. LIBRARY ”Michigan Stat! .1 2; University This is to certify that the thesis entitled Sex Role Portrayals 0n Commercial Broadcast Television presented by Laura Lee Henderson has been accepted towards fulfillment of the requirements for MJL degree in Lnnmunicati on Majbr prowa Date February 24, 1978 0-7639 —- I ll“! Illzllljljllll an ll! ll 1m Will u m (D SEX ROLE PORTRAYALS ON COMMERCIAL BROADCAST TELEVISION By Laura Lee Henderson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1978 ABSTRACT SEX ROLE PORTRAYALS ON COMMERCIAL BROADCAST TELEVISION By Laura Lee Henderson Sex role portrayals on prime time and Saturday morning television were content analyzed. Three categories of beha— vior were defined and recorded from a social learning theory perspective. The three categories were: Dominance/Deference: Gives Orders; Nurturance/Succorance: Needs Support; and Independence/Dependence: Makes Plans. Orders were defined as directives to do,say, or think something. Supports were defined as non-routine physical and emotional needs of assistance. Plans were defined as statements of a method for doing something. The data were analyzed with t-tests and a difference of proportions test. Results showed sex differences within all categories. The data were also analyzed by program type, broadcast time, and viewing preference. The results were discussed in terms of the stereotyping of male and female sex roles as well as the availability of sex role models for imitation. ACKNOWLEDGEMENTS There is no way to begin even properly recognizing all the folks who helped out along the way. My committee, headed by Dr. Greenberg and filled out by Katrina Simmons and Tom Muth, most certainly deserves special thanks. Each of my committee members served me a different function for which I was both delighted and grateful. Beyond my committee it gets tough because I want to put everyone first for they all contributed to this thesis and its completion in their own special way. My parents gave me their pride. Joyce gave me her ear and made me feel it was worthwhile. Steve, Ed, and Nicky bolstered my confidence (and my alcohol consumption). Kris, Bruce, and Kruce Arms restored my sanity. Nancy, Bill, Paul, Ron, Mary, Kathy and the entire second floor kept tabs on my progress and cheered me on. Suzy answered my questions and helped me through the maze of graduate school memos. Willa, my typist, was incred- ibly patient with my unwillingness to proofread with any speed. Although this work has only one name on it, it was not accomplished by just one person. All these people, plus many others -- graduates, undergraduates, faculty and friends -— contributed to the final product. As a part of Project Castle this thesis was supported by Grant 90-0-635 to ii iii Michigan State University, Office of Child Development, Department of Health, Education and Welfare, and the data was collected through the efforts of Karen Cottledge, Melanie Williams, Chris Kesserling, Jim Day, Ellen Hadley, Helen Short, Ken Zgraggen. and Mike Zgraggen. Special thanks goes to M. Mark Miller for getting me started in graduate school and making me feel like I had a future. TABLE OF CONTENTS CHAPTER ONE - THEORETICAL AND RESEARCH PERSPECTIVE Dominance/Deference Nurturance/Succorance Independence/Dependence CHAPTER TWO - METHODS Overview of Procedures Dominance/Deference: Gives Orders Nurturance/Succorance: Needs Support Independence/Dependence: Makes Plans Training Procedures The Sample Reliabilities Analyses CHAPTER THREE - RESULTS Dominance/Deference: Gives Orders Post Hoc Analyses: Gives Orders Program Types Broadcast Time Viewing Preference Nurturance/Succorance: Needs Support Post Hoc Analyses: Needs Support Program Type Broadcast Time Viewing Preference Independence/Dependence: Makes Plans Post Hoc Analyses: Makes Plans Program Type Broadcast Time Viewing Preference iv 12 16 21 25 25 27 29 33 36 37 38 42 47 48 52 52 60 63 68 68 74 so 85 87 96 9L: Summary CHAPTER FOUR - DISCUSSION APPENDICES Appendix A- General Instructions for the Use of Coding Forms Appendix B - Viewer Training Packets Appendix C - 1975-76 Sample of Television Programs Appendix D - t-values and Standard Deviations REFERENCES 97 102 113 126 133 136 162 TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE 10: 11: 12: 13: 17: : Means, : Means, : Means, LIST OF TABLES Intercoder Reliability Estimates Means, t-tests, and z-scores: Orders Category, All Shows Means, t-tests, z-scores: Order Types by Program Type Means, t-tests, z-scores: Order Receivers and Followed by Program Type Means, t-tests, z-scores: Order Types by Broadcast Time Means, t-tests, z-scores: Order Receivers and Followed by Broadcast Times Order Means, t-tests, z-scores: Types by Viewing Preference Means, t-tests, z—scores: Order Receivers and Followed by Viewing Preference Means, t—tests, z-scores: Needs Support Category, All Shows Means, t—tests, z-scores: Physical Support Types by Program Type Means, t-tests z-scores: Emotional Support Types by Program Type Means, t-tests, z-scores: Support Asked and Responded to by Program Type Means, t-tests, z-scores: Support Given and Respondents by Program Type t-tests, z-scores: Physical Support Types by Broadcast Time t-tests, z-scores: Emotional Support Types by Broadcast Time t—tests, z-scores: Support Asked and Responded to by Broadcast Time Means, t-tests, z-scores: Support Given and Respondents by Broadcast Time vi #0 51 54 55 58 59 61 62 67 69 7O 72 73 75 76 78 79 TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TMABLE 18: 19: 20: 21: 22: 23: 24: 25: 26: 27: 28: 29: D1: D3: D5: D6: vii Means, t-tests, z-scores: Support Types by Viewing Preference Means, t-tests, z-scores: Support Asked, Responded to, Given and Respondents by Viewing Preference Means, t-tests, z—scores: Plans Category, All Shows Means, t—tests, z-scores: Makes Plans and For Others by Program Type Means, t-tests, z-scores: Plan Executors and Outcomes by Program Type Means, t-tests, z-scores: Make Plans and For Others by Broadcast Time Means, t-tests, z—scores: Outcomes by Broadcast Time Plan Executors and Means, t-tests, z-scores: Makes Plans and For Others by Viewing Preferences Means, t-tests, z-scores: Plan Executors and Outcomes by Viewing Preference Direction of Difference: Orders, Plans, Support: t-tests and z-scores: All Shows Direction of Difference: t-tests and z-scores: Orders, Support, Plans, All Shows: Outcomes Direction of Difference: t-tests and z-scores: Orders, Support, Plans, All Shows and Outcomes t-values and Standard Deviations: Orders Category, All Shows t-values and standard deviations: Order Types by Program Type t-values and standard deviations: Orders Receivers and Followed by Program Type . t-values and standard deviations: Order Types by Broadcast Time t-values and standard deviations: Orders Receivers and Followed by Broadcast Time t-values and standard deviations: Order Types by Viewing Preference 82 83 86 88 89 92 93 95 96 105 106 107 136 137 138 139 140 141 TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE D7: D8: D9: D10: D11: D12: D13: D14: D15: D16: D17: D18: D19: D20: D21: D22: D23: viii t-values and standard deviations: Orders Receivers and Followed by Viewing Preference t-values and standard deviations: Support Category, All Shows t-values and standard deviations: Physical Support Types by Program Type t-values and standard deviations: Emotional Support Types by Program Types t-values and standard deviations: Support Asked and Responded to by Program Type t-values and standard deviations: Support Given and Respondents by Program Type t-values and standard deviations: Physical Support Types by Broadcast Time t-values and standard deviations: Emotional Support Types by Broadcast Time t-values and standard deviations: Support Asked and Responded to by Broadcast Time t-values and standard deviations: Support Given and Responded to by Broadcast Time t-values and standard deviations: Support Types by Viewing Preference t-values and standard deviations: Support Asked, Responded to, Given and Respondents by Viewing Preference t-values and standard deviations: Plans Category, All Shows t-values and standard deviations: Makes Plans and For Others by Program Type t-values and standard deviations: Plan Executors and Outcomes by Program Types t-values and standard deviations: Makes Plans and For Others by Broadcast Time t-values and standard deviations: Plan Executors and Outcomes by Broadcast Time 142 143 145 146 147 148 149 150 151 152 153 154 155 156 157 158 1539 ix TABLE D24: t-values and standard deviations: Makes Plans and For Others by Viewing Preference 160 TABLE D25: t-values and standard deviations: Plan Executors and Outcomes by Viewing Preference 161 CHAPTER ONE THEORETICAL AND RESEARCH PERSPECTIVE With the advent of the women's movement, content analysis has been used in a variety of media to demonstrate the existence of narrowly defined categories of appropriate behavior for men and women. The purpose of content analysis is to specify in a systematic manner the characteristics of a variable of interest. It is generally done to find out "what's there" in a systematic fashion that enables a researcher to make general statements about what constitutes the content of a particular medium. Children's literature, print and television advertising, news treatment, and television programming have been subject to extensive content analysis of sex role portrayals (see Busby, 1975, for a comprehensive review of sex role content analyses). These efforts have produced the general finding that mass media create and perpetuate sex role stereotypes. Although the emphasis has been on the "negative" or "tradi- tional" images of women, researchers have also found a "traditional" stereotype for men as well (e.g., Reeves & Greenberg, 1977, document the limited dimensions on which children differentiate television characters, male and female). The process of stereotyping as articulated by Bowes (1976) zuqd by Carter (1962) involves first the homogenization of an 2 image. Homogenization results when the characteristics of a situation become increasingly associated with, and predictable from, each other. The homily "A woman's place is in the home" demonstrates such homogenization. Homogenization may be followed by the pglarization of characteristics or attributes when they are held in the extreme. Polarization of sex roles occurs when one sex can only be characterized by the polar opposite of the other sex, as in "Men are strong: women are weak." Finally, if the image persists over time it has fixedness. The image becomes stable and resistant to change. Although researchers have demonstrated that stereo- typing is not necessarily a negative process (e.g., Gage, 1952), studies of mass media stereotypes have viewed the stable and polarized images of the sexes as being less than desirable. This concern with stereotyped sex role models stems mainly from the work done by Bandura and others (e.g., Mischel, 1966; Bandura, 1965) demonstrating the effective- ness of observational learning. For many years, differences in the behavior of boys and girls were attributed to innate biological functions. IWore recently, however, the focus has shifted from physio- logical sources of difference to socializing sources of (iifference. Behaviorists have attributed differences in ‘behavior to differential reinforcement of male and female “behavior. Bandura (1971) has departed from traditional :reirfibrcement theory by highlighting the role of 3 observational learning as a source of socialized behavior differences. Bandura's early work with the imitation of aggressive film models (e.g., Bandura, Ross & Ross, 1963) has helped spark the current concern over, and research into, television violence. While stressing the importance of direct exper— ience and reinforcement in the acquisition and performance of behavior repertoires, Bandura has theorized that much of behavior is learned by seeing how others are reinforced for their actions (Bandura, 1971). According to Bandura (1971) learning may be achieved in one of two ways: through direct experience or through observation. Extending a traditional focus on direct experience and reinforcement, social learning theory incor- porates the ability to acquire response repertoires through modeling. As with traditional learning theory, reinforce- ment plays a major role in observational learning. Reinforcement may also be direct or observed. Direct reinforcement occurs when a personis rewarded or punished for performing a behavior. Observed or vicarious reinforce- ment occurs when a person observes a model being rewarded or punished for the performance of a behavior. In general, vicarious reinforcement promotes faster learning than direct Ireinforcement, but not long term responsiveness. Direct and observed reinforcement interact to provide a person with zan.observed standard for judging the relative value of ciirect reinforcement the person receives (Bandura, 1971, p. 25). L: The concern over television role models has arisen due to the effectiveness of vicarious reinforcement in acquiring and regulating behaviors. It is the most obvious medium available for observational learning. Television provides attractive and vivid models from which to learn. The positive or negative reinforcement of a television character's behavior may affect a viewer's tendency to imitate the model's actions. The focus here is on the out- come of the behavior available for imitation. The behavior itself is relatively neutral. It is the reinforcement, or outcome, contingent on the behavior, that increases or decreases the likelihood of imitation by the observer. Behaviors which produce positive outcomes for models are more likely to be imitated than behaviors which produce negative outcomes. The work with filmed and televised violence has shown that violent acts which receive positive reinforcement or have positive outcomes are more likely to be imitated (e.g., Bandura, Grusec, & Menlove, 1966; Bandura, 1965). The stereotyping of sex roles on television is more subtle than the stereotyping of the positive outcomes of violence. This may be partially attributed to the nature of interactions between, and characteristics of, the sexes. Violence is inherently more "action-packed," and there- :fore more obvious than the portrayal of men and women on 'television. The subtlety may also be due to the nature of tflie sex-typing process. Mischel (1966) has articulated a social learning 5 interpretation of the acquisition and performance of sex- typed behaviors. He theorizes that boys and girls learn behavior repertoires for both sexes. During the learning process, however, they also learn to label which behaviors are appropriate or inappropriate for their respective sex due to "the difference in outcome as a function of the performer's sex" (p. 60). Mischel also stresses that response consequences need not be obvious or direct. They may also be inferred. Just as Mischel states "a man does not have to be arrested for wearing a dress in public to learn the consequences for such behavior" (p. 61), a little girl receiving more atten- tion when wearing a party dress can learn that females should be interested in their appearance without receiving a direct compliment for being dressed up. Television, by providing relatively homogeneous role models in many areas, facilitates the labeling process Mischel describes. This homogeneity of portrayal has been demonstrated in terms of occupational portrayals and status (Simmons, Greenberg, Atkin, & Heeter, 1977; Miller & Reeves, 1976), minority role portrayals (Simmons, gt al., 1977), pro- and anti-social behaviors (Gerbner & Gross, 1976; Feshback, 1972), demographic characteristics (Simmons, gt §§L., 1977; Tedesco, 1974; Katzman, 1972), and sex role gportrayals (Busby, 1974; Streicher, 1974; Downing, 1974). Using the definition of stereotype provided earlier, 'television is a major medium for stereotyping. The homo— ganeity has been demonstrated. Polarization is also evident 6 in television portrayals. Given time constraints which force plot lines and characters into simplistic and easily contrasted roles, polarization is necessary. The good guys wear white hats; the bad, black. Men are strong; women, weak (Tedesco, 1974). The continuing production of such roles provides television with the fixedness sufficient to perpetuate stereotypes. Content analyses of sex roles have not been conducted long enough, however, to document their fixedness. It can only be assumed to have