THE ROLE OF TELEVISED SPORTS IN THE SOCIALIZATION 0F POLITICAL VALUES 0F ADOLESCENTS DISSertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ROBERT H. PRISUTA 1977 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ' 3 1293 10409 9563 This is to certify that the thesis entitled The Role of Televised Sports in the Socialization of Political Values of Adolesgents presente by Robert H. Prisuta has been accepted towards fulfillment of the requirements for Ph.D. Mass Media degree in 7 ,/ / / ’ fl}! {Kc/7 .. Ajf/Zj/A Major professor 0-7639 LIBRARY Michigan State University ABSTRACT THE ROLE OF TELEVISED SPORTS IN THE SOCIALIZATION OF POLITICAL VALUES OF ADOLESCENTS by Robert H. Prisuta Background This study examined the role of exposure to and involvement in televised sports in the acquisition and development of political values in adolescent viewers. Previous research found the family, school, and peer group to be significant socializers of political values, with a similar role ascribed to broadcast television programming, and participation in athletic competition. One conclusion has been that both television and sports participation tend to develop conservative, traditional, and authori- tarian values. These values are also related to socio- economic status and demographic factors. Content analysis of selected televised sports events found these values to predominate. This study tested the premise that levels of con- servative political values would be related to exposure to and involvement in televised sports. Robert H. Prisuta Method Survey research, with correlation and regression data analysis, was utilized. Approximately six hundred high school students in three diverse Michigan school dis- tricts were questioned regarding their television viewing, sports activity, political values, demographic backgrounds, and perceived related attitudes and activities of family, peers, and school. Questionnaire items were primarily Likert scaled. Amount of, preference for, and function of tele- vision sports viewing were considered. Political values surveyed included authoritarianism, nationalism, equali- tarianism, individualism, sportsmanship, need determined expression, and value restraint, plus a general additive measure of conservatism. Zero order correlations were calculated between sports television and political values. Various control variables were used to construct partial correlations. Multiple regression analysis was performed to control for other possible socialization agents and construct a linear socialization model. Factor analysis was used for data reduction and dependent variable construction. Causality is inferential, given early childhood exposure to television, and the relatively late develop- ment of complex political values. Generalizability is indirect and inferential, but can be made with some con- fidence given the representativeness of the sample. Robert H. Prisuta Results Descriptive Televised sports is popular at this age level. Viewing is diverse, with many different sports favored. Viewing is loyal and stable, more so than among adults, and purposive, involving the functions of escape and arousal. High television sports viewers tend to be younger, male, and of higher socioeconomic status. They also tend to participate in sports, attend sports events as spectators, and have families and peers interested in sports. Hypothesis Testing Television sports significantly correlates with certain conservative political values, even when con- trolling for other related factors. It is the best pre- dictor of some of these values, especially social and peripheral values such as authoritarianism, nationalism, and need determined expression, as well as general con- servatism. Where sportsmanship becomes central, via par- ticipation, sports television is not significant. Other- wise, however, the value is related to sports viewing. The person oriented, central values of equalitarianism and individualism were not significantly related to the independent variables. General television viewing is also a strong predictor of conservatism. Perceived con- servatism among traditional socialization agents was also Robert H. Prisuta a consistently strong factor, as was the role of the family, and to a lesser extent, school, peers, and S.E.S. Partic- ipation in athletic competition was usually not a signifi- cant factor when controlling for other factors. Conclusions Televised sports is associated with certain con- servative political values, and may be a socializer of those values. The mass media may be playing a significant role in the political socialization process, especially via content which is not overtly political. Further attention should be focused on this possibility. Also, given the legitimization functions of television demon- strated, where content re-enforces system support and loyalty, discouraging dissent, further attention to access and diversity should be considered as specific policy alternatives. Further research, which is more general- izable, exerts more control, uses more sophisticated measurement, and examines a wider age range, is desirable. THE ROLE OF TELEVISED SPORTS IN THE SOCIALIZATION OF POLITICAL VALUES OF ADOLESCENTS BY Robert H. Prisuta A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Communication Arts and Sciences 1977 G I e "‘7 =1?» f 44* Copyright by ROBERT H. PRISUTA 1977 ACKNOWLEDGMENTS I would like to especially thank my dissertation adviser, Dr. Thomas F. Baldwin, for his help and encour- agement at every stage of this project. I would also like to thank the other members of my committee: Dr. Charles K. Atkin, Dr. Martin P. Block, and Dr. Ada W. Finifter, who gave freely of their time and contributed many positive ideas throughout the development of the study. In addition, I would like to express my appreci- ation to the hundreds of high school students surveyed and the school district personnel who cooperated with the study, thus making it possible. Most of all, however, I would like to thank my wife, Barbara, whose continuing support, even through difficult moments, was absolutely essential and greatly appreciated. ii Chapter I. II. III. IV. TABLE OF CONTENTS INTRODUCTION AND RATIONALE . . Definitions . . . . . . Development of Values . . . Socialization and Values . . Audiences and Programming. . HYPOTHESES AND REVIEW OF RELATED LITERATURE . . . . . . General Approach. . . . . Television and Political Socialization The Role of Televised Sports. Hypotheses . . Controls for Possible Confounding Variables . . . . . . ME TH O DOLOGY C O O O O O 0 Data Collection . . The Sample. . . . Statistical Testing. Variable Construction Coding . . . . . Hypothesis Testing . DATA ANALYSIS . . . . . . Descriptive Information . . Amount and Preference for Sports Viewing . . . . . . Types of Sports Preferred . Functions of Sports Telecasts. The Sports Viewer. . . . Value Measures. . . . . Hypothesis Testing . . . . iii |-' “>th 13 13 13 29 44 54 54 58 64 71 75 82 82 82 84 85 88 91 93 Chapter Relationships Among the Variables . . . Confounding Variables Further Analysis . Variations in Locality Cross Tabulation . V. CONCLUSIONS, RECOMMENDATIONS, AND FOR FURTHER RESEARCH. Descriptive Information Hypothesis Testing . Suggestions . . . Policy Implications . Suggestions for Further Research. APPENDICES APPENDIX A. SURVEY TESTING INSTRUMENT Dependent SUGGESTIONS B. DEMOGRAPHIC INFORMATION FOR SAMPLE DISTRICTS SELECTED BIBLIOGRAPHY . . . iv 99 99 131 140 141 151 151 153 162 167 169 172 181 182 10. 11. 12. 13. 14. LIST OF TABLES CORRELATION MATRIX OF THE INDEPENDENT/ PREDICTOR VARIABLES . . . . . . . . FACTOR LOADINGS AND FACTORS . . . . . . FACTORS AND VARIANCE ACCOUNTED FOR. . . . INTER-ITEM CORRELATION COEFFICIENTS . . . FREQUENCY DISTRIBUTIONS OF DEPENDENT VARIELES O O O I C O O O O O 0 SIMPLE REGRESSION EQUATION PREDICTING VALUE LEVELS O O O C O O O O O O O O BIVARIATE PEARSON CORRELATION MATRIX OF DEPENDENT/CRITERION VARIABLES. . . . . REGRESSION EQUATION: SPORTS TELEVISION AND PARTICIPATORY SPORTS. . . . . . . . REGRESSION EQUATION: SPORTS TELEVISION AND SPECTATOR SPORTS C O O O O O I O O REGRESSION EQUATION: SPORTS TELEVISION AND GENERAL TELEVISION VIEWING. . . . . . REGRESSION EQUATION: SPORTS TELEVISION AND TYPE OF SPORT PREFERRED . . . . . . REGRESSION EQUATION: SPORTS TELEVISION AND INSTITUTIONAL FACTORS . . . . . . . PARTIAL CORRELATIONS BETWEEN SPORTS VIEWING MD VALUES 0 O O O C C O O O O O REGRESSION EQUATION: SPORTS TELEVISION AND OTHER SOCIALIZATION AGENTS. . . . . . 65 67 68 76 92 98 100 103 106 109 112 114 116 117 Table 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. CORRELATIONS BETWEEN DEPENDENT VARIABLES AND DEMOGMPHIC FACTORS C C O C O O C . REGRESSION EQUATION: SPORTS TELEVISION AND DEMOGRAPHIC FACTORS . . . . . . . . REGRESSION EQUATION: SPORTS TELEVISION AND AGE 0 O C C C I O O C O O C . REGRESSION EQUATION: SPORTS TELEVISION AND LOCATION O O O O O O O O O O C O PARTIAL CORRELATION SUMMARY TABLE . . . . COMPLETE MULTIPLE REGRESSION ANALYSIS. . . VARIANCE ACCOUNTED FOR BY SPORTS TELEVISION. VARIANCE ACCOUNTED FOR BY ALL TELEVISION VIEWING O O O O O O O C O O O O MULTIPLE REGRESSION ANALYSIS FOR PERSON AND SYSTEM ORIENTATIONS . . . . . . . . MULTIPLE REGRESSION ANALYSIS FOR SOCIAL AND INSTITUTIONAL ORIENTATION . . . . . . VARIANCE ACCOUNTED FOR BY TELEVISION VIEWING CORRELATION MATRIX OF FINAL DEPENDENT/ CRITERION VARIABLES . . . . . . . . ZERO CORRELATIONS BY AREA. . . . . . . PARTIAL CORRELATIONS BY AREA. . . . . . CROSS TABULATION: SPORTS TELEVISION AND AUTHORITARIANISM . . . . . . . . . CROSS TABULATION: SPORTS TELEVISION AND NATIONALISM . O C O C O O C C O CROSS TABULATION: SPORTS TELEVISION AND NEED DETERMINED EXPRESSION/VALUE RESTRAINT . . CROSS TABULATION: SPORTS TELEVISION AND EQUALITARIANISM . . . . . . . . . vi 120 121 124 126 128 129 130 132 135 138 138 139 140 142 146 146 147 148 Table 33. CROSS TABULATION: SPORTS TELEVISION AND INDIVIDUALISM . . . . . . . . . . . 149 34. CROSS TABULATION: SPORTS TELEVISION AND SPORTSMANSHIP . . . . . . . . . . . 149 35. VARIANCE AND STANDARD DEVIATION OF DEPENDENT VARIABLES . . . 155 vii CHAPTER I INTRODUCTION AND RATIONALE This study dealt with the relationship between televised sports events and political values. It con- sidered the possibility that such events may be part of a process in which television acts as a socialization agent of those values. The study is, therefore, an attempt to deal with the overall problem of ascertaining how it is that a young person acquires the values, attitudes, beliefs, behavior patterns, and skills necessary to both function effec- tively within society and allow society to function in a stable manner from generation to generation. It is felt that much of this process comes about via some form of social learning in which the individual reacts to cues supplied from individuals and institutions in his environ- ment, as opposed to being purely genetic or instinctual. As Siegel (1969) has stated, Every civilization is only twenty years away from barbarism, for twenty years is all we have to accomplish the task of civilizing the infants who are born into our midst each year. These savages know nothing of our language, culture, religion, values (my emphasis), or customs. The infant knows notfiing about communism, fascism, democracy, civil liberties, the rights of the minority as contrasted with the perogatives of the majority, respect, decency, ethics, morality, conventions, and customs. The barbarian must be tamed if the civilization is to survive. Over the centuries, man has evolved methods of accomplishing this. Our methods of “socializing” the barbarian hordes who invade our community every year rely on their remarkable learn- ing abilities, as the young child learns through observation and imitation. Therefore, this study chose to examine one part of this vast process, examining one object of sociali- zation; values and value systems (the latter being organized general combinations of specific values), and one process by which these values may be transmitted and displayed for observation and subsequent imitation, the telecasting of organized competitive sports events. Definitions A value, as defined by Rokeach, is a ”type of belief, within a larger system, about how one ought or ought not to behave," and has been considered an abstract evaluation of a specific behavior (Frost & Sims, 1974). The importance of values to the human condition has been stressed by Maslow who says, "a state of being without a system of values is psychopathogenic" possibly creating more harm than classic neuroses or psychoses. Skinner (1971) states that values deal ”not with what man can do, but with what he ought to do," classify- ing behaviors as "good" or “bad.