 113 APPENDIX D Appendix D RACIAL IDENTIFICATION SCORING SCALE same; 0 opposite same; 2 opposite .4 ll 0 O 0 same; 1 opposite same; opposite ~51 00 N ll ....1 2 = 2 same; 2 opposite U'l II ._| same; 1 opposite = 2 same; 1 opposite same; 0 opposite (I) \l 03 II —I = 2 same; 0 opposite 114 APPENDIX E Question 1. Who 2. Who 4. Who 7. Who 10. Who 12. Who 14. Who 16. Who 18. Who 19. Who 20. Who Appendix E QUESTIONS ON OCCUPATIONAL ROLES goes to work every day? can be a police Officer? can can can can can can can can can be a be a be a be a be a be a be a give be a lawyer? nurse? doctor? school principal? crook? teacher? secretary? the news on television? judge? 115 APPENDIX F Appendix F QUESTIONS ON HOUSEHOLD ROLES Question 3. 5 6. 9 11. 15. 17. Who cooks the food for the family? Who is home more Often than at work? Who takes care Of little children at home? Who takes care Of the people at home when they are sick? Who cleans the house? Who washes the family's clothes? Who shops for food for the family? 116 REFERENCES REFERENCES Bandura, Albert. "Social Learning Theory Of Identificatory Processes," in David A. Goslin, Handbook Of Socializa- tion Theory and Research, Rand McNally and Company, Chicago, 1971. PP. 213-262. Bandura, Albert. Social Learninngheory. General Learning Press, Morristown, New Jersey, 1971. Bandura, Albert.~ Origins Of Behavior, Social Learning Theory. Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1977. Bandura, A., Grusee, J. E. and Menlove, F. L. "Observational Learning as a Function Of Symbolization and Incentive Set." Child Development. 1966, 37:499-506. Bandura, A., Ross, 0. and Ross, 5. A. "Transmission of Aggressive Models." Journal of Abnormal Social Psychology. 1961, 63, 575-582. Bandura, A. and Walters, R. H. Social Learning_and Personality Development. Holt, Rinehart and Winston, NewFYOrk, 1963. Banks-McGhee, and Cherry A. A Content Analysis Of the Treatment of Black Americans on Television," Social Education, Vol. 41, NO. 4, April, 1977, pp. 336-339. Bennett, E. M. and Cohn, L. R. Men and Women: Personality Patterns and Contrasts. Genetic Psychological Mono- 9r Eh. 1959, 60, 101-1530 Brown, Daniel G. "Sex Role Preference in Young Children," Psychological Monographs: General and Applied. V01. 70, NO. 14, Whole NO. 421, 1956, pp. 1-19. Campbell, Donald T. and Stanley, Julian C. Experimental and Quasi-Experimental Designs for Research. Rand McNally and Company, Chicago, 1963, p. 25. Cook, Harold and Smothergill, Daniel W. "Racial and Sex Deter- minants of Imitative Performance and Knowledge in Young Children,“ Journal Of Educational Psychology, Vol. 65, NO. 2, 1973, PP. 211-215. 117 118 Culley, James D., Lazer, William, and Atkin, Charles K. "The Experts Look at Children's Television," Journal of Broadcasting, Vol. 20:1, Winter, 1976, pp. 3-20. DeFleur, Melvin L. "Occupational Roles as Portrayed on Tele- vision," Public Opinion Quarterly. Vol. 31, NO. 1, Spring, 1964. DeFleur, Melvin L. and DeFleur, Lois B. "The Relative Contribu- tion of Television as a Learning Source for Children's Occupational Knowledge." American Sociological Review. Vol. 32, NO. 1, 1967. PP. 777-789. Dobkin, Robert A. "Nearly Half Of U. S. Wives Now Working Out- side Home," Lansing State Journal, Lansing, Michigan, March 8, 1977. Dohrmann, Rita. "A Gender Profile of Children's Educational T.V." Journal Of Communication, Autumn, 1975. V01. 25, NO. 4, pp. 56-65. Felitzen, Cicelia V. and Linne, Olga. "The Effects Of Television on Children and Adolescents, Identifying with Television Characters." Journal of Communication. Vol. 25, NO. 4, Autumn, 1975. Pp. 51-55. Frueh, Terry and McGhee, Paul E. "Traditional Sex Role Develop- ment and Amount of Time Spent Watching Television," Developmental Psychology. Vol. 11, NO. 1. 1975, p. 109. Hale, Gordon A., Miller, Leon K. and Stevenson, Harold W. "Incidental Learning Of Film Content: A Developmental Study," University Of Minnesota. Child Development. Vol. 39, NO. 1. March, 1968, pp. 69-77. Hartup, Willard W. and Zook, Elsie A. "Sex Role Preferences in Three and Four-Year Old Children." Journal Of Consulting Psychology, 1960, Vol. 24, NO} 5, pp. 420-426. Hicks, David J. "Girls' Attitudes toward Modeled Behaviors and the Content of Imitative Private Play." Child Develop- ment. Vol. 42, No. 1, March, 1971, pp. 139-147. Hoffman, Martin L. and Hoffman, Lois W., editors. Review of Child Development Researoo. Russell Sage Foundation. New York, V01. 1, 1964, pp. 137-164. Kagan, J. "The Concept Of Identification," Psychological Review. 1958, 65, 296-305. 