existed in the past. The fixedness of violence has more support. And, with the continuing assessment of violence on television (Greenberg, Atkin, Edison, Korzenny, Heald, & Wakshlag, 1977; Gerbner & Gross, 1976), the medium itself is being stereotyped as antisocial. In the context of social learning theory, the persis- tence of such homogeneous role models is expected to have an effect on viewers. The outcomes of male behavior are different from the outcomes of female behavior. The effect of this vicarious reinforcement of stereotyped role models is to provide viewers with a homogenous (stereotyped) behavior repertoire, especially in performance areas where the observer has no previous direct experience with the role model or the behavior. This effect has been most clearly demonstrated in the area of occupational roles. Miller & Reeves (1976) have shown that for children, viewing traditional television occupational roles (i.e., polarized by sex), decreases the 7 perceived appropriateness of cross-sex occupational role portrayals (e.g., a female police officer). On the other side of the coin, children who view counterstereotyped role portrayals see the cross-sex occupational status as being more appropriate. It should be noted that television pro- vides many occupational role models with which children have little or no direct experience (Simmons, et al., 1977). Content analysis of television programming for sex role stereotypes is a relatively recent phenomenon. The majority of studies on television sex roles have been pub— lished since 1974 (e.g., the Journal 9: Communication, Spring, 1974, published a symposium on women). It is interesting to note that content analytic studies of other media tend to precede those done on television (e.g., Chase, 1972; Courtney & Lockeretz, 1971; Nilsen, 1971). However, since researchers turned their attention to television, they have left no area of programming untouched. Educational television, news, prime time, day time, cartoons, and adver— tising have all been subject to the scrutiny of vigilant coders. Game shows appear to be the only program type left to be analyzed. Sex role variables tend to fall into five general categories in television content analysis. Most studies look into many of these categories at the same time to :provide a general picture, or profile, of the program type(s) 'under study. These categories are: (1) head counts, (2) demographic characteristics, (3) occupation, (4) physical characteristics,(5) personality traits. 8 (1) Head Counts. It is almost prerequisite to tally the number of male and female characters appearing on tele- vision in one form or another. Busby (1974) found a male to female ratio of 2.5 to 1 in cartoons. Miller & Reeves (1976) report their results in percentages for prime time program- ming: 72% male; 28% female (a ratio of 2.57 to 1). Other studies (e.g., Turow, 1974; Sternglanz & Serbin, 1974) report similar findings for prime time and cartoon programming. Men and women appear with almost equal frequency in daytime serials and as product representatives in television adver- tising (Katzman, 1972; Courtney & Whipple, 1974). Also included in this area of analysis is the number of men and women appearing in major and minor, or regular and supporting roles (Miller & Reeves, 1976; Busby, 1974; Courtney & Whipple, 1974; Long & Simon, 1974). The percentages in this area generally correspond to overall male/female percentages. (2) Demographic Characteristics. Common to many studies is the recording of such variables as age, race, marital status, and parental status. Long & Simon (1974) report that more than half of the women on children's and family (programs are married and more than two-thirds of the married ‘women have children. Tedesco (1974) reports similar findings .in.prime time programs, with women also being reported as younger than men . Other studies reporting women being younger than men irnzlude Dominick & Rauch (1972, advertising), and Downing (197Wq daytime serials). The resultant profile of a 9 television female is a woman who is most likely youthful, married, and a mother. Race is not a significant variable, due mainly to the small percentage of non-white television characters, particularly non-white females (Simmons, gt gt., 1977; Downing, 1974). (3) Occupation. Although occupation can be considered a demographic variable, it merits a separate category due to the large amount of attention it has received. Almost every study of television sex roles includes some form of occupational analysis. The basic finding is that few women have jobs on television. Tedesco (1974, prime time) reports 40% female employment; 64% male employment. Long & Simon (1974) report 18% female employment in children and family programs. Downing (1974,day time serials) reports 80% female employment; 89% male employment. Her figures, however, include the category "housewife" which accounts for almost a third of female employment in daytime serials. When they are portrayed as being employed, the women are most likely to be single and confined to a much narrower range of job types than men. Men are more likely to have professional or managerial positions, while women are most likely to have in-home or clerical positions (Miller & Reeves, 1976; Long & Simon, 1974; Katzman, 1972). (4) Physical Characteristics. The variables coded as Iflrysical characteristics have included physical attractive- rness, clothing, body traits (e.g., weight, height), physical atrtivities, and size and color (in cartoons). Long & Simon (197W) found that the majority of women they coded were 10 tall, thin, attractive, and well-dressed. Busby (1974) found females to be smaller than males in cartoons. Tedesco (1974) reports women to be more attractive and less physically active than men. (5) Personality Traits. Tedesco (1974) and Busby (1974) have done the most extensive coding of personality traits. Tedesco had coders rate each character on 15 semantic differential scales. Busby used 40 semantic differential scales. Other studies have coded characters on a smaller range of traits (e.g., Miller & Reeves, 1976; Long & Simon, 1974; Hennessee & Nicholson, 1972). Traits receiving the most research attention have been independent/dependent, dominant/submissive, strong/weak, and intelligent/unintelligent. Men are rated more positively on these variables than women. Other polar opposites on which men are rated positively and women negatively are: violent/ peaceful, rational/emotional, and active/passive. (The first trait listed is considered the "positive" end of the variable.) By way of summary, these studies have delineated the characteristics of male and female television role models. While Bandura, Mischel, and others place emphasis on the Inodel's characteristics in the facilitation of response imitation, the outcomes (reinforcement contingencies) of the model's behavior play an important, if not more important, :role in social learning. A relatively small number of studies have looked into tflie amount and kind of reinforcement, direct or inferred, 11 that television sex role models receive for their actions. Turow (1974) observed advising and ordering interactions between the sexes on daytime and prime time television. Turow used "episodes" of advising and ordering (treated as an aggregate) to study "the relationships between knowledge, activity, and sex of characters" (p. 138). He found that television men and women conform to traditional stereotypes in their advising and ordering behavior. This stereotyping of knowledge points to the develop- ment of a sex role standard, by which viewers may judge the relative value of the direct reinforcement received for their own advising and ordering. Women, by the television standard, espouse typically "feminine" knowledge, e.g., in categories such as love, the family, and the home. Men exhibit knowledge in "masculine'areas, such as business, law, and government. Turow has demonstrated Mischel's and Bandura's concepts of inferred or indirect reinforcement of televised models by showing the relationship between a character's sex and his/her espousal of a certain kind of knowledge. . Sternglanz & Serbin (1974) recorded a wide range of behaviors found in television cartoons. They went a step further in their coding by recording the consequences of the behavior (consequences were scored as positive, neutral, or negative). The outcome of male behavior was more often positive than negative or neutral. The outcome of female “behavior was more often negative or neutral. Sternglanz .& Serbin interpret these neutral outcomes to mean "that in ,general their behavior had no environmental consequences" (1% 713)- 12 The content analysis undertaken for this thesis attempted to follow interactions between television characters (both same sex and between sex) from the initial portrayal of the behavior through to its outcome. Based on previous re- search (especially, Sternglanz & Serbin, 1974), three cate- gories of behavior were chosen as representative of behaviors most often stereotyped by sex, culturally and in their tele— vision portrayal. These categories are: Dominance/deference, Nurturance/succorance, and Independence/dependence. DominanceZDeference Dominance is a broad concept which has been applied to instances of bullying, influence attempts, toughness, direct- ive behavior, machiavellianism, and leadership (see Maccoby & Jacklin, 1974, for a comprehensive review). In general, there are no sex differences in dominance, although male children tend to be slightly more dominant in terms of influence attempts than female children (e.g., Sutton-Smith & Savasta, 1972). With adults, it appears that only a general statement can be made about dominance. Maccoby & Jacklin (1974) point out that men generally have higher formal status as well as generally higher status (p. 262). Due to this generally higher status, men tend to take dominant roles more often. Although dominance is not a sex-typed trait in the “real world," content analysis of television programming sfluows that dominance exists as a sex-typed trait for male 13 and female characters. Turow (1974) reports that 70% of the advising and ordering episodes recorded during prime time television were initiated by men. That is, men gave 70% and women gave 30% of the advice and orders directed to a member of the opposite sex. Turow found that daytime programming is not as blatantly sex-typed on this variable, although men initiated 56% of the episodes recorded. Sternglanz & Serbin (1974) recorded both dominance and deference in children's cartoons. Males were not shown to be significantly more dominant than females, but females were much more deferent than males. The definitions formu- lated by Sternglanz & Serbin served as a starting point in operationally defining dominance. Dominance. To influence or control others, to per- suade, prohibit, dictate. To lead and direct. To restrain. To organize the behavior of a group. Deference. To follow directions or example (imitate) of a leader (except to ridicule) - either willingl or under threat. To admire or compliment. (p. 712 These definitions were found to be too broad, including forms of behavior that may not necessarily be dominant or de- ferent. In doing the content analysis for this thesis, verbal dominance, in the form of directive, or ordering, behavior was recorded. Given the simplistic nature of role portrayals on television, occupational status confers greater authority on men than on women (Miller & Reeves, 1976; fPurow, 1974: Katzman, 1972). With this in mind, two general ‘types of directives were identified according to the role fromlwhich the directive was given: authority orders and 14 simple orders. Authority orders are directives given by a person in authority to a subordinate. This authority may be con- ferred on the person by his/her occupational status, e.g., a doctor ordering a nurse; by the nature of his/her position as a social agent, e.g., a police officer ordering a citizen; or by parental status, as when a parent orders a child. Simple orders are directed toward another peer. Peers are defined as persons interacting with equal status along any dimension, e.g., marital, as husband/wife; social, as friends; occupational, as co—workers. It was found after initial testing of the scheme that threats can sound very much like orders. Therefore, a threat order category was added to the analysis. Threat orders are directives with statements concerning bodily harm attached. Tedesco (1974) and Simmons, gt gt. (1977) found that men are more often cast as criminals than women. Therefore, men would be more likely to use threats as a form of dominance. Due to the deference demonstrated by women in previous content analyses, it was felt that women would be more likely to explain or justify their directives. Such ex- planations serve to modify the directive,tone it down in essence, to improve its chances of being followed or carried out. Consistent with the non—authoritative female stereo- type, women television characters would be expected to explain or justify more orders than men. General hypotheses concerning the giving of orders 15 were formulated based on these conceptions. Due to the un- equal proportion of male to female television characters, hypotheses, in all categories, are stated with the assump- tion that "more" or "less" is used with respect to the rela- tive frequency of male and female characters on television. That is, chance occurrence is based on doing a behavior more or less than the expected proportion of occurrence. H1: Male characters will give proportionately more authority orders than female characters. H : Male and female will give proportionately equal numbers of simple orders. H3: Male characters will use proportionately more threat orders than female characters. H4: Female characters will explain proportionately more of their orders, authority or Simple, than male characters. In keeping with social learning theory, the outcomes of the directive behavior were recorded. In this analysis, the outcome of an order is the degree to which the order was followed. A character whose orders are followed is positive- ly reinforced for being dominant. A character whose orders are not followed is negatively reinforced for being dominant. In line with the sterotyping of television sex roles, it is expected that more male orders will be followed than female orders, regardless of the type of order given: H5: Orders given by male characters will be followed proportionately more often than orders given by female characters. Of interest in these dominance interactions is also the sex of the character being ordered. Turow (1974) restricted 16 his analysis to cross-sex orders. Sternglanz & Serbin (1974) recorded dominance and deference as separate behaviors. In this analysis the sex of the character receiving the order was recorded regardless of the sex of the order giver. Viewing directive behavior as an interaction between a possibly dominant and possibly deferent character ties the two forms of behavior together. With this perspective it is possible to assess whether female characters are more often cast in roles deferent only to male characters, or whether they are cast in roles deferent to dominance in general. It is hypothesized that females will be more often cast as deferent to both sexes: H6: Proportionately, male characters will order other male characters more often than female characters will order male characters. 