“ This labeling of ”good" or ”bad" classifies something in terms of its re-enforcing effects. Things which are "good” are posi- tively re-enforcing, while that which is "bad" is nega- tively re-enforcing. This re—enforcement function, Skinner maintains, is the most important aspect of a value, as behavior considered beneficial is encouraged. Re-enforce- ment is a major component of television content's effects on human attitudes and behaviors (Klapper, 1960). Kluckhohn (1952) describes values as "culturally weighted preferences for things, ideas, people, insti- tutions, and behaviors,“ expressing “preferences for modes of conduct and end-states of existence.” Kerlinger (1973) defines a value as a perception and evaluation of what is important, expressing the "good, the bad, the shoulds, the oughts, etc. of human behavior." Laswell (1948) stresses the more functional role of values, describing them as "categories of relationships that are recognized objects of gratification.” That is, complex, ordered, patterns of behavior which lead to generally desired goals, such as "power“ and ”respect.” Rokeach also refers to a value as a known and favored behavior based on the goals of the individual. Another dimension of values is stressed by Katz who states that a value is an ”organized and hierarchical system of specific attitudes.“ Dennis (1973) states that values are complex, having affective, cognitive, and evaluative components. While the concept of 'value' is complicated, this study focused on the evaluative components of values in which certain states, relationships, and behaviors are evaluated positively and then preferred and sought by the individual. Development of Values Laswell states that such values are "shaped and distributed by institutions, including communications.” While Atkin (1975) states that "unlike parental and school agencies, the mass media do not produce messages with the intent of politically socializing new members of society,” other more radical spokesmen feel that commercial media in the United States actively promote those values from which they will derive the most benefit (Skornia, 1965; Johnson, 1970). Laswell also states that values are transmitted through "ideology . . . the myths of any given society.” This mythic role in contemporary society has been ascribed to sports and sports broadcasting by media and social critics in recent years (Real, 1975; Novak, 1976). Dahl (1973) states that the transmdssion of and incentive for values and value development is weakened in larger societies. Himmelweit (1955), however, finds that values can be influenced by television if consis- tently presented and if those receiving the messages are at a level conducive to value development. Easton and Dennis (1965) find that value processes related to government do not become fully developed until adolescence, a finding similar to that of Davies (1968) regarding the development of nationalism and related values and attitudes. It is at this age level that early family- based patterns of authoritarianism are eroded, and the individual begins to deal with the social system on a realistic scope and basis. More generally, several child development studies (Taba, 1949; Havighurst, Robinson, & Dorr, 1946; Siegel, Conn, Pepinsky, & Rubin, 1956; Durkin, 1959) indicate that the development of an organized, con- sistent, and sophisticated value system does not begin until adolescence. Ryan (1977) applies Kohlberg's (1968) theory of Cognitive-Developmental Value Acquisition to suggest the significant role television may play in this regard. Socialization and Values An individual acquires both values and value sys- tems through a process of socialization, in which appro- priate beliefs, values, attitudes, and behaviors are learned (Comstock & Lindsey, 1975). Each generation must be taught such basic social patterns and values, if the system is to survive. The concept of socialization views this learning process as primarily institutional, wherein the individual is considered to be a product of society's institutions. An institution, as defined by English and English (1958) is: . . . an enduring organization of some aspect of collective life (social, political, economic, or religious) controlled by rules, customs, rituals, or laws. While the organization consists of persons, the pattern of their relationship is such a way as to be relatively independent of the individual. Defleur (1970) also states that institutionali- zation is the stabilizing of widespread patterns of actions related to some cultural trait or combination of traits, usually felt to be functional to the social system. English and English further state that ”abstract concepts, such as values, are usually not the result of the individual's own valuing; they are social products that have been imposed upon him and slowly internalized (i.e., accepted and used as his own criteria of worth).' Rokeach further states that it is just as meaningful to speak of institutional values as individual values, as social organizations specialize in the transmission of selected values from generation to generation. "In its broadest conception, socialization refers to the sum total of past experiences an individual has, that in turn, may be expected to play some role in shaping his future behavior“ (Inkeles, 1969). Effective sociali- zation is a precondition to organized social life. Every social organization, therefore, is to some degree an agent or producer of socialization. Political socialization involves the acquisition of information, and attitude and value formation, which helps the individual make sense out of the political objects and institutions in the surrounding environment, and ultimately influences behavior (Rubin, 1976). It helps orient the individual to the political environment (Fowlkes, 1974). Easton (1957) defines politics as the making and executing of authoritative decisions for society. It is an organized process of utilizing power to distribute the scarcity of demand satisfaction. In contemporary society, such responsibility is placed upon the institution of government and only government can act authoritatively on the whole society (Dye, 1972). Thus, that which relates to or is a part of government and its decision- making processes may be considered part of the political system and the subject of the political socialization of its members. The magnitude and pervasiveness of the political system, and its influence on all members of society, makes it the most important of the social systems (Finifter, 1972). In a restatement of socialization concepts applied to information systems and political concerns, Ellul states that message systems are basically propo- gandization networks in which "contemporary media are not information channels, but purveyors of sociological propaganda," or "mass propaganda." He defines this propa- ganda as "those influences which are in basic accord with each other and lead spontaneously in the same direction . . . are organized along spontaneous patterns and rhythms . . . and produce a certain general conception of society, a particular way of life." Thus, certain cultural ele- ments inherently stress certain values related to the overall sociopolitical system (Christians, 1976). The complexity and importance of the socialization process, and the role of television in it, is now receiv- ing the highest priority among communication scholars (Comstock & Lindsey, 1975). More specifically, political socialization study has been termed one of the newest and most active areas of social science research (Atkin, 1975). Kraus and Davis (1976), in their analysis of political socialization research to date, cite several weaknesses of previous studies, which this project attempts to address. One of these is a tendency to underestimate the role of mass media in the political socialization process, concentrating on more traditional socialization agents, probably due to the greater attention the subject has received in political science, as Opposed to communi- cation, research. Another is the tendency to concentrate on content more overtly political, as opposed to that which may contain inherent political messages not salient to the individual, and thus processed via incidental learning. A third is the tendency to narrow the focus of the oper- ational definition of "political socialization" to more direct aspects of this process, such as attention to political information, voting behavior, party identifi- cation, etc. In addition, the authors call for future research which attempts to deal with the complex interactions between the media and other established socialization agents. Therefore, by dealing with a form of programming which is relatively nonsalient politically, and consider- ing abstract political values and value systems, and conducting analysis which lends itself to discerning more complex relationships between numerous variables, this study hopes to move in the direction suggested by the authors. Audiences and Programming The study will focus on the telecasts of competi- tive sports events, the popularity and economic value of which is virtually common knowledge. Audiences for even relatively unpopular Sports events number in the millions of households, with audience estimates for popular telecasts like the annual "Super Bowl" ranging between seventy and ninety-five million. The economic value of the programming has led to a gradual increase in the number of events telecast. 10 Children and adolescents make up a significant portion of the audiences for such events, usually com- prising from 20 to 30 percent of the total (Ogilvy & Mather, 1975). Other audience data (Nielsen, 1976) reveals further the extent of the "under 18" audience. The average audience for this group, over all network- telecast sports events, was 2,610,000. Also, the younger audience is more consistent in its viewing patterns, with less variance than adult audiences between different sports events. Average audience levels for sports events telecast by the networks in 1975, among children and adolescents, was as follows: Boxing (Ali-Norton) 7,030,000 Superstars 4,740,000 Football 3,510,000 Skiing 3,110,000 Multi-Sports (i.e., ABC's Wide World of Sports and CBS Sports Spectacular) 3,190,000 Horse Racing 3,010,000 Auto Racing 2,770,000 Baseball 2,440,000 Basketball 2,060,000 Bowling 2,050,000 Track 1,730,000 Golf 1,520,000 Hockey 950,000 Tennis 850,000 The single program attracting the largest audience was the Super Bowl, which attracted 12,260,000 viewers of this age level. During 1975, television networks presented 336 separate sports event telecasts for a total of 1,075.7 hours of such programming. These figures do not include 11 local and regional coverage, indicating that actual exposure and audience levels are much higher than these figures show. Given the large audiences for sports telecasts, and the frequency with which they are presented, it is quite possible that this type of programming is a sig- nificant part of the socialization process. Therefore, it is felt that this study can contribute to the body of knowledge and theoretical development related to the socialization process and the role of mass communication in that process. The study will be organized in the following manner. It will consist of five chapters, a bibliography, and appendices. The first chapter, or introductory chapter, will present the basic problem for consideration and rationale for the study. It will also define terms and concepts to be used in the analysis, as well as briefly discussing the concepts to be considered in later analysis. The second chapter will discuss general theory and specific hypotheses to be tested in the study, state those hypotheses, and review the related literature and previous research upon which those hypotheses are based. The third chapter will discuss the methodological procedure by which concepts will be operationalized, 12 data collected, data treated, hypotheses tested, and descriptive information derived. The fourth chapter will present the results of the processes described in Chapter III. It will consist of descriptive information, and the testing of hypotheses, as described in Chapter II. These data will be analyzed and discussed in Chapter V, which will also draw conclusions based on the data. The conclusions will deal with generalizations of consistent data findings, upon which recommendations will be made. These recommendations will deal with theoretical concerns, policy