119 Kagan, Jerome. "Acquisition and Significance Of Sex Typing and Sex Role Identity," Review of Child Development Research. Vol. 1. Edited by Martin L. HOTfman and Lois W. Hoffman, Russell Sage Foundation, New York. 1964, pp. 137-164. Kohlberg, Lawrence. "Stage and Sequence: The Cognitive Develop- mental Approach tO Socialization." In David A. Goslin, Handbook Of Socialization Theory and Research. Russell Sage Foundation, Rand McNally and Company, Chicago, 1971, pp. 347-480. Lane, Myrna. “Sex Typing in Children as a Function Of the Parents' Sex Role Identification." Master's Thesis, Michigan State University, 1968. Long, Michele and Simon, Rita J. "The Roles and Statuses Of Women on Children and Family Television Programs." Journalism Quarterly. Spring, 1974. PP. 107-110. Maccoby, Eleanor E. and Jacklin, Carol N. "Sex Typing and The Role Of Modeling." The Psychology of Sex Differences. Stanford University Press, StanfOrd, California, 1974, pp. 277-302. Maccoby, Eleanor E. and Wilson, William C. "Identification and Observational Learning from Films." The Journal of Abnormal and Social Poyohology. Vol. 55, 1957, pp. 76-86. Masingale, Eula Mae. "Father-Absence as Related to Parental Role-Play Behavior." Doctoral Dissertation, Michigan State University, 1971. McGhee, Paul. "Television as a Source Of Learning Sex Role Stereotypes." Paper presented at Biennial Meeting of the Society for Research in Child Development, Denver, Colorado, April 10-13, 1975. Ed 111 528. Mehrens, William A. and Lehmann, Irvin J. Measurement and Evaluation in Education and Psychology. Holt, Rinehart and Winston, Inc., New York, 1973. Miller, Mark M. and Reeves, Byron. "Dramatic T. V. Content and Children's Sex Role Stereotypes." Journal Of Broad- casting. 20:1, Winter, 1976, pp. 35-50. Minuchin, Patricia. "Sex Role Concepts and Sex Typing in Child- hood as a Function of School and Home Environment." In Social Learnin in Childhood,yReadingo in Theory_and ApplTCation, th d by_M. Gelfand. Brooks ColeTPUb- lishing Company, Belmont, California, 1969, pp. 269-281. 120 Mischel, Walter. "A Social Learning View of Sex Differences in Behavior." In The Development Of Sex Differences. Edited by Eleanor E.7Macc0by, StanfOrd’UnTVerSTty Press, Stanford, California, 1974. Mussen, Paul H. "Early Sex Role Development." Handbook of Socialization Theory and Research. Edited by David A. Goslin, RusséTl Sage Foundation, Rand McNally and Company, Chicago, 1971, pp. 707-731. Nie, Norman H.; Hull, C. Hadlai; Steinbrenner, Karia; and Bent, Dale H. Statistical Package for the Social Sciences, SPSS. Second Edition, McGraw Hill Book Company, New York, 1975. Ormiston, Linda H. and Williams, Sally. "Saturday Children's Programming in San Francisco, California: An Analysis Of the Presentation Of Racial and Cultural Groups on Three Network Affiliated San Francisco Television Stations." Speech presented to the Federal Communica- tions Commission's Hearing on Children's Television. Washington, D. C. January 8, 1973. Ed 071440. Rabban, M. "Sex-Role Identification in Young Children in Two Diverse Social Groups." Genetic Psychology Monographs. 1950, 42, 81-158. Roberts, Churchill. "The Portrayal of Blacks on Network Tele- vision." Journal of Broadcasting. Vol. XV. No. 1. Winter, 1970-71. pp. 45-53. Simmons, Katrina W.; Greeberg, Bradley S. Atkin, Charles K. with Carrie J. Heeter. "The Demography Of Fictional Television Characters in 1975-76." Department of Communications, Michigan State University. Castle Report Number Two. February, 1977. Stein, Aletha Huston and Friedrich, Lynette Kohn. Impact of Television on Children and Youth. The Univers1ty 0 Chicago Press, Chicago and London, 1975, pp. 57-63. Sternglanz, Sarah H. and Serbin, Lisa A. VSex Role Stereo- typing in Children's Television Programs." Developmental Psychology. 10:5, September, 1974, pp. 710-715. Unger, Arthur. "T. V. -- It Stops Trains, Changes Time, Adjusts History . . . and Alters Life?" Christian Science Monitor. May 26, 1977. 121 Verna, Mary Ellen. "The Female Image in Children's Television Commercials." Journal of Broadcasting. 19:3, Summer 1975. pp. 301-309. Walpole, Ronald E. Introduction to Statistics. Second Edition. MacMillan Publishing Company, Inc., New York. Collier MacMillan Publishers, London, 974, p. 267. Walters, Harry F. "What T. V. Does to Kids." Newsweek. February 21, 1977, pp. 63-76. 3835.92 HGWfiOWbbm wOCZUPEOZ Erma Wu 3,.» .25 H3308 . . . baafimom mam woeiomm batman :Zmafioaozm mgiomm >283: Ema U33” 03255.53 ”335:! .35 bmawswfimao: 4.5.888 00326389 W388? mxSSmwo? . . 35:93 . . . .25 560258 8 <05 Oeomaumaos :2 56:25 mmmumtsh ENE :22 magnum ”HERBS acoU—WSSRBO :oflmELOME 3383862 MTF MEMBERSHIP APPLICATION "I1111111.1”1.1111111111111111“