7. Female characters will be the receivers of orders proportionately more than males Will be the receivers of orders, regardless of the sex of the order giver. Nurturance/Succorance Murray (1938) used the terms nurturance and succorance to describe giving aid to others and receiving aid from others. Nurturance was considered a response to succorance, and the succorant person was considered weaker, younger, and dependent. Nurturance is generally considered to be care- taking behavior. It has been almost exclusively labeled a female role behavior, and is often used synonymously \Nith "maternal" behavior. Maccoby & Jacklin (1974) review psychological research studying nurturant behavior by men and women. Included in the review is a discussion of altruism, a concept similar to 17 nurturance. Maccoby & Jacklin describe the body of research on altruism as dealing with "helpful, supportive behavior that a person may direct toward a variety of other persons" (p. 221). The emphasis in this research is most often on voluntary offers of aid or voluntary assistance to another. Sears, Rau, & Alpert (1965) defined nurturant behavior as "voluntarily guiding or assisting another with the intent of being helpful or performing a service" (in Maccoby & Jacklin, 1974, p. 223). The definition used in this content analysis combines aspects of the "maternal" and the "altruistic" characteristics of nurturant behavior. The person in distress need not necessarily be weak or young, nor is the nurturant behavior given only on a voluntary basis. Deaux (1976) details a number of studies concerning male and female personality traits. A consistent finding emerging from her review has been that, culturally, women are considered to be dependent and emotional, while men are considered independent and active. Deaux also comments that television is a force in creating and perpetuating stereo- types. Maccoby & Jacklin (1974), however, could find no sex differences in their review of studies on nurturant behavior. The stereotyping of nurturant/succorant behavior on television has been recorded systematically in only one content analysis. Sternglanz & Serbin (1974) recorded rnxrturance and succorance as separate behaviors: 18 Nurturance. To nourish, aid or protect a defenseless other. To express sympathy. To "mother" a child. To give information to help carry out a project (see Succorance). Succorance. To seek aid, protection, sympathy, infor- mation to help carry out a project. To cry for help. To plead for mercy, to adhere to another person for security. To be dependent. (p. 712) They found male characters to be significantly more succorant than females, but there was no sex difference for nurturance. As with dominance/deference, these definitions were found to be too broad due to the inclusion of information seeking behavior, and were later defined more explicitly in their operational definition. Other studies touch on nurturance/succorance only peri- pherally. Long & Simon (1974) report that a traditional view of women is presented in television commercials in that women are dependent and have socio-emotional family roles. Busby (1974) reports that women are more emotional, fragile, and dependent on others, while men are more adventuresome, sturdy and self-reliant. Turow (1974) reports that women in daytime and prime time programming are most knowledgeable in "feminine" areas such as personal problems, the family, and the home. Men display more knowledge in areas like business, crime, and coping with danger. The resultant research profile of nurturance/succorance is consistent with the stereotypes held in this culture. (Deaux, 1976; Donelson, 1975; Broverman, Vogel, Broverman, Clarkson, & Rosenkrantz, 1972, all discuss a cultural stereo- type of "warm" females and "competent" males). If television 19 stays true to the cultural stereotypes of active, adventurous men and dependent, emotional women, there will be differences in the ways men and women are portrayed in terms of succorance. Male television characters will be more likely to find them- selves in physical danger, while female characters will be more subject to emotional distress. The findingsof Sternglanz & Serbin (1974) appear to conflict with this line of reasoning (i.e., males, in general will be more succorant). They, however, made no distinction between physical and emotional succorance. They also restricted their analysis to seeking aid, which eliminates the previ— ously defined succorant "condition" wherein aid might not be requested. A succorant person, for the purposes of this analysis, need only evidence a need for supportive behavior, i.e., be in trouble, have a problem. Nurturance is the response to that succorant need. It may be given voluntarily or it may be asked for. A nurturant response is a response with which another person attempts to relieve the danger or distress. A nurturant respondent need not be successful to evidence nurturance. Very simply, a person in a succorant condition has a problem and a person who responds to that problem is being nurturant. A person znust be in a succorant condition in order for nurturance to occur. Although it would be expected that nurturance would be 21 female activity, the overabundance of male characters on 'television may negate any sex differences in the response 20‘ to succorant characters. Males may also be more likely to respond to typically male problems, i.e., coping with physi— cal danger. It is, however, hypothesized that female charac- ters will do more than their share of nurturing others, especially those who find themselves in emotional trouble. H8: Male characters will be portrayed in physically succorant conditions proportionately more than female characters. 9. Female characters will be portrayed in emotionally succorant conditions proportionately more than male characters. H : Female characters will respond to succorance with 10 . nurturance proportionately more than male charac- ters. As with dominance/deference interaction, nurturance/ succorance is also considered in terms of its outcomes. In this instance, the recorded behavior is the succorant condi— tion and the outcome is the nurturant response. A nurturant response giving aid to the succorant character is positive reinforcement for needing aid or expressing a need for aid. A response not giving support to succorance is negative re- inforcement. If female characters are nurtured only when in emotional distress, and male characters only when in physical danger, the characters are being differentially reinforced for be- havior according to the sex of the performer. H11. Male characters will be nurtured (receive positive ' reinforcement) for physical succorance more than female characters. H12: Female characters will be nurtured (receive posi- tive reinforcement) for emotional succorance proportionately more than male characters. 21 Independence/Dependence The purpose in this study for recording instances of independent behavior stems from previous content analyses in the area. Male and female television characters have been rated on their independence as a personality trait (Busby, 1974; Tedesco, 1974; Long & Simon, 1974), but not as a behavior. Sternglanz & Serbin (1974) defined two categories of behavior similar to the concept of independence: Achieve- ment-construction and Autonomy. Achievement—construction. Planning and carrying out one's own plans, building, stating desire to overcome obstacle, to surpass self or others. Autonomy. To resist influence or coercion. To defy an authority or seek freedom in a new place (emigrate). To strive for independence. (p. 712) They found male characters to be more constructive than female characters. There was no sex difference for autonomy. Their definitions are again too broad and also include an element of dominance. The intent here was to define a di- mension of behavior which showed independence that need not be closely associated with dominance. Defining independence poses a difficult task. The problem with independence is that it is rarely defined. Most often its polar opposite, dependence, is defined and studied (e.g., Maccoby & Jacklin, 1974; Mischel, 1966). .About the most that can be said about independence is that it is not being dependent. Although previous research (e.g., Busby, 1974; Tedesco, :22 1974) has shown male characters to be more independent than female characters, the characteristics of such behavior have not been defined. In searching for a way to assess the many shades of meaning independence can take on, a number of concepts were considered that appear to be indicative of independence. This culminated in the development of a category labeled "makes plans." Making plans was originally conceived as a decision making category. Decision making is a part of being able to plan actions. In order to plan one must decide on a course of action to be implemented and also decide how to imple- ment that action. Making plans also incorporates the notion of leadership in that not all plans may be carried out alone. A planner may have to solicit the aid of others in order to implement that plan. The concept of leadership is most often associated with small group behavior. It is not necessary, however, to lead a group to show independence of thought, autonomy, or constructive behavior. Creativity and originality of thought are also elements of plan making. Making a plan and attempting to carry it out show a measure of being able to think on one's own without seeking the advice or opinions of others. Plan Inaking also shows independence with the realization by the lean maker that something needs to be done. For this analysis, then, independence is defined as 'the ability to recognize the need for, and decide upon, 23 a plan of action when it is called for. "Makes plans" attempts to tap this category of behavior by defining a plan as a statement of a method for doing something. The typical assumption has been that girls show more dependency than boys, and research has shown that girls are indeed a bit more dependent (Mischel, 1966). This assump- tion infers the reverse as well, that boys will be more independent or less dependent than girls. Given the stereo- typic portrayals on television, it is hypothesized that male characters will exhibit the independence of thought and action defined as inherent in plan making more than females. 13 Male characters will make plans proportionately more than female characters. H For this category, the behavior outcome is the success or failure of the plan. Positive reinforcement for indepen- dent behavior is seeing a plan through to its completion. Negative reinforcement is having a plan fail. If female characters are portrayed in traditional roles, they will not be positively reinforced for independent behavior and, therefore, more of their plans will fail. H14: Male characters will formulate proportionately more ultimately successful plans than female characters. In summary, the basic focus of this study is on the behaviors and outcomes that may show sex role stereotyping on television. The behaviors recorded in this analysis were chosen on the basis of previous contest analysis studies, as well as research into cultural attitudes towards sex roles. 24 This study attempts to go beyond previous research into television sex roles by incorporating social learning theory into the content scheme. Social learning theory emphasizes not only the behaviors available for imitation, but also the outcomes of these behaviors. If differential reinforcement of the sexes on television is present as hypothesized, viewers are being presented with stereotyped sex role models. CHAPTER TWO METHODS Overview of Procedures The three major categories for the sex role content analysis presented here were developed, tested, revised, and implemented at the same time. Conceptually, however, they were treated as, and still remain, distinct categories requiring separate consideration. The development of each category scheme will be presented as a separate procedural task. This overview is presented here to provide a basic outline for the development and execution of the sex role content analysis scheme as a whole. From the conceptual definitions of the major categories, operational definitions were developed. For each conceptual category, broad operational categories were defined and types of behavior within each category were delineated: Dominance/ deference was operationalized as giving orders; Nurturance/ succorance as needing support; Independence/dependence as making plans. Following the broad category definitions, behavior sequences for each category were outlined. These sequences zattempted to logically structure the expected occurrence of 'behaviors conceptually following the occurrence of the main cnategory behaviors. The behavior sequence for giving orders 25 26 was outlined: who gave what kind of order; to whom was it given; and, was the order followed? The behavior sequence for needing support was outlined: who needed what kind of support; if someone responded, who was it; and was support given? The behavior sequence for making plans was: who made what kind of plan; was it for him/herself and/or others; who carried out the plan; was it a success or failure? The con— sequences of each behavior sequence were also considered a part of the logical structure. With a rough form drawn up for each category scheme, another graduate student,M. Mark Miller (now a faculty mem- ber of the School of Journalism, University of Wisconsin, Madison), and I tested the scheme by viewing portions of videotaped television programs. Changes and additions were made in areas where we found difficulty in making decisions regarding the recording of behaviors and in the use of the forms. The scheme was tested a second time for five to six weeks by two undergraduate students. After consulting with them about the scheme as well as their training, the coding forms and variables were revised a final time in order to incorpor- ate improvements suggested by their work with the scheme. Revisions of the scheme will be noted in the following dis- cussion of the variables coded and their operational defini- tions. A new group of coders was then trained in the final *version of each category scheme. This group did the formal Iwecording of sex role behavior sequences. Training procedures 2'2 will be outlined following the operational definition of each category. DOMINANCE/DEFERENCE: GIVES ORDERS As previously stated, the Dominance/deference behavior category became the orders category in the content analysis scheme. Giving an order was defined as the beginning of a Dominance/deference behavior sequence. The television char- acter is the unit of analysis; therefore, all definitions are given in terms of what the character does. Characters were defined as persons, animals, or things with speaking parts in a television program. The sex, male or female, for each character was recorded. In this category, giving an order is defined in this fashion: GIVES ORDERS: The character gives a directive for others to do, say, or think something. Orders are coded accord- ing to how the directive is or is not modified by the ordering character. The order types are mutually exclu- sive and exhaustive. Originally, three types of orders were defined. During the initial testing of the scheme we had difficulty distin- guishing some kinds of threats from orders. Before pretesting the scheme we added a category Threat Orders to the scheme. Order Types: Threat: An order given with a statement that physical harm will result for the character being ordered if it is not complied with. Example: Get in the car or you won't know what hit you. Authority: An order to be complied with because of occupational position (e.g., boss), social agent (e.g., police officer, nurse), or parent. If a character has been explicity made a delegate of any of the above he/she is capable of giving an authority order. Example: Get back to work. 28 Simple: An order given among equals or peers; husband/wife, brother/sister, friends, co-workers, etc. An order given by someone in an authority posi- tion may be considered a simple order if the characters are interacting as peers, e.g., in a special setting. An order is simple unless clearly given as a threat or an authority order. Example: Hurry up! Explained: Any of the above orders may be further modified by the inclusion of a justification for why an order should be followed. This justification must be made immediately prior to, or following the giving of an order. Examples: Threat Explained: I get nervous when I have to hold a gun. So if you make a sudden move you might get hurt. Authority Explained: Be quiet, Jane. You're disturbing the other students. Simple Explained: Come back here. I want to tell you why I said that. Other variables in the behavior sequence: Receiver: The receiver in the orders behavior sequence is the character(s) to whom the order was given. Followed: An order was considered to have been followed if the receiver carried out the order as it was given. If the order was not carried out by the receiver as given, the order was not followed. Consequences: Consequences were defined as any action or verbal statement that happened as a result of the order being followed or not followed. They were coded as present or absent, with space for comments as coders found appropriate. Coding Forms An example of an Orders coding form and directions for its use may be found in Appendix A. In general, letter codes (e.g., "A" for authority orders) were used for record- ing order types. "Check columns," columns which required a check mark or an "X", were used for recording the sex of 'the character and whether the order was followed. All order 29, types, followed and not followed orders, and order conse- quences were briefly described by a quote, paraphrase, or action following any letter code or check column. These descriptions were provided as a check on coder accuracy. A final column named "Notes" was included in the form during its final revision. Coders could write in comments about each behavior sequence if they encountered trouble in re- cording that behavior sequence. Behavior sequences were viewed and recorded in "time segments." Initially, a time segment was two minutes long. The first time segment began with the start of the program. Time segments were numbered consecutively, used to organize coding, and to record when the order was followed. Coders found it too difficult to operate a time clock along with their coding duties. In place of a two-minute time segment, they used the counter on the videotape machine and viewed programs in time segments of 50 as determined by the counter. A videotape count of 50 on the machines used roughly approx- imates two minutes. NURTURANCE/SUCCORANCE: NEEDS SUPPORT Needing Support, the second general category, was defined as the beginning of a Nurturance/Succorance behavior sequence. As with Giving Orders, the character is the unit of analysis. It was defined as: NEEDS SUPPORT: A character is in danger or distress. This category does not include routine requests for Assistance or social courtesies. It does include non- routine requests or needs which are relevant to program plots, subplots, or character development. The support types are not mutually exclusive and may, therefore, oc- cur in any combination during character interaction or 30 scene. They are exhaustive in terms of the range of behaviors to be coded. Originally, four types of need for support were defined: Physical Internal, Physical External, Humiliation, and Inabi- lity to Cope. During initial testing, Physical Confinement was added to the physical needs for support. Physical Con- finement was added because involuntary confinement fell into neither of the two physical support types. Psychological support and Concern for Others were added to the emotional needs for support during the second revision of the scheme when our coders told us they had difficulty fitting these needs into the existing emotional needs for support. The decision to add these support types was made on the basis of viewing the segments of television programs in which it was apparent that a character was in a succorant condition for which we had no category. The distinction between routine and non-routine requests for assistance was made during the final revision of the scheme. The coders had become overenthusiastic in their work and were including statements such as "Can you tell me the time?" as needs for support. The routine/non-routine distinc- tion focused the coding on our original variables of interest. Support Types: Physical External: A character is in danger of being killed, injured, or beaten. The threat of physical harm comes from outside the character. Example: A character is in danger of being caught in a cave-in, landslide or other natural dangers. Physical Internal: A character is suffering from a disease, illness or internal malady. The threat of 31 physical harm comes from within the character. Example: A person has cancer. Physical Confinement: Characters are jailed, trapped, or held against their will. One character's movements have been restricted by another. The confinement is involuntary. Example: Robbers lock people in a bank vault. Inability to Cope: A character states that he/she has a problem that he/she cannot solve; is in need of ego support. The source of emotional distress or self-inadequacy for the character comes from within the character. Example: A person can't get along with boss, parents, spouse, etc. Humiliation: A character states a fear that he/she will be disliked or held in low esteem by others. The character is afraid of being "put down." Example: A person fears that someone will reveal that they are homosexual, have a criminal record, or other socially unacceptable characteristic. Concern for Others: A character discusses help for a friend, relative, or associate with a third person (person needs support because someone else is in trouble). Note that at least three people are in- volved: The person expressing concern, the person to whom concern is expressed, and the person in trouble. Example: Person notes that someone is late and expresses worry that he/she is lost. Psycho Support: A character has a problem because of the actions of others but does not express inabil- ity to cope, fear of humiliation, or concern for others. Example: Person's spouse has left them. Another set of support categories was added during the final revision at the request of the coders. The following categories are occurrences that are not essential to plot development; support is easily given; needs are momentary and easily resolved within a scene or interaction. These categories were developed to help coders make a conceptual distinction between pertinent interactions and behaviors, and less relevant interactions. Compliments: A character receives praise, compliment, or encouragement when not requested or particularly needed. Example: You did that very well. 32 Approval: A character seeks confirmation of ideas, opinions, or actions. Example: Do you like my new dress? Disappointment: A character is visibly upset, de- pressed, or blue because efforts are unsuccessful. The emotional setback is temporary, almost fleeting. Example: What do you mean you can't meet me at the airport? The above variables do not appear in later analyses and discussions due to their use as a coder aid and their essentially "non-problem" (not involving distress or danger) status. The other variables in the behavior sequence are: Asks Support: A character in need of support may ask or not ask for aid in relieving the need. Therefore, coder identification of a support need is not depend- ent on the character asking for help. Respondent: A respondent is defined as a character who recognizes that another character is in need of support. The responding character shows in some way, through physical and/or verbal action, that he/ she knows that another character has a problem. A respondent need not provide support in order to be identified as a respondent. Support Given/Not Given: Support is given when the respondent attempts to provide aid to relieve that particular need for support. Support is not given when the respondent does not or cannot provide the aid necessary to relieve the character's need. Consequences: Any behavior that results from the attempt to give aid or from the absence of aid is defined as a consequence. They were coded as pre- sent or absent, with space for comments as coders found appropriate. Coding Forms An example of a Support coding form and directions .for'its use may be found in Appendix A. As with the orders cnategory, letter codes (e.g., "PI" for Physical Internal) xvere used for recording support types. Check columns and 33 describe columns were used in a parallel fashion as well. Coding of support behavior sequences was also organized by time segment. INDEPENDENCE/DEPENDENCE: MAKES PLANS The final general category of behavior is the Plans category. Formulation of a plan was defined as the begin- ning of an Independence/dependence behavior sequence. Again, the character is the unit of analysis. MAKES PLANS: A description of a method for achieving a goal. Making a plan is not merely a statement of intent to do something (e.g., I'm going to make an appointment). It is the statement of a method for doing something (e.g., I'm going to make a doctor's appointment to see if I can do something about my sinus headache). The plan must be explicit in state- ment or action. The categories are considered mutually exclusive and exhaustive. The "other" category is to be used only if the plan cannot be classified in any other way. Originally, a very small number of plan types were defined for the analysis. After the final revision, however, 14 different plan types had been defined within the scheme. This wide variety of plan types reflects the diverse nature of the television programs coded. The plans behavior sequence was the most difficult for the coders to record and the large number of plan types both helped and hindered their efforts. They helped by making it easier for a coder to recognize a plan. They hindered by making the scheme more cumbersome than the Orders and the Support schemes. Examples for euach plan type are not given because the type of plan is gin example in itself. A sample plans coding for and directions .for'its use may be found in Appendix A. 34 Plan Types: Makes Appointments: A character states that he/ she will set up a time to meet with someone for a purpose. Housework or Maintenance: A character states that he/she will clean, repair, or fix-up, etc., for a specified reason. Social Affairs/Family: A character states he/she will arrange an outing, party, etc., for family members or relatives. Social Affairs/Non-family: A character states he/she will arrange an outing, party, etc., for others outside of the family, or for the family and out-siders (non-relatives), e.g., friends, co-workers. Business Deals: A character states that he/she will prepare a contract, or arrange a meeting in order to conduct business. Strategies: A character states how he/she will achieve a goal through movement and/or placement of other people. Criminal Activities: A character, identified as a criminal, states how he/she intends to carry out an illegal action. Criminal Apprehension: A character states how he/she intends to catch a criminal, or stop an illegal action. Construction: A character states how he/she will build or put together something. Acquisition: A character states how he/she intends to get something legally that he/she doesn't presently have. Rescue: A character states how he/she will attempt' to find, release, or rescue someone who is trapped, lost, captured, etc. Behavior Intent/Personal: A character states how he/she will do something for him/herself, e.g., I'm going to take a walk so I can clear my head and get some exercise. The Executor (carries out): 35 Behavior Intent/Interaction: A character states how he/she will do something with another person, e.g., I'm going right out there and straighten him out. A future interaction is explicit in the statement of method. Other: If a statement of method cannot be class— ified by any of the above categories, it is coded as an "other" then described. other variables in the behavior sequence are: For Whom: A plan may be formulated for execution by, and the benefit of, the character making the plan, for other characters, or for both the plan— making character and others. Complexity: The complexity of the plan was rated on a scale of 1 to 5 as defined below. Complexity was added during the final revision in order to free coders from repeatedly recording a very com- plex plan. 1 = Very Simple, immediately executed; instruc- tions implicit in the plan, e.g., I'll go get the mop to clean up that mess. 2: Simple, executed within the same time seg- ment; very simple set of instructions, e.g. Officer Smith, you cover the back door; Fred and I will break in through the front and catch the burglar. 3 = Moderate, instruction necessary for execu- tion which takes place within one or two time segments. It advances the story line but is not a major contribution to the plot, e.g., plans for a surprise party which re- quire moving and hiding people, coordinating lights, etc. 4 = Complex, large set of instructions; carries across more than two time segments. Some of the plot will revolve around the plan and execution, e.g., a bank robbery plan that requires time, people, blue prints, etc. 5 = Very Complex, requires a major portion of the show to plan and execute. Most of the plot will revolve around the plan and its execution, e.g., a Mission Impossible plot. The executor is the character who carries out the plan. The executor may be the character who originally made the plan, or another character acting as executor. Execution: Our coders felt the need to record whether the execution of the plan was seen or not seen. They 36 found that many plans are formulated but that their execution and outcome are only implied. If the execution of the plan is portrayed during the program, it is seen. If implied within the plot or assumed by the characters, but not por- trayed, the execution is not seen. Consequences: The consequences of formulating and executing a plan are its success or failure. Some plans both succeeded and failed, sometimes more than once. Each success and/or failure of a particular plan were recorded. TRAINING PROCEDURES Coder training consisted of a series of discussion and practice sessions. Before viewing any programs, the coders discussed the variables with the investigators in order to become familiar with them. The coding procedure may be sum- marized as follows: Discussion of conceptual definitions for each category. Discussion of operational definition (category types and other variables) for each category. Instruction in the use of coding forms. . Practice with the schemes. Discussion of problems and questions concerning the schemes. \n-P‘Ki) NH This sequence was repeated a number of times over a two week period of training. Coders worked in pairs, initially to discuss what they were recording. Following the train- ing period they continued to work in pairs but were coding independently for reliability purposes. Coder training packets were developed during the training period to act as aids in the recording of behavior sequences. These may be found in Appendix B. The criterion for the coders to meet was to identify and record behavior sequences as defined. This meant that the group of coders as well as their trainers had to agree upon the identification and recording of sample behavior 37 sequences. Agreement between the two investigators was reached during the development and initial testing of the schemes. It was necessary then to bring the coders into agreement with us, with each other, and with the definitions given in their training packets. The reliability discussed pin a later section reflect how well the criterion was met. The formal coding was done by a group of four under- graduate, female coders. They were all work study students. It would have been most desirable to have a mixed-sex group of coders. We attempted, but were not able, to hire an equal number of male and female coders. The final group of coders was all female. One male coder was involved in initial training sessions but left the group before formal coding began. The formal data collection was conducted during the months of January to June, 1976. THE SAMPLE Data were collected on one videotaped episode each of all prime time (8-11 p.m.) and Saturday morning (8 a.m. - 12 p.m.) regular network programming. Selected midseason replacement shows (e.g., "One Day at a Time") were also coded. Replacement shows were selected mainly on the basis of having female lead characters. The original sample of shows was aired and videotaped during a four-week period in the fall of 1975. Midseason replacements were videotaped in early 1976. Not included in the sample of shows coded were variety 38 shows,documentaries or news programs, movies, and special programs. The total number of shows coded, 79, represents 59.5 television hours. Characters included in the final analysis were all characters with speaking parts. A demographic analysis (Simmons, Greenberg, Atkin, & Heeter, 1977) of this sample of television programs identified 1212 speaking characters; 885 were males (73%) and 327 were females (27%). It was possible for the coders to record nonspeaking characters (e.g., a character receiving an order who doesn't speak). In such cases only the sex was recorded and used in analysis. Characters with nonspeaking parts who exhibited a main category behavior (e.g., needing support) were not included in the final analysis. It was also possible for groups of people to be recorded (e.g., a group of police officers receiving an order). If all the characters in the group were of one sex it was re- tained in the analysis and treated as one character. Mixed groups were not retained in the analysis. Groups were not capable of exhibiting a main category behavior by definition and so were not recorded. RELIABILITIES Reliabilities were calculated on all behavior sequence variables coded and retained in the analysis. About half of the sample programs were double coded, i.e., two coders viewed the same show independently. Intercoder reliabili- ties were computed as follows: 39 1. All variables were indexed for each character identified by a coder as having exhibited a main category behavior. 