questions, and suggestions for further research. The appendices will contain information of study subjects, as well as the testing instrument used in the study. CHAPTER II HYPOTHESES AND REVIEW OF RELATED LITERATURE General Approach It is the general theoretical assumption of this study that television sports programming acts as a socializer of political values. Television and Political Socialization Several studies have established the function of television content as a socialization agent. Many of these concentrated on news and public affairs programming, and the cognitive role such content plays in increasing information levels and interest among children and adolescents (Chaffe, Ward, & Tipton, 1970; Dominick, 1972; Hollander, 1971; Johnson, 1973; Tolley, 1973). Rubin (1976) also found such content to have affective components, finding distrust of government to be a function of exposure to television news and public affairs among that age group. Two major methods by which young people learn from television have been discerned and discussed in 13 14 both theoretical and research literature. The first of these is cognitive theory, in which specific information is represented in television programming and transmitted to the child, who then processes and retains it. It has been this specific process of information transmission that has been most widely researched in the past, espe- cially with regard to the impact of television with regard to political socialization. Often, such research has centered on the role of television as a source of political information. More important in the process of value development, however, is the role of television in social learning. Here, television content is thought to provide a "role model” which the child imitates and identifies with. This integration and identification causes the behavioral role, and its associated attitudinal, value, and belief systems, to be integrated into the child's personality. This learning can be either formal or incidental (struc- tured and intended or, on the other hand, unstructured and accidental). This learning process also has affec- tive components in which the behaviors, attitudes, etc. presented are viewed within the program context as either good or bad. The process is also a generalized one, given the fact that ”status conferral' and "legitimi- zation“ functions have been attributed to television. In this process, it is felt that there is a tacit 15 approval of the content expressed in the fact that it has been presented, over and above its treatment within the program. The audience, it is reasoned, therefore, "knows” the content is important, and at least tacitly approved, given the fact that those with programming power have chosen to present it. Emotional components are also important with regard to social learning, with such emotional ”arousal" felt to be a stimulus to increased acceptance of the message. Also considered are the degree to which the viewer personally empathizes with and feels ego-involved with the outcome. Horton and Wohl (1956) have found television to be effective in stimulating “para-social interaction" between the viewer and televised characters. That is, the viewer, while consciously realizing the televised character is either fiction, or too remote from the viewer for a genuine personal interaction, still becomes involved on a personal basis with it, sharing emotions, responding with conversation, etc. This was further noted by Lang and Lang (1953) who found viewers became emotionally involved in televised events, to a greater extent than those experiencing the event personally, due to the power of television to involve its viewers in the content. Therefore, social learning theory seems most relevant for this study, stating as it does that a viewer, 16 when presented with a role model via television content, will identify with it, become involved in it, and over time adopt and/or view favorably behaviors and/or atti- tudes associated with that role model. This will espe- cially happen when the viewer is excited, ego-involved, and the message presented is incidental to the overt function of the programming. It thus seems that tele- vision sports events and associated persons are serving as role models for those who view, making such programming a vehicle for social learning in areas removed from the basic subject matter. Several authors (Siegel, 1969; Atkin, 1975) have found social learning to be a factor with regard to general television programming. However, in addition, several authors have noted the same function for sports, espe- cially spectator sports. Denney (1969) has discussed the role modeling function of spectator sports presented via television, with Miller and Russell (1971) commenting upon its ego—involving function. Beisser (1967) discusses the rationale for viewing sports as a traditional influence, maintaining continuity in a rapidly changing world. With clear lines of authority and clear goals, allies, enemies, etc., defined, the field lacks "cultural ambiguity.” The position of athletes as role models was ascertained via sociometric analysis by Dickey (1974) as being relatively high. 17 Also, Kohlberg's (1968) theory of moral develop- ment may apply. Kohlberg pictures moral values as having six stages, progressing from a direct, personally oriented "obedience-punishment” orientation, to a more general, abstract all-encompassing "conscience-principal” orien- tation. As people grow and interact, they go through these stages of moral thinking, although the level among individuals of any particular chronological age may vary. The constant unfolding of moral dramas, in the case of sports events, teaching such moral principles as sports- manship, teamwork, obedience to authority, patriotism, competition, and other aspects of the “American way,“ may thus be causing these values to develop within the individual viewer. Affective functions were also noted with tele- vision viewing in general. Jackson-Beek (1976), Thornton (1976), and Tan (1976) all found high levels of viewing to be associated with certain political values, which could be considered conservative, authoritarian, and traditional, such as government support, the desire for ”security,” passive and conformist attitudes, and will- ingness to see government action taken to solve perceived personal and social problems. Tan found such values to be a function of exposure to television. Jackson-Beek's study involved secondary analysis of National Opinion Research Center national survey data. 18 With controls for age and sex, a relationship between television viewing and political conformism and govern- mental support was found. Tan, while concentrating on nonwhite minority groups in his survey, found, via regression analysis, that television use for both enter- tainment and information was a strong predictor of per- ceived efficacy of government and government support. Thornton's survey found high television viewing associated with such values as the need for “security” and ”stability" with viewers tending to prefer more authoritarian values. Concentrating on violent television entertainment programs, Gerbner and Gross (1976) and Somers (1976) found acceptance of an authoritarian environment to be related to exposure to such programming. In studies of specific programs, Surlin (1976), Wilhoit and de Bock (1976), Vidmar and Rokeach (1974), and Pierce (1977) found All in the Family to act as a re-enforcer of authoritarian, anti-equalitarian, and conservative values to viewers so predisposed. Vidmar and Rokeach's survey of both Canadian and United States audiences found that viewers who were highly ethnocentric, less equalitarian, and more racially biased viewed like qualities displayed by the show's main character, Archie Bunker, as less negative, than viewers without these attitude traits. In fact, many of the viewers felt there was nothing wrong with 19 such manifestations of these attitudes as ethnic slurs and racial stereotypes and approved of their use in the program, which they felt added to its enjoyment. Also, such viewers tended to view the Bunker character more positively, considering Bunker to get the better of the situation in many instances and viewing others, not Bunker, as the butt of humor. Thus, much of the more subtle satire designed to ridicule such attitudes was missed by those sharing them. The authors feel that "selective perception" was at work, a process in which ambiguous messages are interpreted in such a way as to re-enforce the predispositions of those viewing. Wilhoit and de Bock found similar, although not as pronounced, tendencies with a Dutch audience for the program. In this case, more selective exposure was prac- ticed, with those of higher intolerance, authoritarian, and ethnocentric attitudes tending to watch the program less. However, those that did view also practiced selec- tive perception to some extent, tending to view Bunker as a more "reasonable" character. Surlin's study of Georgia audiences for the program also found a re-enforcing effect generated by the character. He finds that the program, by making the attitudes discussed more salient, tends to raise these beliefs to a level of consciousness and, by making the viewer thus more aware of them, re-enforcing them. 20 Pierce finds, in his content analysis and theoreti- cal overview, that the character draws support for other reasons. The character is conservative, traditional, and without ambiguity. He is predictable and does not change, despite changes in society and pressure from others for him to alter his cultural background. He does not reject the background, however, but maintains it and thus draws support from those with similar belief systems. Such studies demonstrate the capacity for tele- vision programming to generate, transmit, and re-enforce traditional, ethnocentric, and authoritarian values, even when its goal is to satirize those values and modify them. The Role of Televised Sports Similar attention is now being focused on the role of sports telecasts in the socialization process. In a theoretical critique and content analysis by Real (1975), televised sports were found to contain ”expressions and values and functions of the larger social structure." Sports telecasts were found to express several cultural values proper to American institutions and ideology. Via content analysis, he demonstrated that coverage of actual action made up a small percentage of the total program time, with subjective analysis and 21 commentary making up a much larger segment, and thus providing ample opportunity for television coverage to inject its own impact, over and above that of the sport itself. Drawing upon Ellul's theories of “mass propaganda," he states that the content may be develoPing similar attitudes and values in the population. He feels such telecasts are reflective of the basic conservative values present in our society. Dis- cussing such aspects as the strict regulation of the activity, the emphasis on property and competition, the male domination, and the authoritarian nature of most sports organizations, he states that sports is ”almost a caricature of the discipline of modern corporate-military society," with ”organizational discipline both on and off the field." He feels that televised sports reflect the ”sexual, racial, and organizational priorities of American social structure.“ Nationalism is another value promoted by televised sports, according to Real, with patriotic/military and ethnocentric themes predominating both the sport itself and surrounding activities. These events are considered to be modern "myths'l which "reflect and sacralize the dominant tendencies of our culture, thereby sustaining social institutions and lifestyles." 22 Real states that these values are transmitted in two ways: through reproduction of the basic sports events themselves, which contain certain inherent values, and through the interpretation, symbolism, and color added to the event by the coverage. The primary function of such events, he feels, is their similarity to traditional mythic activities. ”Larger than life super heroes“ are represented in the contest, with the skill and physical power of the athlete stressed over his commonality with other individuals. In this way, the viewer is drawn into the role of vicarious participant, thus becoming partisan, and investing ego, money, etc., on one side or the other. In this manner, the outcome becomes important, even to relatively noninvolved individuals, and the competitive nature of the event is stressed. Communal focus is also a factor, as a sense of collective participation develops. Identities develop out of athletic identification. A primary outcome here is the ethnocentrism and extreme symbolic nationalism usually associated with athletic events. ”The fact that all major sports are basically American national sports, leven when playoffs are called the 'WOrld' Series, etc., may account for more than a small part of the national cohesion and identity.” 23 Also, Real points out that most sports center around winning property by competition. Field position in football, the geographical domination of the hockey rink or basketball arena, and the circling of bases in baseball, as well as measuring fielding ability by the amount of territory covered by the fielder, are all factors here, all related to the basic social/legal concept of prOperty rights and measuring success and power by property controlled. Authority is also a strong proponent of such organized sports activity. Most modern sports are strictly regulated by "official” clocks, with very spe- cific time limitations. This regulation is a factor not only for the "periods” of time within which competition may take place but within which certain activities may be carried out (i.e., the ten, three, and twenty-four- second time limitations on various basketball activity and similar requirements in football and baseball). Authority figures are also prominent. Highly symbolic, and separated from the others with specific uniforms and roles and positions, they make decisions not through consensus or arbitration or negotiation, but rather through the direct authority of the individual in charge. This, in contrast to unorganized ”sandlot' sports, in which authority is often unstructured. 24 In addition, authority is stressed within the sports organization. Leadership positions are relatively absolute. Employee (athlete) rights such as collective bargaining and freedom of employment have only recently been applied on a partial level, and only after disruptive strikes and conflicts with owners. The individual's role is subsumed by the organization, as is made obvious by the symbolic use of uniforms, etc., to create a feeling of strict organization. Even such concerns as dress and hairstyle are controlled, and stressed by television coverage, to a great degree. Nationalism is also stimulated by the pageantry and symbolism surrounding the event, including military bands, martial music, the overt display of the flag, the playing of the national anthem, flag salute, use of the flag throughout the game, various military demonstrations, half-time entertainment with patriotic themes, etc. Throughout the event, therefore, one is caught up in the feeling that the event is somehow "naturally American” and thus proper. As a recent commercial jingle states: We go together. In the good old U.S.A. Baseball, hot dogs, apple pie And Chevrolet As Real states, a sports event is "replete with martial music, precision drills, uniforms, and massive formations." 25 Social conservatism predominates as well, given the demographics of the participants and developers of sports programming. Male domination is common knowledge. Not only are male sports covered predominantly, but the owners, managers, and coaches are all male, and announcers and commentators are almost totally male. Moreover, while playing itself has become integrated (at least at certain positions, and to some extent), management and coaching, and commentators and announcers, remain predominantly, if not exclusively, white, thus reflecting a more traditional anti-equalitarianism common to the culture. Thus, televised sports become, according to Real, ”aggressive, strictly regulated games fought between males who use both violence and technology to gain control of pr0perty for the economic gain of individuals within a nationalistic entertainment context.” Therefore, it closely resembles, and in fact re-enforces, the cultural system of which it is a part. Also, again drawing on Ellul, the consistency and saturation of sports programming over time may be a factor. As game follows game, in a constant, year-round parade of various sports, each with a consistent theme, a desensitization to the values presented takes place, blurring the ideas inherent in the individual game, and shifting the focus to ”sports“ and what it represents, rather than focusing on individual games, personalities, etc . 26 Similar conclusions are drawn by Novak (1976). Televised sports are viewed as ”symbolic realities involving elements of business, entertainment, politics and gossip." He feels these telecasts perform a social function. Describing them as narrative forms, he ascribes to them an integrative function, stating that ”without narrative forms, a culture flies apart; sorting out the relevant from the irrelevant becomes impossible; living loses zest.” He echoes Real's concept of sports as "mass prOpoganda.' ”The narrative forms that recount athletic struggles supply millions with a sense of form. These forms express implicitly realities of law, fairness, effort, and spirit,” he adds, stating that sports ”drama- tize the myths of our society." Also included is stress on standards "inherent in the structure of the game,” which border on ritual, and exist separate from participants or viewers. He states that such broadcasts “dramatize our sense of order.” Novak discusses affective power of television, stating that "the power of the drama is-so intense that one is drawn into the game,” creating a sense of partici- pation.” He likens sports events to drama, with a coherent narrative, a beginning, middle and end in which emotions become tangible. One of the most common 27 ways sports develops such influence is through its inherent symbolism and the manner in which it can trans- mit symbolic political information. He feels that sports provides a medium by which basic social characteristics may be carried out in ritual form. Thus, sports, like other cultural forms, are designed to be integrative, drawing people together and transmitting traditional values. Among the values Novak feels are transmitted by such activity are standards of fairyplay and law- abidingness which he feels are dramatized in sports. By being drawn into the contest, participating vicariously, and learning from the events which transpire, he likens the effect of sports events to that attributed to ”morality plays“ or tragedies in other cultures. Reflection on the events which take place is possible, as well as a "catharsis" to some extent. This vicarious role playing allows the individual to develop the values stressed by the role. The possible consequences of the interaction of sports, the media, and political control have become salient to the general public, becoming the subject of best-selling novels and popular films, such as those by Harrison (1974). 3 Sports participation itself has been found to be a conservatizing influence in the few studies done on the subject, supporting the theoretical analysis 28 previously done. Stern (1971) found participation led to greater efficacy, confidence in established order and channels to solve perceived problems, and "a general acceptance of society's authority structure” with an "establishment orientation." He theorizes that such changes come about as a result of individuals being forced into a competitive environment in which they must perform specific tasks in order to achieve success. Thus, they become success and achievement oriented, and more supportive of a "laissez faire" environment, in which individuals must solve their own problems, as opposed to a social welfare state where the individual receives additional support. Williams (1973) also found interest in athletics inversely related to critical thinking, and associated with a desire for social achievement, in his survey of college students. Coutts (1973) also found college students participating in sports in physical education classes became more authoritarian and less equalitarian, essentially for the same hypothesized reasons as those of Stern. Phillips and Schafer (1971) found high school athletes to be more upward mobile and less likely to be in trouble with the law. This research would support both conservative spokesmen such as Rafferty (1971) and radicals such as Scott (1971) who feel sports participation is a con- servatizing influence. Rafferty praises this influence, 29 discussing sports as an environment in which youth can learn ”respect for authority,” how to "follow orders,” I'patriotism," 'manliness,” and other traditional values. Scott differs somewhat, attributing this influence of sports more to the current structure of sports organi- zation in the United States rather than its inherent nature. Traditionally, successful figures in both business and politics have attempted to identify themselves with sports as a way of building popularity and reaffirming its role in traditional American lifestyle and values. Hypotheses While the preceding literature provides legitimacy for the overall theoretical assumptions, several specific hypotheses may be further derived. One of these relates to the specific relationship between sports on television and the associated general authoritarian environment pre- viously discussed. Hypothesis 1: Higher levels of authoritarianism are associated with higher levels of exposure to televised sports. This hypothesis is derived from both the previous material, and research in the recreational field which associates participatory sports with this value as demon- strated by Coutts (1973), Williams (1973), and Stern 30 (1971). This research supported similar discussion by others (Scott, 1971; Meggyesy, 1970; Rafferty, 1971) as well as popular literature (Gent, 1972). Authoritarianism is characterized as a conceptual framework which seeks to reduce ambiguity, and accept the actions and statements of those in position of power, as well as the display of intolerance for dissent and/or opposition to that power. Information to test the hypothesis was collected through survey procedure. The operationalization of two concepts, television exposure and authoritarianism, was necessary. Likert summated rating format questions were utilized for this purpose (see methodology section). Con- cepts and dimensions utilized were inspired by instruments used by Kelley (1970) and Sonandres (1968) in their sports studies, as well as that utilized by Atkin (1976) in his study of television effects. The first series of questions dealt with television sports exposure and its various dimensions. The first question dealt with the frequency of sports viewing. 1. Which of the following statements comes closest to describing the amount of television sports viewing that you do? 1. I watch most sports events telecast. 2. I watch many televised sports events. 3. I watch sports events occasionally. 4. Rarely do I watch a televised sports event. 5. I almost never watch a sports event on television. 31 The next questions dealt with affective dimensions impor- tant due to the finding of Himmelweit (1955) that uncriti- cal attachment to the medium and its content is related to acceptance of the message. 2. Watching sports on television is something I really like to do. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. I enjoy the announcers and commentary on sports tele- casts because they add a great deal to the program. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Another series of questions dealt with the function sports viewing serves for the viewer: 4. I watch sports on television because I learn a lot from it. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 5. I watch sports on television because it relaxes me, and helps me to forget about my everyday problems. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 6. I watch sports on television because it's exciting, and gets me emotionally involved. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 7. I watch television sports only if there is nothing better on. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Finally, a question assessed the type Of sport, whether individual or team, which is preferred: 8. What is the sport you most enjoy watching on television: (1) Baseball (6) Swimming (2) Tennis (7) Hockey (3) Basketball (8) Skiing (4) Golf (9) Track and Field (5) Football (10) Other 32 These questions performed two functions. By examining various dimensions of the activity, these dimensions were combined to determine the degree of involvement in the process. To establish the general independent/predictor variable necessary in the initial hypothesis ("exposure to televised sports”), responses to questions 1, 2, 3, 4, 5, and 6 were added with 7 also included, for a possible score of 7 to 35, the assumption being that someone who views often, enjoys viewing, becomes emotionally involved, etc., is more involved in the process than someone reacting less positively along one or more of these dimensions and whose behavior is less consistent. Question 7 was subtracted, as it is indicative of more passive and less favorable behavior. Question 8 was applied to the ”type of sport preferred“ index which will be discussed later. Questions related to the dependent variable were derived from the study by Coutts, who utilized the Bales and Couch Value Dimension Test to measure authoritarianism and three other political value dimensions. Withey (1965) developed a more concise ”Dimensions of Value" index from this test, which will be replicated here. Via factor analysis, Withey determined which questions from the original index fall into which particular factors, related to the four values under consideration. He then took the three highest loading items on each 33 factor and adapted them into the short form, which is self- administered and takes approximately five minutes to com- plete. Through additive Likert scoring, which is the format of the questions, the scores on each dimension may be added, providing a range of score on each of the dependent/criterion variables of 3 to 15. Such repli- cation was utilized where possible, given the advantages in time, cost, cross validation, interpretation of find- ing, etc., inherent in the process (Riecken & Boruch, 1974). The questions utilized to operationalize authori- tarianism are as follows: 1. Young people sometimes get rebellious ideas, but as they grow up they ought to grow out of them. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 2. You have to respect authority, and when you stop respecting authority your situation isn't worth very much. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. Obedience and respect for authority are the most important things in character that children should learn. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Hypothesis 2: Higher levels of nationalism are associated with higher levels of exposure to televised sports. In addition to previous authors, Clark (1976) and Anderson (1976) have discussed the relationship between participatory athletics and nationalistic 34 attitudes, which can be characterized by loyalty to and identification with the nation-state of which the indi- vidual is a member. Stern (1971) found similar results in his study of Chilean youth, with Lypsyte (1976) find- ing this to be a multi-national aspect of sports develop- ment in his historical analysis. Frommer (1974) has dis- cussed the possible impact of television in this area. Measures of the independent/predictor variable were the same as those utilized in Hypothesis 1. Measures of the dependent variable were replicated from the National Involvement Scale of DeLamater, Katz, and Kelman (1969). This scale attempts to isolate the three components of nationalism: symbolic, normative, and functional. The first involves national symbols, the second rewards and sanctions, and the third, material benefits. These three components were derived from the work of Katz and Kahn (1966), who viewed the above as the three dimensions by which any role system is integrated. These dimensions are values, norms, and roles, respec- tively. As a result, the portion of the instrument designed to measure the value of nationalism deals with national symbols. It is the most abstract and affective, as opposed to the more cognitive and behavioral charac- teristics of the other dimensions. The questions were modified slightly to fit the Likert format. The initial eight items were reduced to 35 three for the purposes of questionnaire brevity and efficiency. Alternatives selected were those which displayed the widest distribution across each set of response alternatives, maximizing the potential variance in the measurement of the variable. The questions are as follows: 1. A person who says that he takes no particular pride in our armed forces is not being a good American. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 2. A person who does not stand when the Star Spangled Banner is played is not a good American. TIT STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. Foreigners have no business criticizing American's internal problems, such as those concerning the economy or racial issues. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Hypothesis 3: Higher levels of value restraint, and lower levels of need-determined-expression, are associated with higher levels of exposure to televised sports. “Value restraint" and 'need-determined-expression' are opposite concepts. The former is the preference for an ethical framework in which long-term, traditional ethics and values are maintained over more expedient alternatives, and loyalty to institutions and organi- zations, and their well-being, takes precedence over the needs, wants, and well-being of the individual. The latter is a valuing of short-term, self-actualizing 36 approaches which place individual development ahead of more general social guidelines and restraints. ”Need Determined Expression“ is, therefore, simi- lar to the popular concept of "situational ethics,” or the concept of a "value free” environment. Short-term factors are considered in any decision rather than long- term consequences. A value system favoring “need deter- mined expression' would view expediency and an “ends justify the means" approach to problem solving over a more deliberate, considered, and traditional approach which considered ramifications of short-term actions, even if those short-term approaches were beneficial at the outset. "Value Restraint” is a framework which attempts to take into account the long-run and the long-term ramifications of activities, attitudes, etc. It is more general in its assumption that there are enduring, more or less absolute values of what is good and bad, proper and improper, etc. Therefore, it views the individual's role as being one of propagating and carrying on those values, even if that propagation results in short-term inconvenience for the individual. Value Restraint assumes social stability as preferred, whereas need determined expression assumes a flexible, self- actualized society is more functional. 37 The relationship between these value systems and sports has been tested by Coutts and Stern as part of the previously referred to study in which the Withey Dimensions of Value Test was used to measure this particular value. The questions are in the same format as other value dimension questions: 1. The solution to almost any human problem should be based on the situation at the time, not on some general idea of right or wrong. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 2. Do what you want to do that's fun and worry about the future later. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. Since no values last forever, the only real values are those that fit the needs of society right now. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE High levels of agreement would indicate high levels of need-determined-expression with low levels indicating strong value restraint. Hypothesis 4: Lower levels of equalitarianism are associated with higher levels of exposure to televised sports. As was the case with the previous values, this hypothesis is also derived from the general theory of Real and Novak, and empirical studies of participating sports by Stern and by Coutts, using the Withey Dimensions of Value Test. Equalitarianism is a value which embodies tolerance and acceptance of individuals, and values equal 38 treatment, as opposed to a more rigid and stratified process of dealing with individuals, based on social and/or psychological characteristics or attributes. This concept was thus operationalized by Withey, Bales and Couch, and Coutts as follows: 1. A group of people that are nearly equal will work a lot better than one where people have bosses and ranks over one another. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 2. Everyone should have an equal chance and an equal say in most things. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. Everyone should have what he needs. The important things we have belong to all of us. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Hypothesis 5: Higher levels of individualism are associated with higher levels of exposure to televised sports. Individualism values accomplishments and activi— ties of the single person, acting alone and on his own, without support from government and other institutions. Stern noted its association with participating sports in his study. In it, he found athletes to assume a "survival of the fittest" approach to social issues, which tended to support a competitive system, as opposed to that utilizing governmental welfare action to aid the dis- advantaged. This value was the other value dimension 39 tested by Coutts in his examination of the relationship between sports and values and completes the Withey Dimensions of Value Test. 1. We should all admire a man who starts out bravely on his own. ( 1) STRONGLY AGREE ( 2) AGREE ( 3) NEUTRAL ( 4) DISAGREE (5) STRONGLY DISAGREE 2. In life a person should for the most part '90 it alone,” working on his own and trying to make his own life. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL ( 4) DISAGREE (5) STRONGLY DISAGREE ' 3. One should not depend on other persons or things, the center of life should be found inside oneself. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Hypothesis 6: Higher levels of sportsmanship are associated with higher levels of exposure to televised sports. The American Society of Health, Physical Edu- cation, and Recreation (1974) defines the concept of sportsmanship as that which values excellence in per- formance within a context of respect for competitors and the rules of the game. They ascribe a great deal of influence to television in the acceptance or rejection of this value. Shusedorf (1969) has also commented on the role such coverage may play, without drawing specific conclusions on the direction of the effect. . Several authors have paralleled the value of sportsmanship with more direct political values, and 40 have commented on its effect on the political system (Margolis, 1976; Michener, 1976; Tutko & Burns, 1976). Laboratory experiments have established a causal link between television sports content and behavioral and cognitive dimensions of the concept of sportsmanship. Poulos (1976) found that children viewing programs in which sports advocating positive play behavior (such as sharing) were aired engaged in games much more coopera- tively afterwards and were less competitive than subjects viewing regular programming. Stein and Bryan (1972) found that those who saw a model follow rules, not cheat, and succeed in a game were much more likely to both remember and to practice such behavior than those who saw more negative behaviors either discussed or acted out. The relationship between this relatively indirect political value and the political value system of the individual becomes apparent at the general level. Sports- manship, like such dimensions as value restraint, calls upon the individual to support institutional procedures even at the expense of short-term personal success which may result from the violation of rules and regulations and departure from traditional standards of conduct. Competition, while valued, must take place only within a specific institutional framework. Such a frame- work is designed to perpetuate the format of competition, 41 with the success or failure of any particular participant in the system a secondary issue. Both our political and economic systems have parallel value structures. A representative democratic political system features competition for power via election to office, with candidates for the office com— peting with each other for a scarce resource. For each office, many may compete, but only one may win. Simi- larly, in the economic sector, corporations and individual entreprenuers compete for the scarce reward of consumer dollars. In any and all of these institutionalized compe- titions, the actors are faced with a choice. They may obey the "rules of the game," following traditional patterns of behavior, and/or formalized codes of conduct, thus remaining loyal to the institution at the risk of losing the contest; or they may violate such codes or traditions for short-term expediency, increasing their chances of victory while also increasing the chance of sanctions by the institution should they be caught vio- lating the rules. The sense of values and priorities the actor brings to his participation will influence the degree of loyalty to either self-interest or the interests of the institution. A great deal of discussion in recent years has centered on this phenomenon in light of the "Watergate" 42 revelations in which short-term outcomes favorable to those in political power were given higher priority than the upholding of both legal and traditional norms of the electoral process. The "win at any cost” approach applied to the political system, as adopted from the sports world, has been similarly discussed. Thus, this value's ties to sports competition and its role as a subset of the overall value of institutional loyalty and tradition makes it an appropriate value for examination in this context. This interaction between business competition, ethical values, and sports has been studied by Maccoby (1976) who describes an individual who combines such qualities as "the games- man," the most common personality type now in leadership positions in American business and government. In this portion of the instrument, part of a sportsmanship index developed by Johnson (1966) was replicated. It is also Likert scaled. In selecting questions from the overall pool, several criteria were applied. The primary one was variance of the response distribution. Those questions with high phi coefficients, thus showing maximum discrimination, were first selected. Next, questions which dealt with a variety of participants in the sports process over a variety of sports activities were selected. 