2. The indices were created by summing across each _ variable type recorded by a coder for a particular character. For instance, the number of authority orders recorded by a coder for one character be-: came the authority orders index for that character. All dichotomous variables (e.g., orders followed/ not followed) were broken out and treated as two variables (e.g., the number of orders followed and the number of orders not followed). 3. Coders were randomly designated as Coder A or Coder B. The variable index for Coder A was then correl- ated with the variable index for Coder B. The cor- relation coefficient for each variable was used as the estimate of intercoder reliability. Coder A was then used for the formal analysis. Reliability estimates may be found in Table 1. Behaviors recorded as consequences, with the exception of a plan's success or failure, were dropped from the analysis before reliabilities were calculated due to their extremely low frequency of occurrence. The reliability estimates show that the Orders category was the category within which there was the most agreement between coders on all variables. The Support category shows the second highest level of intercoder reliability. The Plans category is the least reliable of the three. This is not too surprising. The Orders category is the simplest category conceptually and operationally. It had the fewest number of main category variables to be coded as well as the fewest number of behaviors in its sequence. The Plans category was the most complex conceptually and operationally. Initial reliabilities showed less than marginal reliability for the identification of a plan type. Also, the total number of plans recorded was rather small 40 Table 1 Intercoder Reliability Estimates Category: Orders N = Variable authority orders simple orders authority explained simple explained female receivers male receivers orders not followed orders followed all orders Category: Support N = Variable physical internal physical external physical confinement ego support concern for others psycho support support not asked for support asked for support not responded to support responded to female respondents male respondents support not given support given all physical support all emotional support all support Plans N = Variable Category: plans plans made for self plans made for females plans made for males simple complexity moderate complexity high complexity female executor male executor plan success plan failure 277 characters 294 characters I‘ .69 .67 -75 .50 .74 .48 .54 .63 .67 .68 .79 .63 .66 .66 .81 .60 .68 133 characters r .74 .14 .43 .65 .27 .26 .34 .31 .61 ~59 41 in comparison to the two other main categories. Because of this, plan types were collapsed into one variable, Plans. The reliability estimate (.74) for this variable shows that while coders could not agree on the type of plan a character was formulating, they did agree that the character was making a plan. All variables coded in the Orders and Support categories were retained in the analysis. Threat and Threat Explained Orders had been dropped before reliability estimates were computed due to their extremely low frequency. Although the Support variables Ego Support and Psycho- logical Support show, at best, marginal reliability (.50 and .48, respectively), they have been retained due to their importance to theanalysis. Ego Support is a composite vari— able. Humiliation and Self-inadequacy were collapsed into one variable because of their low individual frequencies and their conceptual similarity. Four variables were dropped from the Plans analysis due to extremely low reliability; Plans made for self and the three complexity variables. Although complexity was originally a five-point scale, initial frequencies showed that coders were using basically a three-point scale. Very Simple and Simple plans (1 and 2), were combined into Simple Complexity. Complex and Very Complex (4 and 5) were combined into the High Complexity variable. Moderate Complexity (3) remained the same. Two Plans variables were retained in the analysis in 4.2 spite of their low reliability estimates: Plans made for other females (.43) and female executors (.31). These vari- ables will be retained because the corresponding male vari- ables reach an acceptable level of reliability (.65 and .61, respectively), and because of their conceptual interest for this study. ANALYSES As with most content analyses, the data collected in this analysis are nominal. This level of measurement does not allow for comparison of results across categories. The creation of additive indices (frequencies) within categories changes the structure of the data, however, from a nominal level to an interval level. As long as inferential statis- tics assuming an interval level of measure are computed only for the indices within categories, the results are meaningful. The hypotheses generated in Chapter One suggest that there are two populations of interest in this content anal- ysis. First, there is the group of characters who exhibit at least one main category behavior. An example of an hypo- thesis concerning this population is: Male characters will need physical support more than female characters. In general, this type of hypothesis asks whether there is any difference between the sexes in the number of behaviors each sex exhibits. Assuming a normal distribution for this population, the t-test for difference of means will answer the questions these hypotheses pose. The results of these t-tests will show whether there is a difference in the per 43 character rate of behavior exhibition. Or, more simply, a significant t-test of these hypotheses will allow the infer- ence that the rate (or average number per character) of a behavior by one sex is very different from the rate of behav- ior by the other sex. The means in each category of behavior will show which sex has a higher rate. The second population of interest is the entire popula- tion of television characters with speaking parts. The characters who exhibit main category behaviors are a subset of this larger population. This larger population is not normally distributed. The number of characters found in each main category (691, Orders; 613, Support; 212, Plans) is at best half of the total sample under consideration. The distribution of all characters, then, is skewed toward zero. An inferential statistic assuming a normal distribution would be mislead- ing in this instance. The hypotheses generated about this population also ask a different question than those of the subset of characters who exhibit at least one main category behavior. These hypotheses were formulated with the unequal representation of the sexes on television in mind (males, 73%, females 27%). An example of this type of hypothesis is: In general,physical support needs will be proportion- ately overrepresented as a male behavior and proportion- ately underrepresented as a female behavior. In general,this type of hypothesis asks whether the sexes are performing behaviors in proportion to their rep- resentation in the total population. The expected frequency 44 of behaviors performed by each sex is proportional to the expected frequency of males and females in the population. The statistic used to test these hypotheses is a z- statistic. This z is a normal approximation of the binomial distribution and similar to a Chi-square with 1 degree of freedom (see Hays, 1963, p. 585, for a full discussion of this statistic.) No population distribution assumptions are made. The formula for this 2 is: - f fol e1 V/N 01 observed frequency of female acts where: f fe1 = expected frequency of female acts (27% of total) fe2 = expected frequency of male acts (73% of total) N = total number of acts This z not only provides a test of difference of propor- tions, it also gives an indication as to whether one group is overrepresented or underrepresented in the population. A negative z-score for this test will indicate that female behaviors are proportionately underrepresented. A positive z-score shows that female behaviors are proportionately overrepresented. The reverse is true for male behaviors. That is, a negative 2 means that male behaviors are over- represented; a positive z that male behaviors are under— represented. 45 In summary, then,two statistics have been computed for these data: the t-test, which tests the difference in mean rate of behavior performance by the sexes, and the z—test, which tests the difference in proportion of behavior performance by the sexes. A number of exploratory post hoc tests were also per- formed on the data. The main hypotheses concern television programming in general and do not take into account the different types of programming or time periods available. (A list of television programs coded may be found in Appen- dix C.) These post hoc tests were done in three breakdowns of the data. The data were broken down by: type of programming, broadcast time, and viewing preferences. As defined in the previous demographic analysis of this sample (Simmons, gt gt., 1977) three program types were analyzed: situation comedies, action adventure shows, and other show types, which include family and medical shows. Saturday morning shows could have been split into cartoons and noncartoons for this category. This split was not made because the number of shows, and therefore, characters, falling into these categories became too small for meaningful analysis. The broadcast time of the shows was split into three time periods: Saturday morning, 8—9 p.m. (popularly known as "the family hour"), and 9-11 p.m. Viewing preferences were split into two categories: preferred shows (Top 40 viewing) and non—preferred shows (Low 60 viewing). This split was made on the basis of a 46 survey conducted in the Lansing area in the fall of 1975 with children in the fourth, sixth, and eighth grades. Shows viewed weekly by 40% or more of the sample were desig— nated preferred shows. The remaining shows were labled non- preferred shows. The results of this data analysis will be presented as follows: Each main category (Orders, Support, Plans) will be presented as a separate analysis, with the results of the general (all shows) t- and z-tests presented first. Post hoc test results will be presented within each main category. A summary table presenting general data from all three cate- gories will be discussed at the end of the results chapter. CHAPTER THREE RESULTS The analysis of the data, as discussed in the previous chapter, provides two kinds of evidence for each hypothesis to be tested. The test of difference of means (t-test) com- pares the average number of behaviors performed by male and female characters in each content category. The difference of proportions test (z-test) compares the extent to which behaviors performed by each sex are representative of the extent to which each sex is present in the total sample of TV characters. Negative z-scores indicate that male behav- iors are overrepresented, while positive z-scores indicate that female behaviors are overrepresented in the sample of television characters. Consideration of both types of evidence should not only provide support or non-support for the hypotheses generated in the first chapter; it should also give an indication of the extent to which sex role stereotyping is present in the sample of TV content. This chapter is organized around the three content categories: Dominance/deference, Nurturance/succorance, and Independence/dependence. Within each category results are presented for the hypotheses discussed in Chapter One, 47 MB; followed by results of the post hoc analyses for program type, broadcast time, and viewing preference. DOMINANCE/DEFERENCE: GIVES ORDERS The results of the "all shows" analysis of the Orders category appear in Table 2. Table 2a presents the results of the Order Types analysis; Table 2b, the Receivers analysis; Table 2c, the Orders Followed analysis. Table 2a shows that, across all shows, male characters gave nearly three times as many orders per male character as did the female characters ("all orders"). Thus for each male, the rate of order giving is much higher than for each female. The large and negative 2 indicates that, furthermore, the total number of orders given by females was much less than would have been expected from the distribution of male and female characters in the population. The distribution of order—giving across all shows is 20% female and 80% male. That is, 20% of the orders given were given by females. Both tests, then, support the hypothesis that, in general male television characters will give more orders than female television characters. This result holds regardless of the type of order a character gives. It was hypothesized that male and female characters would be equal in their order giving behavior for Simple and explained orders. These hypotheses were not supported by the data, as can be seen by the significant t-tests and large and negative z-scores in Table 2a. 49 Negative z-scores indicate that male behaviors are over- represented in the sample of television characters. All differences are statistically significant beyond the .05 level. Support is provided in Table 2b for those hypotheses concerning receivers of orders. In the orders behavior sequence, the male-male sequence occurs three times as often as a female-male sequence. That is, in ordering other characters, males order other males at a higher rate than females order males. Males also order other males in a higher proportion than would be expected from the popula- tion distribution. The male-male sequence occunsin 82% of the orders behavior sequences. When females are on the receiving end of an order, the rate issflightly lower (closer to 2:1), but the proportion of females being ordered is nearly that of what would be expected in the population. The insignificant z-score shows that females receive orders from both sexes in proportion to their presence in the population. Although the average number of female receivers is quite different for males and females, this could be a function of the larger number of orders given by males. The effectiveness of an order, whether it is followed or not followed, is demonstrated in Table 2c. An effective order (one that is followed) occurred three times more for a male order giver than for a female order giver. That is, the rate of followed orders was much higher for males than 50 for females. The large and negative 2 shows that the total number of orders followed was not distributed as would be expected in the population, with males having more than their share of orders followed (82% of the orders followed were given by males). In the case of an ineffective order (one that was not followed), the male rate is again much higher than the female rate. The proportion of ineffective orders given by males and females conforms, however, to the proportion expected from the population distribution of males and females. Therefore, support is found for the hypothesis that orders given by male characters will be followed more often than orders given by female characters, but only in terms of the rate of occurrence. In summary, the findings for the Orders category are these: -Males have a higher rate of order giving than females, regardless of the type of order given; -The total number of orders given by males is much greater than would have been expected from the distribution of males and females in the population; —On the average, males are the receivers of orders given by males more than receivers of orders given by females; -Females receive more orders on the average from males than from females, but in proportion to the expected distribution of males and females in the population; -Orders given by males are followed at a higher rate than those given by females. -Male orders are also not followed at a higher rate than female orders, but the proportion of orders not Table 2 Means, t-tests, and z-scores: female N male N 294 395 Table 2a: Orders Category, All Shows Order Types Significance .