1. A football coach left the bench to change the position of the marker dropped by an official to indicate where the ball went out of bounds. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE 43 A baseball player was called out as he slid into home plate. He jumped up and down on the plate and screamed at the umpire. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE At a basketball game, fans tried to bother an opposing player trying to shoot a foul Shot. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE During a football game, a player tried to get an opponent into a fight, so that the opponent would be thrown out of the game. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE During a golf match, a fan deliberately made noise while a player was trying to concentrate on an important shot. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE During a tennis match, a player called his opponent's shot out of bounds, although it actually fell within the line. (1) STRONGLY APPROVE (2) APPROVE (3) NEUTRAL (4) DIS- APPROVE (5) STRONGLY DISAPPROVE Finally, a general additive variable based on the conventional concept of ”conservatism? was utilized to test the overall relationship. In constructing the variable, an individual high in authoritarianism, nationalism, value restraint, sportsmanship, and indi- vidualism, and low in equalitarianism, was deemed more "conservative“ than others, given loyalty to tradition, institutions, and the support of rules and constraints, plus a desire to place greater emphasis on individualistic, as Opposed to egalitarian, considerations. 44 Therefore: Hypothesis 7: Higher levels of conservatism are associated with higher levels of exposure to televised sports. Controls for Possible Confounding Variables Exposure to sports television, and the possession of certain political values, does not occur in isolation but is associated with other personal behaviors and attributes. Thus, it was necessary to control for these other factors if the true relationship between the variables under consideration was to be properly assessed. Several areas were taken into consideration including participatory gports. Hypothesis 8: These values will be associated with higher levels of exposure to television sports when controlling for participation in competitive athletic events. Several authors, in addition to those previously cited, have discuSsed the role participatory sports plays in this regard (Milton, 1972; Torkilsden, 1967; Miller, Russel, 1971; Kelley, 1970; Denney, 1969; Saunders, 1972). Thus, involvement in such sports activity was measured and controlled for to separate the effects generated by participation from those generated by viewing. 45 This variable was considered along two dimensions. One of these was an affective dimension: l. I enjoy sports activities, and try to participate in them in my spare time. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Also considered was the frequency of activity: 2. I get a chance to participate in sports (1) Almost every day. (2) Several times a week. (3) About once a week. (4) About once every two weeks. (5) Less than once every two weeks. In addition, the sport itself was considered. 3. What is the sport you most enjoy playing? (1) Baseball (6) Swimming (2) Tennis (7) Hockey (3) Basketball (8) Skiing (4) Golf (9) Track and Field (5) Football (10) Other The first two questions were summed with the third used as part of a later analysis. Again, the instrument of Kelley (1970) was of value in formulating questions. Hypothesis 9: These values will be associated with higher levels of exposure to television sports when controlling for other television viewing. Previously discussed research has identified the role and association of several types of programming with these values. Specifically, violent programming, realis- tic situation comedies like All in the Family with strong 46 authoritarian personalities generating identification, news and public affairs viewing, and television viewing in general have all been identified as socializers. Therefore, it will be necessary to construct a variable which measures the degree of this involvement with tele- vision in contexts other than sports programming. 1. I enjoy watching television. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE During the average day, I watch of tele- vision. (1) Less than 1 hr. (2) about 1 hr. (3) about 2 hrs. (4) about 3 hrs. (5) 4 hrs. or more I especially enjoy programs with strong characters, like Archie Bunker in All in the Famil . (1) STRONGLY AGREE (2) A REE NEUT (4) DISAGREE (5) STRONGLY DISAGREE . I like programs like police shows and westerns, that have a lot of fighting and shooting in them. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE I like to watch news and public affairs programs on television. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE These questions were added to measure overall television viewing identified as socializers. By con- trolling this variable, the unique contribution of tele- vision sports can be more readily assessed. Hypothesis 10: These values will be associated with higher levels of exposure to television sports when controlling for the in-person attendance at sports events. a 47 Although most research done in the area (Jordan, 1950; Horowitz, 1974; Belson, 1959) indicates that atten- dance at sports events is a function of exposure to those events on television, at least in part, some evidence of reverse causation exists (Saunders, 1972; Greendorfer, 1974). Also, a factor is the concept of different values emerging from participatory, as opposed to passive, sports environments (American Society of Health, Recreation, and Physical Education, 1963). The variable of sports attendance was controlled to determine how much of a factor viewing is when holding attendance constant. The following questions, also inspired by the work of Kelley (1970), were used to operationalize the con- cept of in-person sports attendance: 1. I like to attend sports events and watch others play. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 2. I attend spectator sports events. (1) Once a week or more. (2) Several times a month. (3) A couple of times a month. (4) Once a month. (5) Less than once a month. Also utilized will be a question on the sport preferred. 3. What is the sport you most enjoy watching in person? (1) Baseball (6) Swimming (2) Tennis (7) Hockey (3) Basketball (8) Skiing (4) Golf (9) Track and Field (5) Football (10) Other 48 Other social institutions, established by previous research as effective and significant political sociali- zation agents, were examined and their effect controlled for. These other institutions have been identified as school, family, and peer group (Atkin, 1975; Dennis, 1973; Fowlkes, 1974). Both Atkin and Dennis also cite numerous other studies stating that one or more of these insti- tutions are effective agents in the political sociali- zation process as do Kraus and Davis (1976) and, to a lesser extent, Jennings and Niemi (1974). The following hypothesis is derived: Hypothesis ll: These values will be associated with higher levels of exposure to television sports when controlling for the effect of other established socialization agents; namely school, family, and peer group influences. ' Given the dimensions to be examined in the study, those of conservative values, sports, and political socialization, questions concentrating on impact in this area dealt with these three dimensions for each of the socialization agents. To derive a measure of the effect of the socialization agent in question, the scores for those responses were summed by dimension. 1. Sports is something my family is very interested in. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 49 2. My friends are very interested in sports. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 3. Sports is very important at school. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 4. My family is very interested in political affairs. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 5. My friends are very interested in political affairs. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 6. Topics such as political affairs and civics are very important subjects at school. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 7. My family is very conservative politically. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 8. My friends are very liberal when it comes to politics. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE 9. My school stresses conservative political ideas. (1) STRONGLY AGREE (2) AGREE (3) NEUTRAL (4) DISAGREE (5) STRONGLY DISAGREE Hypothesis 12: These values will be associated with higher levels of exposure to televised sports when controlling for the attribute variables of sex, socioeconomic status, and race. Differences in political values and attitudes associated with sex differences have been noted by such authors as Easton and Dennis (1967), Lefkowitz, et a1. (1973), and Greenstein (1961). Similar differences with respect to race have been noted by Greenberg (1970) and 50 Aberbach and Walker (1970), and with respect to S.E.S. (Key, 1964; Wallace, 1964; Alford, 1963; WOfinger, et a1., 1964; McClosky, 1962; Jaros, Hirsh, & Felton, 1968). Differences along the same lines have been noted with regard to television viewing (Greenberg & Dominick, 1969; Katzman & Wachman, 1970), and sports interest (Sonandres, 1968; Greendorfer, 1974; Kelley, 1970; Smith, 1972). Questions dealing with these factors appeared late in the questionnaire, due to their personal nature, and the risk of some nonresponse and negative reaction. The questions were relatively straightforward as follows: 1. Your ethnic background is: (1) American Indian (2) Black (3) Spanish American (4) White (5) Asian American (6) Other 2. Your family's income is . . . (1) Much above average ($21,000 per year or above) (2) Above Average ($16,000 - $20,000 per year) (3) Average ($11,000 - $15,000 per year) (4) Below Average ($6,000 - $10,000 per year) (5) Much below average ($5,000 per year) 3. What does the main wage earner in your family do for a living? Categories for the question on family income were based on the latest Health, Education, and Welfare statistics (1977) regarding family income standards. The question dealing with occupation was open-ended, given the wide variety of potential occupations, and possible lack of understanding of categories at this age, which preclude precoding of responses. Once open-ended 51 answers were obtained, they were coded with the procedure established in the North and Hatt Occupational Index (1947). These two variables were combined to form a more general socioeconomic status measure. 