~ Females* Males of t 2 Authority Orders 0.27 1.33 (.0001 -7.68C Simple Orders 0.92 2.12 <.0001 -1.97a Authority Explained Orders 0.06 0.28 <.0001 -3.21b Simple Explained Orders 0.20 0.57 <.0001 —2.42b ALL ORDER TYPES 1.45 4.29 <.0001 —7.190 Table 2b: Receivers Significance of Females Males t z Female Receivers 0.41 0.94 (.0001 -1.42 Male Receivers 0.91 2.99 <.0001 -7.35C Table 20: Orders Followed Yes (followed) No (not followed) .0 ap 5 bp .01 c .0001 P Significance of Females Males t z 0.94 3.17 <.0001 -7.84C 0.53 1.21 (.0001 -1.43 *For all tables, the column females designates the mean. number of behaviors originated by female TV characters. designates the mean number of behaviors for male TV characters. The column male 52 followed for both sexes is not significantly different from the proportion expected from the distribution of males and females in the population. POST HOC ANALYSES: GIVES ORDERS Tables 3—8 present the results of the post hoc analyses for program types (Tables 3-4), broadcast time (Tables 5-6), and viewing preference (Tables 7-8). These post hoc tests are a way of looking for the sources of the large sex differences found in the "all shows" analysis of the Orders category. Program Types Table 3 shows the results of tests done on the types of orders given in Situation Comedies, Action Adventure programs and medical and family dramas (Medfam dramas). Saturday Morning programming is not included in this break- down. Overall, the differences in the average number of orders given by males and females in any program type are small. The male rate for Action Adventure programs for all orders (Table 3b) is almost twice that of the female rate. This may be attributed to the large difference in the average number of authority orders given by males and by females. The other order type which shows a difference in the rate of male and female order giving is in Table 3c. The male rate of simple explained order giving in Medfam Dramas is higher than the female rate. This difference is not large enough, however, to affect the overall rate of order giving in Medfam Dramas. 53 The proportions of orders given by males and females in each of these program types show some interesting contrasts. Table 3a shows that in Situation Comedies, order giving is overrepresented as a female behavior (34%). While the aver— age number of orders given by males and females in Situation Comedies is almost the same, female orders are present in a greater proportion than would be expected in the TV character population. This is particularly true for female simple orders. The large and positive z-score for this variable shows that, in Situation Comedies, simple order giving is a female attribute. On the other hand, Table 3b provides the opposite results for Action Adventure shows. Order giving is a male attribute for all orders (88%), authority and authority explained or— ders (93%) and simple orders (87%). Only in simple explain- ed orders do the proportions of male and female orders equal what would be expected from the population distribution of males and females. Results for the receivers of orders are similar to those for order types. Table 4a shows no difference in the average number of orders received by either sex in Situation Comedies. But, females give orders to other females at a much higher proportion than would be expected in the population (46%). In Action Adventure programs (Table 4b), the results parallel those found for all shows: males (88%) and females (89%) are ordered by males at a higher rate and out of proportion to what is expected in the population. There is no difference in the receivers of orders in Medfam Dramas (Table 40). Table 3: Order Types by Program Type Table 3a: Means, 54 t’teSth Situation Comedies z—scores: Females Males Significance (N=32) (N=73) of t z score Authority Orders 0.69 1.03 n.s. -0.92 Simple Orders 2.59 1.55 n.s. +4.82C Authority Explained Orders 0.13 0.37 n.s. -1.64 Simple Explained Orders 0.69 0.62 n.s. +1.10 ALL ORDER TYPES 4.09 3.56 n.s. +2.84b Table 3b: Action Adventures Females Males Significance (N=34) (Ns152) of t z-score Authority Orders 0.47 1.59 (.01 -7.56C Simple Orders 1.59 2.28 n.s. -6.08C Authority EXplained Orders 0.21 0.32 n.s. —2.42b Simple Explained Orders 0.50 0.43 n.s. -1.24 ALL ORDER TYPES 2.77 4.62 (.05 —9.66C Table 30: Medfam Dramas Females Males Significance (N=28) LNF64) of t z-score Authority Orders 1.39 1.72 n.s. -0.18 Simple Orders 2.04 1.88 n.s. +1.52 Authority Explained Orders 0.21 0.16 n.s. +1.16 Simple Explained Orders 0.18 0.50 <.05 -1.85a ALL ORDER TYPES 3.82 4.25 n.s. +0.58 55 Table 4: Means, t-tests, z-scores: Order Receivers and Followed by Program Type Table 4a: Situation Comedies Females Males Significance (N=32) (N=Z3) of t z—score Female Receivers 1.88 0.96 n.s. +4.94C Male Receivers 1.97 2.23 n.s. +0.30 Yes (followed) 2.53 2.58 n.s. +1.20 No (not followed) 1.63 1.07 n.s. +3.36b Table 4b: Action Adventures Females Males Significance (N=34) (N=152) of t z-score Female Receivers 0.53 0.97 <.05 -4.72C Male Receivers 2.00 3.44 v.05 -8.52C Yes (followed) 1.91 3.43 (.05 -8.66° No (not followed) 0.91 1.32 n.s. -4.61c Table 4c: Medfam Dramas Female Males Significance (N=28) (N=64) of t z-score Female Receivers 1.14 1.42 n.s. -0.20 Male Receivers 2.46 2.55 n.s. +0.89 Yes (followed) 2.38 3.11 n.s. -0.83 No (not followed) 1.54 1.18 n.s. +2.28a 56 The effectiveness of order giving also shows some inter- esting results. While female order giving occurs at a high proportion in Situation Comedies, so does the proportion (40%) of female orders that are not followed (Table 4a). This is also true for female orders not folowed (36%) in Medfam Dramas (Table 4c). In Action Adventure programming, however, male orders are followed at a higher rate and in a higher proportion than would be expected (89%). Male orders are also unsuccessful in a higher proportion (87%), but the rates for unsuccessful orders are about the same for each sex (Table 4b). Broadcast Time The results for broadcast time follow a similar pattern (Tables 5-6). It should be noted that most Action Adventure shows are aired during the 9—11 p.m. time period (Tables 5-6c); and many of the Situation Comedies appear between 8 and 9 p.m. (Table 5-6b). Saturday Morning programming (Table 5a) shows no differ— ence in the rate of all orders given by males and females, but does show differences for authority and authority explain- ed orders. The large and negative z-scores show, however, that regardless of the type of order given, order giving is overrepresented as a male behavior (83%). Table 5b shows a proportional difference for all orders given during 8-9 p.m. (77% male). This may be attributed to the overrepresentation of authority order giving as a male behavior. Otherwise, there is no difference in the average 57 number of the expected proportion of orders given by the sexes. Table 50 shows average and proportional differences in the male direction for order giving in the 9-11 p.m. time period. While Saturday Morning and the 8-9 p.m. periods show no differences in the average number of order received by males and females (Tables 6a & b), males proportionately receive more orders from males in both time periods (83%, 80% respectively). The male-female orders behavior sequence is proportionately overrepresented (88%) during Saturday Morning programming, but not during the 8-9 p.m. period. Table 6c shows somce curious results which may be due to females outnumbering males in the 9-11 p.m. period. While males order both sexes at a much higher rate than females, only the male—male orders behavior sequence is overrepresent- ed (82%) as indicated by the large and negative z-score. There is no difference in the proportions of females receiv- ing orders from males or females. Table 60 shows that the 9-11 p.m. period is the only time period during which the average number of orders follow- ed and not followed is greater for males than for females. While successful orders from males are proportionately over- represented (83%) during the period, there is no proportional difference for orders not followed. Tables 6a & b show no differences in the rates of effectiveness for orders given during Saturday Morning and the 8-9 p.m. period. Male orders are followed in a greater 58 Table 5: Means, t-tests, Order Types by Broadcast Time Table 5a: Saturday Morning z-scores: Females Males Significance (N=25) (N=106) of t z—score Authority Orders 0.12 0.94 (.01 -5.53C Simple Orders 3.00 2.43 n.s. -1.85a Authority Explained Orders 0.04 0.23 «.05 -2.68b Simple EXplained Orders 0.56 0.76 n.s. -2.57b ALL ORDER TYPES 3.72 4.35 n.s. -5.45C Table 5b: 8-9 p.m. Females Males Significance (N=46) (N=152) of t z—score Authority orders 0.83 1.24 n.s. -3.69b Simple Orders 2.39 1.92 n.s. -0.11 Authority Explained Orders 0.22 0.27 n.s. -1.25 Simple Explained Orders 0.50 0.51 n.s. -1.11 ALL ORDER TYPES 3.94 3.94 n.s -2.79b Table 5c: 9 - 11 p.m. Females Males Significance (N=223) (N=132) of t z-score Authority Orders 0.18 1.76 <.0001 -4.66° Simple Orders 0.38 2.11 <.0001 -1.66a Authority Explained Orders 0.03 0.32 (.0001 -1.94a Simple Explained Orders 0.09 0.48 (.0001 -0.49 ALL ORDER TYPES 0.68 4.66 <.0001 -4.36C 59 Table 6: Means, t-test, z-scores: Order Receivers and Followed by Broadcast Times Table 6a: Saturday Morning Females Males Significance (N=25) (N=106) of t z—score Female Receivers 0.36 0.61 n.s. -2.88b Male Receivers 2.64 3.11 n.s. -4.64C Yes (followed) 2.56 3.23 n.s. -5.13C No (not followed) 1.16 1.18 n.s. -2.35b Table 6b: 8 — 9 p.m. Females Males Significance (N=46) (N?157) of t z-score Female Receivers 1.15 0.81 n.s. +0.67 Male Receivers 2.39 2.80 n.s. -3.65b Yes (followed) 2.54 2.90 n.s. —3.49b No (not followed) 1.44 1.13 n.s. 0.00 Table 6c: 9 - 11 p.m. Females Males Significance (N=223) (Ns132) of t z-score Female Receivers 0.26 1.37 <.0001 -1.02 Male Receivers 0.40 3.11 <.0001 -4.53C Yes (followed) 0.43 3.43 <.0001 -5.10C No (not followed) 0.27 1.33 <.0001 —0.59 60 proportion, however, in both time periods (84%, 80%, res- pectively). There is no difference in the proportion of orders not followed during the 8-9 p.m. period, but a greater proportion of male orders are not followed during Saturday Morning programming (81%). Viewing_Preference The results of the breakdown of shows by viewing prefer- ence can be found in Tables 7 - 8. It sould be noted that the viewing preferences of our sample (those shows for which 40% or more of the respondents indicated they viewed regular- ly), tend to conform to broadcast time patterns. That is, many of the preferred shows were aired Saturday morning or before 9:00 p.m. Table 7a shows no difference in the rate of order giving by the sexes, regardless of the order type for preferred shows (Top 40). Table 7b shows a large difference in the rate of order giving for non-preferred shows (Low 60). There are proportional differences, though. Order giving is overrepresented as a male behavior in both tables (78%, 81%, respectively). Only simple orders in preferred shows and simple explained orders in non-preferred shows are present in proportions that would be expected from the popu- lation distribution. Table 8a shows no difference in the rate of orders received by either sex in the preferred shows, but the male- male orders behavior sequence is overrepresented (81%) as indicated by the large and negative z—score. Table 8b 61 Table 7: Means, t-test, z-scores: Order Types by Viewing Preference Table 7a: Preferred Shows Females Males Significance (N=54) (N=185) of t z-score Authority Orders 0.63 1.18 n.s. -4.82C Simple Orders 2.56 1.90 n.s. +0.72 Authority Explained Orders 0.17 0.30 n.s. -2.49b Simple Explained Orders 0.48 0.61 n.s. -2.29a ALL ORDER TYPES 3.83 4.00 n.s. -3.52b Table 7b: Non-Preferred Shows Females Males Significance (N=240) (N=210) of t z-score Authority Orders 0.19 1.47 <.0001 -5.98C Simple Orders 0.55 2.31 <.0001 -3.18b Authority Explained Orders 0.04 0.25 <.0001 -2.28a Simple Explained Orders 0.13 0.52 (.0001 -1.14 ALL ORDER TYPES 0.91 4.55 <.0001 -6.51C 62 Table 8: Means, t-tests, z-scores: Order Receivers and Followed by Viewing Preference Table 8a: Preferred Shows Females Males Significance _ (N=54) (N=185) oft z-score Female Receivers 1.26 0.84 n.s. +1.21 Male Receivers 2.13 2.68 n.s. -3.45b Yes (followed) 2.52 2.94 n.s. -4.07C No (not followed) 1.43 1.15 n.s. -0.13 Table 8b: Non—Preferred Shows Females Males Significance _; jN—240) (N-210) 0 ‘t z-score Female Receivers 0.21 1.03 (.0001 -2.89b Male Receivers 0.63 2.26 <.0001 -5.77C Yes (followed) 0.58 3.37 (.0001 -6.88C No (not followed) 0.33 1.26 <.0001 ‘ -1.82a 63 shows differences in both rates and proportions for orders followed and not followed in non—preferred shows. Male orders are bosh followed and not followed at a higher rate than female orders and are overrepresented as a male behav- ior (84%, 77% respectively). In summary, post hoc findings for the Orders category shows: -Only during Situation Comedies is order giving overrepresented as a female behavior, but that behav- ior is not particularly effective in that female orders are not followed out of proportion to what would be expected in the population. —Action Adventure programming and the 9-11 p.m. time period show the highest rates of male order giving behavior. -In all time periods, male orders are proportion— ately overrepresented as being effective (i.e., fol- lowed). -Saturday Morning programs show the only difference in proportion of orders not followed, with ineffective orders being overrepresented as a male behavior. -Preferred Programs show no difference in the rate of order giving by the sexes, but overrepresent order giving as a male behavior. -Non-preferred programs show a rate and a propor- tional difference in order giving in the male direction. -Preferred and non-preferred programs both overrep- resent male orders as being more effective than what would be expected in the population. NURTURANCE/SUCCORANCE: NEEDS SUPPORT The results of the "all shows" analysis of the second content category, Needs Support, appear in Table 9. Table 9 shows that, across all shows, males were in need of support more often than females, but that needing support occurred as a female behavior in a greater 64 proportion than what would be expected from the population distribution of males and females. The reason for these seemingly contradictory results can be found in Tables 9b and 9c. Table 9b shows that males need physical support, on the average, three times as often as females. That is. the rate of male need for physical support is much higher than that of the female rate. The large and negative z-scores for all types of physical support needs further show that males needed support in a much larger proportion (82%) than would have been expected in the population distribution of males and females. Both tests, then, support the hypothesis that male characters will be portrayed in physically succorant conditions more than female characters. Support is also shown in Table 9c for the hypothesis that female characters will be portrayed in emotionally succorant conditions more than male characters. While there is no difference in average number of emotional support needs of males and females, there is a large difference proportionally. 40% of the emotional support needs are female while female characters constitute only 27% of the population. This difference is indicated by the large and positive z-scores. These differences in need for support hold regardless of the specific type of support needed, be it a physical type or an emotional type. Although not hypothesized, there are clear sex differences in whether the character needing support asked for it 65 (Table 9d). Males did not ask for support at a much higher rate than did females. There is no difference in the aver- age number Of supports asked for by males and females. The difference of proportions test shows, however, that not ask- ing for support was overrepresented as a male behavior (78% of supports not asked for) and that asking for support was overrepresented as a female behavior (38% of supports asked for). Table 9e shows that males both were and were not respon- ded to at a higher rate than females. Table 9f shows that males both were given and were not given support at,a higher rate than females. The difference of proportions test on theses variables shows, however, that female support needs were responded to in a proportion greater (38%) than that expected from the population and support was given to females in a proportion (35%) greater than that expected from the population (27%) of television characters. These tests give only indirect support for the hypo- thesis that females will be nurtured for emotional succor- ance more than males. This is due to the fact that the data were not structured so that a direct test of this hypothesis could be made. It can be inferred that indirect support is shown, however, from the evidence that females are typically portrayed as needing emotional support. Since this is the case, it may be inferred that the responses and the support given is directed mainly toward females in an emotionally succorant condition. 6:6 The same is true for the hypothesis that males will be nurtured for physical succorance more than females. In this instance, however, there is indirect non-support for the hypothesis. The difference of proportions test shows that males were not responded to in greater proportion than would have been expected in the population. Support was not given in proportions equal to those expected from the population. Drawing the same inference, there is no support for this hypothesis. Table 9g shows that the rates of females responding to either sex are about the same. Males,however, average more responses to males in need of support than to females. The large and positive z-scores show that females were responded to by either sex in a greater proportion than were males. These results do not support the hypothesis that females will respond to succorance with nurturance more than males. They do Show, though, that females get a bigger response from both sexes than do males when they need support. In summary, the findings for the Support category show: -Males have a higher rate and proportion of physical need for support. -Females have a higher proportion of emotional need for support. -Asking for support occurred as a female behavior in a proportion greater than expected, while not asking for support occurred as a male behavior in a proportion greater than expected from the distribution of males and females in the population. -Response or non-response and giving or not giving support occurs at a higher rate for males needing support. 67 Table 9: Means, t-tests, z—scores: Needs Support Category, All Shows Significance Females Males of t z-score Table 9a: ALL SUPPORT TYPES 2.16 3.47 <.0001 +3.20b N = N = 357 Table 9b: Physical Support Types Significance Females Males of t z-score Physical Internal 0.21 0.62 <.0001 -2.86b Physical External 0.25 0.85 <.0001 -4.35C Physical Confinement 0.10 0.28 (.0001 -1.81a ALL PHYSICAL SUPPORT TYPES 0.56 1.75 <.0001 -5.450 Table 9c: Emotional Support Types Significance Females Males of t z-score Ego Support 0.68 0.70 n.s. +6.25C Concern for Others 0.34 0.27 n.s. +6.03C Psycho Support 0.58 0.75 n.s. +3.65b ALL EMOTIONAL SUPPORT TYPES 1.60 1.72 n.s. +8.98C Table 9d: Support Asked Significance Females Males of t z-score Yes (support asked) 1.20 1.35 n.s. +7.23c No (support not asked) 0.67 1.66 <.0001 -3.11b Table 9e: Support Responded to Significance Females Males of t z-score Yes (response) 1.61 2.10 <.05 +6.16C_ No (no response) 0.26 0.91 <.0001 -4.57C Table 9f: Support Given Significance Females Males 0f t z-score Yes (support given) 1.16 1.52 .<.05 +5.14C No (support not given) 0.70 1.44 1.0001 -0.94 Table 9g: Respondents Significance Females Males of t z-score Female Respondents 0.39 0.44 n.s. +4.11C Male Respondents 1.14 1.50 <.05 +4.95C 68 Females are responded to and given support in a higher proportion than would be expected in the population, while no response to males was overrepresented. -Response to a female need for support occurs more than would be expected from the population distribution of males and females. POST HOC ANALYSES: NEEDS SUPPORT Tables 10-19 present the post hoc analyses of support needs for program type, broadcast time, and viewing prefer- ence. Program Type Consistent with the "all shows" results are the finding for physical and emotional support types in Situation Come- dies. Tables 10a and 11a show that there is no difference by sex in the rates of needing emotional support, but there is a large difference in needing physical support. These results hold regardless of specific support type needed. Tables 10a and 11a also show that needing physical support is overrepresented as a male behavior (92%), while needing emotional support is a female behavior (41%). Tables 10b and 11b show a similar pattern of results for Action Adventure programs. Females have a much higher rate of needing emotional support. There is no sex difference in the rate of needing physical support. Nonetheless, emotion- al support is again overrepresented as a female behavior (40%) and physical support as a male behavior (81%). Tables 100 and 11c show the same patterns of rate and proportional difference for Medfam Dramas, but not to the extent seen in Situation Comedies and Action Adventures. 69 Table 10: Means, t-tests, z—scores: Physical Support Types by Program Type Table 10a: Situation Comedies Females Males Significance .. (N=38) (N=70) 0f t zgggore Physical Internal 0.03 0.26 (.05 -2.08a Physical External 0.11 0.49 (.05 -2.21a Physical Confinement 0.0 0.09 (.05 -1.74a ALL PHYSICAL SUPPORT TYPES 0.13 0.83 (.01 -3.41b Table 10b: Action Adventures Females Males Significance or (N=40) (N=125) of t z-Score Physical Internal 0.60 0.54 n.s. —.23 Physical External 0.60 0.95 n.s. -2.81b Physical Confinement 0.10 0.34 n.s. -2.68b ALL PHYSICAL SUPPORT TYPES 1.30 1.82 n.s. -3.23b Table 10c: Medfam Dramas Females Males Significance A (N: 40) (N=125) of t z-score Physical Internal 0.26 0.67 n.s. -1.41 Physical External 0.37 0.47 n.s. +.39 Physical Confinement 0.15 1.37 n.s. -4.19C ALL PHYSICAL SUPPORT TYPES 0.78 1.28 n.s. -.49 Table 11: t-test, z-scores: Emotional Support Types by Program Type Table 11a: Ego Support Concern for Others Psycho Support Situation Comedies ALL EMOTIONAL SUPPORT TYPES 3.55 Table 11b: Ego Support Concern for Others Psycho Support Females Males Significance (N=38) (N=70) of t z-score 1.47 1.21 n.s. +3.42b 0.74 0.34 n.s. +4.38C 1.34 1.27 n.s. +2.47b 2.83 n.s. +5.56c Action Adventures ALL EMOTIONAL SUPPORT TYPES 2.70 Table 11c: Ego Support Concern for Others Psycho Support ALL EMOTIONAL SUPPORT TYPES 4.15 Females Males Significance (N=40) (N=125) of t z-score 1.25 0.38 (0.01 +5.49C 0.60 0.29 n.s. +2.34b (3.85 0.60 n.s. +1.08 1.02? (0001 +L1’097d Medfam Dramas Females Males Significance (N=27) (N=51) of t Z—score 1.59 0.86 n.s. +4.87C 0.93 0.49 n.s. +3.47C 1.63 0.94 n.s. +4.45C 2.29 Simple Explained Followed?: u '11 Followed NF Not Followed 128 SUPPORT-CODES Needs Support: Enter as many codes as necessary Physical Internal = Physical External Physical Confinement (involuntary) Inability to Cope (self-inadequacy) Humiliation = Concern for Others Psycho Support Gives Support: PI PE PC CC CO PS Enter in "describe" when applicable Compliments = Approval Disappointment = C A 129 PLANS-CODES Statement of Methods (makes plans): Makes Appointments = A Housework or Maintenance = H Social Affairs Family = SAF Non-Family = SAN Business Deals - BD Strategies = ST Criminal Activities = CAC Criminal Apprehension = CAP Construction = C Acquisition = ACQ Rescue = R Behavior Intent Personal = BIP Interaction = BII Other = 0 Complexity: Consequences: Enter Number Success Very Simple = 1 Failure Simple = 2 Moderate = 3 Complex = 4 Very Complex = 5 130 Definitions for Sex Role Content Analysis ORDERS: Directive for others to do, say or think something. Types of Orders Threat: an order given with a statement that physical harm will be done if it is not complied with. Authority: an order to be complied with because of occupational position (boss), social agent (police- man, nurse, doctor), or parent. In some cases by explicit delegate of the above. Simple: an order given among equals or peers: e.g., husband and wife, brother/sister, friends, co- workers, etc., unless clearly one of the above. Explanation: a justification for why an order should be followed other than threat or authority. 131 Definitions for Sex Role Content Analysis NEEDS SUPPORT: occurs when person is in danger or distress. Does not cover routine request for assistance or social courtesies. Types of Need for Support Physical External: person is in danger of being killed, injured, or beaten Examples: person is about to be shot, knifed, etc. person is in danger of being caught in cave-in, land- slide, etc. person is being chased by potential assailant Physical Internal: person is suffering from disease, illness or internal malady. Examples: person has cancer person has hepatitis android has malfunctioning circuits Physical Confinement: person is jailed, trapped or held against their will. Examples: robbers lock persons in bank vault person is jailed person is trapped in mine cave-in Inability to Qgpg: person states that s/he has a problem that s/he cannot solve; is in need of ego support Examples: person can't get along with boss, parents, spouse, etc. person needs money person has lost something and can't figure out how to find it 132 NEEDS SUPPORT-2 Humiliation: person states fear that they will be dis- liked or held in low esteem by others. Examples: person feels that others will think s/he is dumb, irres— ponsible, or funny looking person fears that someone will reveal that s/he is homosexual, have a criminal record, etc. person fears that others will make fun of s/he Concern for Others: person discusses help for friend, relative or associate with a third person. NOTE: at least three people are involved: the person expressing concern, the person to whom concern is expressed, and the person in trouble. person needs support because someone else is in trouble. Examples: person notes that someone is late and expresses worry that s/he is lost person asks ideas to help a friend who is depressed person seeks assistance in rescuing someone who is trapped or captured by others Psycho Support: person has a problem because of the actions of others but does not express inability to cope, fear of humiliation, or concern for others Examples: person's spouse has left them person's son/daughter has flunked out of school person's dog is causing trouble in the neighborhood The following categories are occurrences that are not essential to plot development; support is easily given; needs are momen- tary and easily resolved. Compliments: person receives praise, compliment or encour- agement when not requested or particularly needed. Approval: person seeks confirmation of ideas, opinions, or actions. Disappointment: person is visibly upset, depressed, or blue because efforts are unsuccessful: a temporary emotional setback. APPENDIX C 1975—76 SAMPLE OF TELEVISION PROGRAMS Name of Show All in the Family Barbary Coast Baretta Barnaby Jones Barney Miller Beacon Hill Bionic Woman* Bob Newhart Bronk Bugs Bunny Cannon Chico and the Man Doc Doctors Hospital Ellery Queen Emergency Emergency Plus 4 Family Holvak Fat Albert Fay Good Times Ghost Busters Happy Days Harry 0 Hawaii Five-O Hong Kong Phooey Isis Invisible Man Jeffersons* Joe and Sons Joe Forrester Top 40 (x) X NNXXX Type Sitcom ActAdv ActAdv ActAdv Sitcom Medfam ActAdv Sitcom ActAdv Cartoon ActAdv Sitcom Sitcom Medfam ActAdv Medfam Cartoon Medfam Cartoon Sitcom Sitcom Noncart Sitcom ActAdv ActAdv Cartoon Noncart ActAdv Sitcom Sitcom ActAdv Time H I I H \o H H I H H Ill H\0H\o H H \O\OCI)\OCD\O\OCD\O I I H 5.1: Sat. 9-11 8-9 8-9 9-11 8-9 8-9 Sat. 8-9 Sat. 8-9 Sat. 8-9 9-11 9-11 Sat. Sat. 8-9 8-9 8-9 9-11 13# Name of Show Josie and the Pussy Cats Kate McShane Kojak Land of the Lost Laverne and Shirley Little House on the Prairie Lost Saucer Marcus Welby Mary Tyler Moore M*A*S*H Matt Helm Maude Medical Center Medical Story Mobile One Movin' On New Adventures of Gilligan Oddball Couple On the Rocks One Day at a Time* Pebbles and Bamm-Bamm Phyllis Pink Panther Police Woman Top 40 NNNN N N Return to the Planet of the Apes Rhoda Rockford Files Rookies Run, Joe, Run Sanford and Son Scooby Doo, Where Are You Secret Lives of Waldo Kitty Shazam Sigmund and the Sea Monsters >4><><>4><>< Type Cartoon ActAdv ActAdv Noncart Sitcom Medfam Noncart Medfam Sitcom Sitcom ActAdv Sitcom Medfam Medfam ActAdv ActAdv Cartoon Cartoon Sitcom Sitcom Cartoon Sitcom Cartoon ActAdv Cartoon Sitcom ActAdv ActAdv Noncart Sitcom Cartoon Cartoon Noncart Noncart Time Sat. 9-11 9-11 Sat. 8-9 8-9 Sat. 9-11 9-11 8-9 9-11 9-11 9-11 9-11 8-9 8-9 Sat. Sat. 8-9 9-11 Sat. 8-9 Sat. 9-11 Sat. 8-9 9-11 9-11 Sat. 8-9 Sat. Sat. Sat. Sat. Name of Show Six Million Dollar Man Space 1999 Speed Buggy Starsky and Hutch Streets of San Francisco Swiss Family Robinson Switch That's My Mama Three for the Road Tom and Jerry/Grape Ape Valley of the Dinosaurs Waltons Welcome Back Kotter When Things Were Rotten *Midseason Replacements 135 Top 40 (X) Type ActAdv ActAdv Cartoon ActAdv ActAdv Medfam ActAdv Sitcom Medfam Cartoon Cartoon Medfam Sitcom Sitcom Time 8-9 8-9 Sat. 9-11 9-11 8-9 9-11 8-9 8-9 Sat. Sat. 8-9 8-9 8-9 Top #0 = Shows viewed regularly by 40% or more of a sample of 4th, 6th, and 8th graders in Fall, 1975. Sitcom = Situation Comedy ActAdv Medfam Cartoon: Animated Cartoon Noncart: Non-animated Cartoon 8’9 : 8-9 pom. 9-11 = 9-11 p.m. Action Adventure Medical/Family Drama Sat. = Saturday morning (8 a.m.-12 p.m.) 136 APPENDIX D t-values and standard deviations Table D1 t-values and standard dev1ations: Orders Category. All Shows (Table 2) Order Type t—value standard deviations females males AUthOFity '6078 1012 2083 Simple ”6043 2023 2067 Authority explained -4.79 0.34 0.80 Simple explained -6.73 0.50 0.91 ALL ORDER TYPES -10.06 2.93 4.48 Receivers t—value standard deviations females males Female receivers —4.25 1.41 1.92 Male receivers -9.11 2.26 3.71 Orders Followed t-value standard deviations females males FOllOWGd ”10025 2001 3063 Not followed -6.27 1.27 1.59 Program Type (Table 3) Situation Comedies Authority Simple Authority explained Simple explained ALL ORDER TYPES Authority Simple Authority explained Simple explained ALL ORDER TYPES Authority Simple Authority explained Simple explained ALL ORDER TYPES 137 Table D2 t-values and standard deviations: Order Types by t-values standard deviations females males ‘0088 1060 2027 +1.74 3.16 1.89 -1084 003,4 1002 +0.38 0.82 1.01 +0.65 3.86 3.87 Action Adventures t-values standard deviations females males '3000 1054 3025 -1.30 2.75 2.94 -0.74 0.77 0.83 +0.57 0.56 0.76 -2.51 3.64 4.87 Medfam Dramas t-values standard deviations females males “0055 1039 3016 +0.27 2.70 2.54 +0.54 0.50 0.41 -2.50 0.48 0.74 -0048 3.82 4025 138 Table D3 t-values and standard deviations: Orders Receivers and Followed by Program Type (Table 4) Situation Comedies t-values standard deviations females males Female receivers +1.39 3.39 2.34 Male receivers —0.44 2.68 3.09 Not followed +1.39 2.08 1.36 Action Adventures t-values standard deviations females males Female receivers -2.28 0.79 1.68 Male receivers -2.26 3.17 4.12 Followed —2.46 3.11 3.83 Not followed -1.50 1.36 1.68 Medfam Dramas t-values standard deviations females males Female receivers -0.63 1.14 1.42 Male receivers -O.12 2.46 2.55 Followed -1.24 2.36 3.11 Not followed +0.82 1.54 1.17 139 Table D4 t-values and standard deviations: Order Types by Broadcast Time (Table 5) Saturday Morning t-values standard deviations females males Authority -3.55 0.44 2.21 Simple +0.75 3.58 2.74 Authority explained -2.19 0.20 0.77 Simple explained -0.99 0.82 1.10 ALL ORDER TYPES -0.72 3.93 4.00 8’9 pom. t-values standard deviations females males Authority -1.13 2.11 2.47 Simple +0.85 3.57 2.42 Authority explained -0.47 0.70 0.81 Simple explained -0.08 0.69 0.75 ALL ORDER TYPES -0.01 4.19 3.85 9-11 p.m. t-values standard deviations females males Authority -5.06 0.82 3.54 Simple '6060 1018 2087 Authority explained -3.94 0.20 0.82 Simple explained -4.63 0.35 0.91 ALL ORDER TYPES -8.10 1.90 5.46 140 Table D5 t-values and standard deviations: Followed by Broadcast Time (Table 6) Female receivers Male receivers Followed Not followed Female receivers Male receivers Followed Not followed Female receivers Male receivers Followed Not followed Saturday Morning Orders Receivers and t—values standard deviations females males -1.27 0.64 1.58 -0-55 3-94 3-56 -o.97 3-03 3-34 -O.97 1-31 1-39 8—9 p.m. t-values standard deviations females males +0.94 2.31 1-59 -o.69 3.60 3.22 -0.73 2.82 3.14 +0.92 2.13 1.42 9-11 p.m. t-values standard deviations females males -4.98 1.17 2.41 -7.01 1.19 4.33 -7.75 1.28 4.34 -6.06 0.85 1.91 141 Table D6 t-values and standard deviations: Order Types by Viewing Preference (Table 7) Preferred Shows t-values standard deviations females males Authority -1.82 1.61 2.90 Simple +1.39 3.24 2.37 Authority explained -1.56 0.42 0.89 Simple explained -1.03 0.77 0.95 ALL ORDER TYPES +0.26 3.82 4.34 Non-Preferred Shows t-values standard deviations females males Authority -6.36 0.97 2.76 Simple -7.69 1.74 2.90 Authority explained -4.03 0.31 0.72 Simple explained -5.92 0.40 0.88 142 Table D7 t-values and standard deviations: Orders Receivers and Followed by Viewing Preference (Table 8) Preferred Shows t—values standard deviations females males Females receivers +1.04 2.76 1.87 Male receivers -1.12 3.01 3.60 Not followed +1.01 1.84 1.48 Non—Preferred Shows t-values standard deviations females males Female receivers -5.71 0.21 1.03 Male receivers -9.06 0.63 3.26 Followed -10.10 0.58 3.37 Not followed -7.06 0.33 1.26 143 Table D8 t-values and standard deviations: Support Category, All Shows (Table 9) ALL SUPPORT TYPES t—values standard deviations females males ALL SUPPORT TYPES -4.64 3.31 3.64 Physical Support Types t-values standard deviations females males Physical Internal —5.19 0.66 1.26 Physical External -6.98 0.72 1.40 Physical Confinement -4.12 0.40 0.68 ALL PHYSICAL TYPES -8.73 1.22 2.14 Emotional Support Types t-values standard deviations females males Ego Support -O.15 1.43 1.64 Concern for Others +1.06 0.93 0.68 Psycho Support -1.50 1.38 1.50 ALL EMOTIONAL SUPPORT -O.54 2.82 2.64 Support Asked t-values standard deviations females males Asked -O.92 2.01 1.81 NOt ASKGd -7006 1027 2018 Support Responded To t-values standard deviations females males Response .203? 