4. What is your sex? (1) MALE (2) FEMALE Hypothesis 13: These values will be associated with higher levels of exposure to televised sports when controlling for the type of sport (individual v. team) preferred. While Greendorfer (1974) noted that sports preference (individual v. team) was associated with S.E.S., most related literature has either dealt with sports as a whole, or related similar values to various types of sports activity, regardless of type or category of sport involved. Nonetheless, this aspect was treated as a potential confounding variable generated by con- sidering spectator, television, and participatory sport preferred. Due to the nature of the socialization process, and changes in both viewing behavior and political values, according to age, as previously cited, an additional possible confounding variable is age. Therefore, the project tested the hypothesis that: Hypothesis l4: These values will be associated with higher levels of exposure to televised sports when controlling for age. 52 To test for age, students were asked to provide their grade level as follows: 1. What is your present grade? (1) 9th (2) 10th (3) 11th (4) 12th Also, given the fact that the survey was conducted in more than one location, as discussed in the next chapter, an additional hypothesis was necessary if fuller generalizability was to be obtained: Hypothesis 15: These values will be associated with higher levels of exposure to televised sports when controlling for geographical location. And finally: Hypothesis 16: These values will be associated with higher levels of exposure to televised sports when controlling for all of the previously mentioned possible con- founding variables. This hypothesis is most consistent with the realistic setting in which such an interaction would take place. With all of the possible confounding effects interacting simultaneously, only by controlling for each can the effect of television viewing be fully isolated. The preceding are all of the hypotheses to be tested in the study, working up a general socialization model. While numerous more specific sub-analyses are 53 possible, they are not within the scope of this study which seeks to test the general model. The completed questionnaire may be found in the Appendix. CHAPTER III METHODOLOGY Data Collection The nature of the problem, and dearth of non- laboratory research in the area, suggested a field study as most apprOpriate. Survey technique was used to gather data. Such an approach can focus on "real world" relationships and processes (Comstock & Lindsay, 1975) as opposed to the somewhat artificial laboratory setting, which may inflate results due to sensitization of sub- jects, removal of extraneous variables, etc. (Hovland, 1959). The number of cases examined can be increased, and more variables and their complex relationships can be studied. It is also more amenable to replication and is appropriate for exploratory studies (Babbie, 1973). A survey which examines a situation at only one point in time cannot ascertain causal direction or exert the control possible in a laboratory experiment. How- ever, from previous theory and research, causal outcomes can be hypothesized and those hypotheses tested, thus 54 55 permitting indirect, inferential, possible causal attri- bution (Emmert & Brooks, 1970). In addition, the time ordering of the variables involved is a factor (Rosenberg, 1968) in that previous research has indicated that exposure to various television content occurs well before the formulation of specific and organized values. Also, the relationship between the variables of political values and exposure to particular types of television content can be regarded as a stimulus-response relationship, given the distinction between exposure and nonexposure and the relative comparability, otherwise, of the subjects involved. Thus, the general hypothesis that certain political values are functions of exposure to certain types of television content can be tested with some con- fidence. The act of assuming reverse causality does not totally eliminate the possible value of the initial hypothesis. Even if possession of such values leads individuals to seek out certain supportive content, such content could be considered to provide re-enforcement and gratification, further increasing the degree to which the values are held. Therefore, demonstration of a relationship between values and sports television could have four possible explanations: 56 l. Causal direction could be from television to values, with the viewer learning such values from television content. 2. Causal direction from values to television, with the viewer seeking out programming supportive of the values he already holds. 3. A combination of (1) and (2) above. 4. A spurious relationship, coming about only as a result of actual relationships caused by other variables or by chance. Statistical significance testing and control for other related variables will be utilized to ascertain the possibility of the fourth alternative. While previous research and theory already cited would support (1) (tele- vision content causality), other theory and research, particularly "uses and gratifications" research (Blumler & Katz, 1974) and ”re-enforcement” theory and research (Klapper, 1960), would support (2) (value possession causality), stating that viewers with a well-developed sense of values would seek out supportive information from media content. Since the time ordering of the variables in question is inferrential and not fully ascertained, such a scenario is possible. However, causal direction in this manner is unlikely without reciprocal causality in the opposite direction, as described in alternative (3). 57 The rationale for this position is derived from the above authors, plus the work of Skinner (1971) and Festinger (1957). The concepts of re-enforcement, dis- sonance, and gratification are involved. Should an indi— vidual possess a highly developed value and seek supportive content in media, this content will still impact upon that value rather than the value remaining static. Should the content be supportive of the value, it may re-enforce it and provide gratification. In this manner, the value will be more strongly held than it was previous to content exposure, suggesting a two-way flow of influence. Con- versely, if the content was not supportive of the value, it would not be providing re-enforcement and gratification. Therefore, with less incentive, the behavior would serve less of a purpose and may be reduced. Also, the content may actually oppose the value assumptions held by the viewer, which may create dissonance or inconsistency, further encouraging the viewer to reduce this type of exposure and providing less incentive for it. Therefore, even if possibility (2) was the initial action, this causal direction could not be maintained for any length of time without a complementary reverse causal flow, sUpporting the initial causal action. Therefore, the possibility that the causal action of (1) is at least a part of the relationship under study exists. This possi- bility is further enhanced by the fact that the values 58 under discussion are not overt and obvious components of the programming under study. In any case, association between viewing behavior and possession of certain political values can be tested and demonstrated if such an association does exist. The establishment of empirical evidence for such a hypothe- sized relationship in a realistic context or, conversely, the refutation of hypotheses by finding no such associ- ation is considered a major and valid accomplishment for exploratory field research in which a relationship hereto- fore not considered is tested (Rossi & Williams, 1972; Riecken & Boruch, 1972). Also important will be descriptive data regarding the sports viewer at this age level. This group has been ignored previously due to the marketing base of such research. However, these data will allow a profile of the high Sports viewer to be developed. The Sample The survey was a self-administered questionnaire, using high school students (grades 9 through 12) in an in-school situation as subjects, administered in March of 1977. Debriefing sessions followed questionnaire administration. This approach allows a rapid accumulation of a large amount of data, as well as a focus on a sig- nificant age group, given the status of both sports viewing and value formation at this age level. Also, 59 at this age level, the variance between high and low amounts of television viewing would be greatest (Baxter, 1960; Schramm & Roberts, 1971), maximizing the systematic variance under study (Kerlinger, 1973). The complexity of the values dealt with could also be optimally under- stood and empathized with at this age level. Children and adolescents also are traditionally the focus of socialization research. Easton and Dennis (1965) find that value processes related to government do not become fully developed until adolescence, a finding similar to that of Davies (1968) regarding the development of nationalism and related values and attitudes. It is at this age level that early family-based patterns of authoritarianism are eroded, and the individual begins to deal with the social system on a realistic scape and basis. More generally, several child development studies (Taba, 1949; Havighurst, Robinson, & Dorr, 1946; Siegel, Conn, Pepinsky, & Rubin, 1956; Durkin, 1959) indicate that the development of an organized, consistent, and sophisticated value system does not begin until adolescence. Ryan (1977) applies Kohlberg's (1968) theory of Cognitive-Development Value Acquisition to suggest the significant role television may play in this regard. The sampling of larger numbers of respondents in this manner will permit greater precision in data 60 analysis (Glass & Stanley, 1970). The procedure will increase internal validity, given the minimization of testing effect, changes over time, impact of outside events, etc. (Stanley & Campbell, 1963). Access to schools, dependent on administration approval, somewhat limited the study's generalizability. This limitation was modified by secondary inference, as described by Cornfield and Tukey (1956) and Bracht and Glass (1968). In selecting school districts in which to admin- ister the questionnaire, therefore, in addition to cooper- ation, an additional factor considered was the type of district. To maximize generalizability, somewhat dif- ferent communities were sought. Selection of areas was also limited to those within daily driving distance, given the limited resources available. Access to state high schools is often handled through the Office of Professional Development in the College of Education. Therefore, consultation with this office was sought in order that a master list of state schools with a predisposition to cooperate with such surveys could be generated. Also, since many schools use the office to screen prospective visiting researchers, the office, in approving of the project, enhanced its acceptability to the districts. After contacting a large number of school districts, arrangements proved 61 to be somewhat difficult to make. However, three school districts, representing somewhat diverse populations, agreed to cooperate in the survey. As a condition of the cooperation, their identity will remain confidential. Brief descriptions of the areas will serve to indicate the characteristics of the populations surveyed. School District A.--A suburban district close to a medium-sized urban area, relatively high socioeconomic status, with a reputation for being somewhat liberal politically. School District B.-—A large urban area, out of which an inner-city high school with a population rela- tively representative of the entire district, was utilized, the city's population being relatively diverse in terms of socioeconomic status, ethnic background, and political beliefs. School District C.