2047 2049 No Response -7.43 0.68 1.43 Given Not Given Female Respondents Male Respondents 144 Table D8 (cont.) Support Given t—values standard deviations females males -2.26 1.94 1-93 -6006 1024 1078 Respondents t-values standard deviations females males -0066 0091 0095 -2017 2002 2012 145 Table D9 t—values and standard deviations: by Program Type (Table 10) Situation Comedies Physical Internal Physical External Physical Confinement All Physical Types Action Adventures Physical Internal Physical External Physical Confinement All Physical Types Physical Internal Physical External Physical Confinement All Physical Types Physical Support Types t-values standard deviations females males —2.03 0.16 0.93 —2056 0039 1013 -2018 0000 0033 -3044 0041 1060 t-values standard deviations females males +0.34 0.98 1.14 -1.96 0.93 1.15 -2.66 0.30 0.83 '1081 1038 2013 Medfam Dramas t-values standard deviations females males -1089 0059 1031 -O-49 0-79 0-99 +0.10 0.46 0.40 ”luau 1019 1086 146 Table D10 t-Values and standard deviations: Emotional Support Types by Program Types (Table 11) Situation Comedies t-values standard deviations females males Ego Support +0.57 1.81 2.93 Concern for Others +1.81 1.22 0.74 Psycho Support +0.14 2.44 2.54 ALL EMOTIONAL TYPES +0.86 4.03 4.50 Action Adventures t-values standard deviations females males Ego Support +2.69 1.98 0.82 Concern for Others +1.48 1.26 0.78 Psycho Support +1.20 1.17 1.07 ALL EMOTIONAL TYPES +3.00 2.86 1.69 Medfam Dramas t-values standard deviations females males Ego Support +1.80 1.80 1.48 Concern for Others +1.47 1.41 0.83 Psycho Support +1.85 1.71 1.22 ALL EMTOIONAL TYPES +2.57 3.38 2.21 Responded to by Program Type (Table 12) Asked Not Asked Response No Response Asked Not Asked Response No Response Asked Not Asked Response No Response 147 Table D11 t—values and standard deviations: Situation Comedies Support Asked and t-values standard deviations females males +1.20 2.42 2.32 -0.93 1.61 2.67 +0.79 3.04 3'67 "2075 0'37 0'91 Action Adventures t-values standard deviations females males +2.25 1.73 1.55 +0.21 1.44 1.84 +2.27 2.28 2.17 -1.04 0.92 1.09 Medfam Dramas t-values standard deviations females males +1.91 2.97 1-99 +0.05 1.48 1.45 +1.91 3.26 2.46 -O.62 0.54 0.83 148 Table D12 t-values and standard deviations: Support Given and Respondents by Program Type (Table 13) Situation Comedies t-value standard deviations females males Given +1.31 2.79 2.66 Not Given -1.77 1.06 2.03 Female Respondents +3.03 1.21 0.47 Male Respondents -O.32 1.79 1.97 Action Adventures t-values standard deviations females males Given +2.09 1.58 1.66 Not Given +0.56 1.44 1.66 Female Respondents +0.66 0.82 0.86 Male Respondents +2.32 2.28 1.71 Medfam Dramas t-values standard deviations females males Given +2.11 2.52 2.06 Not Given +0.66 1.51 1.13 Female Respondents +0.94 1.40 0.94 Male Respondents +1.82 2.70 1.82 149 Table D13 t-values and standard deviations: Physical Support Types by Broadcast Time (Table 14) Saturday Morning t-values standard deviations females males Physical Internal -O.62 1.24 1.47 Physical External -0.96 1.35 1.84 Physical Confinement +1.09 0.92 0.71 ALL PHYSICAL TYPES -O.6O 2.03 2.32 8-9 pom. t-values standard deviations females males Physical Internal -1.12 0.72 0.88 Physical External —1.58 0.83 1.18 Physical Confinement -2.43 0.31 0.71 ALL PHYSICAL TYPES -2.45 1.22 1.82 9-11 p.m. t-values standard deviations females males Physical Internal -3.67 0.44 1.34 Physical External -6.62 0.41 1.08 Physical Confinement -3.09 0.17 0.59 ALL PHYSICAL TYPES -6.54 0.73 2.13 150 Table D14 t—values and standard deviations: Emotional Support Types by Broadcast Time (Table 15) Saturday Morning t-values standard deviations females males Ego Support +0.54 1.13 1.11 Concern for Others +1.42 0.71 0.29 Psycho Support +0.27 1.03 0.98 ALL EMOTIONAL TYPES +0.98 1.87 1.74 8—9 p.m. t—values standard deviations females males Ego Support +0.98 1.56 2.34 Concern for Others +1.30 1.26 0.78 Psycho Support -O.57 1.35 2.01 ALL EMOTIONAL TYPES +0.76 2.92 3.64 9‘11 pom- t-values standard deviations females males Ego Support +0.28 1.38 1.00 Concern for Others -O.49 0.82 0.79 Psycho Support -1.09 1.42 1.17 ALL EMOTIONAL TYPES +2.57 3.38 2.22 151 Table D15 t-values and standard deviations: Support Asked and Responded to by Broadcast Time (Table 16) Saturday Morning t-values standard deviations females males Asked +2.34 1.86 1.53 Not Asked -2.32 1.45 2.32 Response -1091 1092 1081 No Response +1.72 1.29 1.96 8-9 pom. t-values standard deviations females males Asked +0.79 2.40 1.92 Response +0.70 2.65 3.06 No Response -2.14 0.68 0.90 9-11 p.m. t-values standard deviations females males Response -2.93 2.34 2.34 No Response -5.40 0.45 1.11 152 Table D16 t-values and standard deviations: Support Given and Responded to by Broadcast Time (Table 17) Saturday Morning t-values standard deviations females males Given -0.05 1.63 1.60 Not Given +0.30 1.35 2.13 Female Respondents -O.74 0.66 1.07 Male Respondents +1.37 2.01 1.37 8—9 p.m. t-values standard deviations females males Given +1.32 2.19 2.32 Not Given -1.64 1.24 1.61 Female Respondents +1.69 1.25 0.86 Male Respondents +0.32 2.26 2.82 9-11 pom. t-values standard deviations females males Not Given -5.17 1.13 1.55 Female Respondents —1.52 2.79 0.92 Male Respondents -2.85 1.84 1.77 153 Table D17 t-values and standard deviations: Support Types by Viewing Preference (Table 18) Preferred Shows t-values standard deviations females males Physical Internal -4.64 0.63 1.42 Physical External -2.32 1.05 1.72 Physical Confinement -2.59 0.38 0.73 ALL PHYSICAL TYPES -4.76 1.40 2.53 Ego Support +2.53 1.96 1.87 Concern for Others +2.74 1.48 0.71 Psycho Support +2.23 2.15 1.13 ALL EMOTIONAL TYPES +3.59 3.90 2.43 Non-Preferred Shows t-values standard deviations females males Physical Internal -2.44 0.67 1.05 Physical External -6.74 0.51 0.94 Physical Confinement -2.85 0.41 0.62 ALL PHYSICAL TYPES -6.41 1.13 1.57 Ego Support -1.24 1.03 1.35 Concern for Others -O.78 0.54 0.65 Psycho Support -3.71 0.88 1.78 1.88 2.84 ALL EMOTIONAL TYPES -2.98 154 Table D18 t-values and standard deviations: Support Asked, Responded to, Given and Respondents by Viewing Preference Asked Not Asked Response No Response Given Not Given Female Respondents Male Respondents Asked Not Asked Response No Response Given Not Given Female Respondents Male Respondents (Table 19) Preferred Shows t-values standard deviations females males +3.11 2.43 1.70 ‘2018 1075 2'53 +2.51 3.03 2-50 -3.68 0.94 1-74 +2.48 2.61 2.01 -1.70 1.53 2.11 +2.04 1.22 1.06 +2.29 2.75 1.98 Non-Preferred Shows t-values standard deviations females males -2.98 1.68 1.92 -6.08 0.96 1.69 -4.05 1.96 2.48 -5-95 0-55 0-97 -4.40 1.41 1.86 -5.54 1.06 1.31 -2.13 0.73 0.81 -4.00 1.50 2.26 t-values and standard deviations: Makes Plans For Females For Males Female Executors Male Executors Success Failure 155 Table D19 Plans Category, All Shows (Table 20) t-values standard deviations females males -1.42 0.92 1.00 Plans for Others t-values standard deviations females males -1.01 0.35 0.53 -1.03 0.75 0-93 Executors t-values standard deviations females males +4.52 0.67 0.20 ”6017 0068 0092 Outcomes t-values standard deviations females males -1011 0087 0090 ‘OoLl'O 0092 0086 156 Table D20 t-values and standard deviations: by Program Type (Table 21) Situation Comedies Makes Plans and For Others t-values standard deviations females males Makes Plans -2.38 0.24 0.62 For Females -O.72 0.33 0.57 For Males -1.42 0.49 0.55 Action Adventure t—values standard deviations females males Makes Plans -0070 1013 1029 For Females +0.85 0.54 0.55 Medfam Dramas t-values standard deviations females males Makes Plans +2.07 0.97 0.67 For Females -1.20 0.32 0.72 For Males +1.52 0.63 0.51 157 Table D21 t-values and standard deviations: Plan Executors and Outcomes by Program Types (Table 22) Situation Comedies t-values standard deviations females males Female Executors +3.28 0.62 0.16 Male Executors -6.17 0.33 0.54 Failure -1.13 0.44 0.49 Action Adventures t-values standard deviations females males Female Executors +1.72 0.79 0.23 Male Executors -3.04 0.76 1.08 Success -1.30 0.79 1.02 Failure +0.55 0.69 0.91 Medfam Dramas t-values standard deviations females males Female Executors +1.76 0.71 0.29 Male Executors -2.75 0.48 0.75 Success -O.46 1.23 0.87 Failure +0.04 0.70 0.59 t-values and standard deviations: 158 Table D22 Others by Broadcast Time (Table 23) Makes Plans For Females For Males Makes Plans For Females For Males Makes Plans For Females For Males Saturday Morning Makes Plans and For t-values standard deviations females males +0.16 1.29 0.96 “2.20 0.00 0038 +0.29 0.88 1.02 8-9 p.m. t-values standard deviations females males -0.82 0.77 0-74 -OOL1’O 0043 0063 +0.53 0.59 0.67 9'11 pom. t-values standard deviations females males '1060 0075 1026 -0.76 0.34 0-54 -2.14 0082 1007 159 Table D23 t-values and standard deviations: Plan Executors and Outcomes by Broadcast Time (Table 24) Saturday Morning t-values standard deviations females males Female Executors +1.72 0.70 0.14 Lale Executors —1.24 1.00 1.01 Success +0.47 0.94 0.84 Failure +0.33 1.56 1.05 8-9 panic t-values standard deviations females males Female Executors +4.0' 0.69 0.13 Male Executors -6.39 0.38 0.72 Success -O.11 1.00 0.88 Failure +0.14 0.62 0.65 9-11 p.m. t-values standard deviations females males Female Executors +1.74 0.62 0.30 Hale Executors —4.32 0.58 1.05 Success —2.59 0.63 0.97 Failure -1.24 0.02 0.84 t-values and makes Plans For Females For Males flakes Plans Standard deviations: by Viewing Preference 160 Table D24 Preferred Shows Makes Plans and For Others (Table 25) t-values standard deviations' females males -0.67 1.08 1.10 -1.52 0.30 0.63 -O.14 0.78 1.03 Non-Preferred Shows t-values standard deviations females males —1.33 0.77 0.86 +0.13 0.39 0.37 -1045 0072 0080 161 Table D25 t-values and standard deviations: Plan Executors and Outcomes by Viewing Preference (Table 26) Preferred Shows t-values standard deviations females males Female Executors +2.82 0.68 0.10 Male Executors -4.14 0.72 0.90 Success —O.61 1.04 0.95 Failure +0.02 1.15 0.84 Non-Preferred Shows t—values standard deviations females males Female Executors +3.42 0.66 0.28 Male Executors -4.58 0.65 0.88 Success -O.96 0.70 0.83 Failure -O.6O 0.67 0.88 REFERENCES 162 References Bandura, A. 1965 Influence of Model's Reinforcement Contingen- cies on the Acquisition of Imitative Responses. Journal of Personality and Social Psychology. 1 9 589'595 ' Bandura, A. 1971 Social Learning Theory. Morristown, N.J.: General Learning Press. Bandura, A., J.E. Grusec, F.L. Menlove 1966 Observational Learning as a Function of Symbolization and Incentive Set. Child Development. 5, 449-455. Bandura, A., D. Ross, S.A. Ross 1963 Imitation of Film-mediated Aggressive Models. Journal of Abnormal and Social Psychology. 66, 3-11. Beuf, A. 1974 Doctor, Lawyer, Household Drudge. Journal of Communication. 24:2, 142-145. Bowes, J.E. 1977 Stereotyping and Communication Accuracy. Journalism Quarterly. 54:1, 70-76. Broverman, J.K., S.R. Vogel, D.M. Broverman, F.E. Clark, P.S. Rosenkrantz 1972 Sex-role Stereotypes: A Current Appraisal. Journal of Social Issues. 28: 59-78. Busby, L.J. 1974 Defining the Sex Role Standard in Commercial Network Television Programs Directed Toward Children. Journalism Quarterly. 51, 690—696. Busby, L.J. 1975 Sex-role Research on the Mass Media. Journal of Communication. 25: 4, 107-131. Carter, R.F. 1962 Stereotyping as a Process. Public Opinion Quarterly. 26, 77-91. Chase, D.J. 1972 Sexism in Textbooks. Nation's Schools. Deon, 31-35 163 Courtney, A.E., S. W. Lockeretz 1971 Woman's Place: An Analysis of the Roles Portrayed by Women in Magazine Advertisements. Journal of Marketing Research. 8:Feb., 92-95 Dominick, J.R., G.E. Rauch 1972 The Image of Women in Network TV Commercials. Journal of Broadcasting. 16:3, 259-265. Donelson, E. 1975 Sex Differences in Developmental Perspective. Homewood Ill.: Learning Systems Company Frueh, T., P.E. McGhee 1975 Traditional Sex Role Development and Amount of Time Spent Watching Television. Develop- mental Psychology. 11:1, 109. Gage, N.L. ' 1952 Judging Interests from Expressive Behavior. Psychological Monographs. 66:18, Whole #320 Gerbner, G., L. Gross 1976 Living with Television: The Violence Profile. Journal of Communication. 26:2, 173-199. Greenberg, B.S., C.K. Atkin, N.G. Edison, F. Korzenny, G.R. Heald, J.J. Wakshlag 1977 Pro-social and Anti—social Behaviors on Com- mercial Television in 1975-76. Project CASTLE: Report to the Office of Child Development, I-IEWO #10 Hays, W.L. 1963 Statistics. New York: Holt, Rinehart, & Winston. Hennessee, J.A., J. Nicholson 1972 NOW Says: TV Commercials Insult Women. New York Times Magazine. May 28, 12-13+. Katzman, N. 1972 Television Soap Aperas: What's Been Going On Anyway? Public Opinion Quarterly. 36:2, 200-212 Long, M., R.J. Simon 1974 The Roles and Statuses of Women on Children and Family TV Programs. Journalism Quarterly. 51, 107-110 ' Maccoby, E.E., C.N. Jacklin 1974 The Psychology of Sex Differences. Stanford, Calif.: Stanford University Press. 164 McNeil, J.C. 1975 Feminism, Femininity, and The Television Series: A Content Analysis. Journal of Broadcasting. 19. 259-269- Miller, M.M., B. Reeves 1976 Dramatic TV Content and Children's Sex-role Stereotypes. Journal of Broadcasting. 20:1, 35-50- Mischel, W. 1966 A Social Learning View of Sex Differences in Behavior. in Maccoby, E.E. (ed.), The Develop- ment of Sex Differences. Stanford, Calif.: Stanford University Press. 56-81. Murray, H.A. 1938 Explorations in Personality. New York: Oxford University Press. Nilsen, A.P. 1971 Women in Children's Literature. College English. May,918-926. Reeves, B., B.S. Greenberg 1977 Children's Perceptions of Television Characters. Human Communication Research. 3:2, 113-127. Sears, R.R., L. Rau, R. Alpert 1965 Identification and Child Rearing. Stanford, Calif.: Stanford University Press Seggar, J.F. 1975 Imagery of Women in Television Drama. Journal of Broadcasting. 19, 273-281. Simmons, K.W., B.S. Greenberg, C.K. Atkin, C.J. Heeter 1977 The Demography of Fictional Television Charac- ters in 1975-76. Project CASTLE: Report to the Office of Child Development. HEW. #2 Sternglanz, S.H., L.A. Serbin 1974 Sex Role Stereotyping in Children's Television Programs. Developmental Psychology. 10:5, 710-715 Streicher, H.W. 1974 The Girls in the Cartoons. Journal of Com- munication. 24:2, 125-129. Sutton—Smith, B., and M. Savasta 1972 Sex Differences in Play & Power. Paper presented at the Annual Meeting of the Eastern Psychological Association, Boston. 165 Tedesco, N. 1974 Patterns in Prime Time. Journal of Commun- ication. 24:2, 119-124. Turow, J. 1974 Advising and Ordering: Daytime, Prime Time. Journal of Communication. 24:2, 138-141.