--A transitional district, pri- marily rural but becoming suburbanized, with somewhat lower than average socioeconomic status and with a repu- tation of political conservatism. Due to the large diversity of the total sample, it is felt that the results will be fairly representative and generalizable despite the lack of randomness and the confining of the study to the state of Michigan. In addition, several authors have pointed out that in 62 exploratory research such as this the demonstration of the existence of hypothesized relationships is of greater importance than full generalizability (Rossi & Williams, 1972; Riecken & Boruch, 1972). Sample size was approximately six hundred, increas- ing the power of the statistical analysis (Glass & Stanley, 1970), providing a sufficiently accurate confidence inter- val for an exploratory study (Blalock, 1972) and permitting use of more sophisticated data analysis (Kerlinger & Pedhazur, 1973). The use of Likert scales, requiring the respondent either to "strongly agree" or "strongly disagree" over a range of five alternatives, is considered to be fairly reliable and valid (Emmert, 1970; Blalock, 1972; Babbie, 1973). They are used often, being favorably compared to Thurstone equal appearing interval scales. These authors state that values generated from the scales can be treated as interval level data. This is especially true when multiple questions are uSed, with the scores summed to generate an overall value. Probability distributions (Blalock, 1972) and the tendency of random regression toward the mean (Stanley & Campbell, 1963) tend to increase the likelihood of a normal distribution in these instances, permitting the use of parametric data analysis, usually considered more powerful and sophisti- cated than nonparametric (Siegel, 1956). 63 The questions will be treated as indexes rather than scales given their relatively high level of validity and ease of construction, as compared to scales (Babbie, 1973). Statistical Testing A variety of testing procedures was utilized, including zero-order correlation, partial correlation, and multiple regression, from which a predictive statis- tical model was derived. In the model the goal was to predict scores on political value measures, given scores on television exposure and other related variables. The dependent political value measures were considered cri- terion variables while the others were considered pre- dictor variables. Finn (1974) states that such a procedure is appropriate with ordinal data or above with Harris (1975) also concluding that such techniques are robust to vio- lations of normal distribution assumptions. The Statis- tical Package for the Social Sciences (Kim & Kohout, 1975) packaged computer programs were utilized to con- duct the analysis. Before statistical testing could begin, several preliminary steps in data treatment were necessary. One of the first was generation of frequency distributions which are reported for the primary independent variable 64 in the next chapter, and inter-item correlation coef- ficients, reported on pp. 76-78. to check for colinear items in the questionnaire, as much of the analysis involved the additive construction of variables. Analysis revealed strong groupings of items with very little colinearity. Also, intercorrelations among all of the preposed independent/criterion variables were checked to determine the possibility of multicolinearity which becomes more likely the higher the predictor correlations. As can be seen on Table l, extreme high correlations between the independent variables are not common, dimin- ishing the likelihood of a multicolinearity problem. Variable Construction Relationships between the dependent variables and their basic items were checked via factor analysis to note the clustering of individual items into the overall value structure. As previously noted, it was felt that six distinct dependent variables would be generated--sportsmanship, authoritarianism, nationalism, need determined expression, value restraint, indi- vidualism, and equalitarianism. While this was the hypothesized number of factors, an open factor solution, in which the maximum number of factors with an eigenvalue of at least 1.0 were generated, was utilized in order to discern exactly how the variables were factoring. Hypo- thesized groupings were as follows (see p. 66). 655 ooo. omo. Lao. ooo.I oHo.- omo.u ooo. omo.u moo. ooo.I oeo.- moo.I .uoma .u m mo mean. «no. oo~.I ooH.I o~H.I oo~.I ooH.I «so. ooo. HNH.I ov~.I ooo.u coouauoq ooH.I moo. omo. moo. moo. ooo. ooo.u moo. omo. coo. omo. ovum oao.I moo.u ooH.I moo.I voo.u omo. oHH.I ooH.I Hoo.u ooa.u ova.u use omo.u moo. ~os. omo. mos. has. omo. oHH. moo.I ooa. new. xom ooo. ooo. Hoo. moN. ooo. oma.I ooo.u and. mos. ~NH. .m.m.m omN. sow. «so. poo. mom. omH. oo~. oom. oo~. Hoonum Hem. saw. one. osm. mos. aflm. mom. son. unwoa mom. mom. son. ooo. «so. «as. non. Rosana NNH. mus. mos. omm. moo. mom. .ucH .nuom .umoH oao. omo. moo.I mos. moo. .ucH .Hom .unoH oHH. mas. omo. mas. .>uu:u .uucu voH. ooo. mom. >9 .omo now. Now. .oauua mam. .msoauuaa >sm .mmum .umm .00 . . .m m . o .>um: >9 . s . no Go a mum sum «some see u m H a o .cwo a use u a >em Us H mqm<8 mmdm MOBUHDHKQ\BZNOZHQHOZH mmB mo mefidz ZOHB< vs. so. No.I oo. oo. mo. so. on. muozuo co voodoo u.coouuoo so. mo.- no. so. no. so. so. No. .ocosa as oo.uuom oo.I oo. as. GN. o~. oo. ss. os. :30 Aswan so once» museoauuos oo.I mo. os.I so. so.. so. so. om. momma m>oa asses» ssaqum oo.- ss.- oo.I so. os. No. mo. s~. sea a m>mn ossoen ssosou moosm> ozIIvm me. so. ms. mo.I so. as. os.I os. nouns suuo3uueoo us ooIImm Ns. mo. oo.u mN. oo. ~s.I no. es. soon: so unmss soumooo ozII~m mo. oo. me. so. so. ms. es. mm. Emsosusuo cosmsos ozuusm oo. Ns. oo. ss.- so. ss.- on. em. Season sou ocaumunom so. os. so. so.. No. mo. as. so. uoouos seems as mesuaIIoN so.I so.I m~. so. so. ~o.I mo. mm. unannoAEs boos oocmsooBOIIos so.- so.I ms. mo.- oo.I vs.I no. em. susuosusa Oceanus unszIIom ~o.I so. as. no. No.n mo. om. om. cosssonmu sououoo mososuuoo as. as. ss.: mo.- mo. mm. oo.I on. umosom msmnuoa amouumm ms. os. mo. oo.. «o. oo. so.I No. use yoga mssmouuo~ ss. m4. ss. ss.- mo. om. oo.I no. sense umsmsauum~ so. sag oo. ss.- mo.- ss. so. mm. swuoonm umnuon mamauu- os. ov. mm. so. so. ms. mo.I em. use omssao umsasauuso no.I mo.I ss. os. oo. ssm._ no. on. Roxane mo>os nouoouuoo ss> s> > >s sss ss s auwamcoSEoo EOOH ouwoccOwummso WHOUUME UGO NDQHUOOQ HOUUflh N mqmda mmoeudh Qz< mUZHodoq moaodm 68 TABLE 3 FACTORS AND VARIANCE ACCOUNTED FOR Pot. of Factor Label Variance Explained I Nationalism 20 II Sportsmanship (I) 18 III Equalitarianism 11 IV Individualism 11 V Need Determined Expression/ Value Restraint 9 VI Authoritarianism . 19 VII Sportsmanship (II) 11 99 (esti— mated) 69 does not seem to be a "real” factor in that there is nothing, upon examination of the items, which makes it unique from the other "sportsmanship” factor, Factor I. Therefore, given that more items loaded successfully on Factor I, with a higher average correlation, and the fact that one of the items in Factor VI is relatively low, this particular factor and its associated items were deleted from further analysis. In addition, items 38 and 32 were deleted from their respective factors due to their extremely weak loading (coefficients below .30) in apparent disparity from the others. Thus, the factor labels and their average variable loadings emerged as follows: Authoritarianism .64 Individualism .62 Nationalism .55 Sportsmanship .53 Need Determined Expression/ Value Restraint .51 Equalitarianism .42 Not only was the factor analysis indicative of support for the hypothetical structuring of variables, it was also utilized to generate more precision in the dependent variables, as factor score coefficients were utilized to construct the additive measures as described by Kim: X1 = Fa x (Ya - Ma)/SDa + Fb x (Yb - Mb)/SDb . . . 70 where: X1 = additive scale 1 Pa = factor score coefficient of variable a Ya = value of variable a Ma = mean of variable a SDa = standard deviation of variable a, etc. Therefore, the basic dependent variables were constructed as follows: Sportsmanship = .28331 x (Q. 20 - 2.89)/1.033 + .20646 x (Q. 23 - 3.113)/.944 + .31724 x (Q.24 - 3.14l)/.865 + .23014 x (0.25 - 3.393) /.809. Authoritarianism = .17898 x (Q. 26 - l.893)/l.066 + .45630 x (Q. 27 - l.548)/1.044 + .37487 Nationalism = .36425 x (Q. 29 - 2.646)/l.061 + .38009 x (Q. 30 - 2.142)/l.229 + .20307 x (Q. 31 - 2.189)/l.107. Need Determined Expression/ Value Restraint = .26510 x (Q.33 - 2.673)/l.004 + .47347 x (Q. 34 - 2.125) /1.130. 71 Equalitarianism = .22286 x (Q. 35 - 1.867)/l.105 + .22810 x (Q. 36 - .742)/.833 + .30701 x (Q. 37 - l.592)/.94l. Individualism = .45357 x (Q. 29 - 2.019)/.982 + .41039 x (Q. 40 - 1.664)/l.063. 92M Numerical values were assigned to each response according to the values indicated on the questionnaire for a range of from one to five on each item. Coding of the dependent values therefore indicated the following: Sportsmanship.--Four items, for a possible range of 4-20, lower scores indicating less sportsmanship, higher scores indicating more sportsmanship (correlation coefficient signs reversed for clarity). Authoritarianism.--Three items, for a possible range of 3-15, with lower scores indicating high authori- tarianism, higher scores indicating low authoritarianism. Nationalism.--Three items, for a possible range of 3-15, with lower scores indicating high nationalism, higher scores indicating low nationalism. Need Determined Expressiop/Value Restraint.--Two items, for a possible range of 2-10, with lower scores indicating higher need determined expression, and higher scores indicating higher value restraint. 72 Equalitarianism.--Three items, for a possible range of 3-15, with lower scores indicating high equali- tarianism. Individualism.--Two items, for a possible range of 2-10, with lower scores indicating higher individualism. Coding of the independent variables was accomplished as follows: Sports Television.--A measure of perceived exposure to and involvement in televised sports events, compiled by summing items 1 to 7, with item 7 reverse scored, for a possible range of 7-35, lower values indicating greater involvement in and exposure to televised sports events. Participation in Sports.--A measure of degree of involvement in actively playing sports, compiled by summing items 9 and 10, for a range of 2-10. Lower scores indicate greater participation in sports. Sports Attendance.—-A measure of in-person view- ing of sports events, compiled by adding items 12 and 13, for a range of 2-10. Lower values indicate greater involvement in spectator sports. Type of Sport Preferred.--Questions 8, 11, and 14 were coded "0" if the individual preferred a team sport, ”1" if the individual preferred an individual sport, and 73 summed. Thus, lower scores indicate a preference for team sports, higher scores a preference for individual sports. General Television Viewing.--A measure of other potential television socialization factors, compiled by adding responses to items 15 through 19, for a range of 5-25, with item 16 reverse scored. Lower values indicate a higher exposure to and involvement with television in general. Institutional Conservatism.--Compiled by adding items 47, 48, and 49, with 48 reverse scored. Lower levels indicate more perceived conservatism among tra- ditional socialization agents. Possible range 3-15. Institutional Political Interest.--Compiled by adding items 44, 45, and 46, for a range of 3-15, lower scores indicating higher levels of political interest. Institutional Sports Interest.--Compiled by adding items 41, 42, and 43, for a range of 3-15, lower scores indicating higher levels of perceived interest in sports. Family Influence.--Compiled by adding items 41, 44, and 47, for a range of 3-15, lower scores indicating greater family involvement in dimensions under study (sports, political values, conservatism). 74 Peer Influence.--Compiled by adding items 42, 45, and 48, with 48 reverse scored. Possible range of 3-15, lower scores indicating greater peer involvement across dimensions under study. School Influence.--Compi1ed by adding items 43, 46, and 49. Possible range 3-15. Lower scores indicative of greater involvement across dimensions. Socioeconomic Status.--Compi1ed by combining items 52 and 54 (54-52) with 52 recoded as follows: (1 = 30, 2 = 45, 3 = 60, 4 = 75, 5 = 90), thus permitting consis- tency with item 54, coded with North-Hatt Occupational Rating Index, which utilizes a scale of O to 100, higher scores indicating higher socioeconomic status of the occupation. Therefore, overall, possible range of -90 to 70, with higher values indicating higher family socio- economic status (correlation coefficients then reversed for clarity). Sex.--Question 51, as stated. Race.--Question 53, with 4 recoded as 0, others recoded as 1, thus dividing the population on a white/ nonwhite basis. Age.--Operationalized by grade level, in Question 50. Lower values indicate lower grade level on a 9th 75 through 12th grade range (correlation coefficients then reversed for clarity). Location.--Coding as a dummy variable assigned respondent according to which of the three tested school districts survey was conducted. District ”A” = (0), District ”B" = (1), and District "C“ = (2). This coding, while basically nominal, also provided an ordinal ranking of the areas according to their perceived conventional conservatism (from low to high). Hypothesis Testing Hypotheses 1 through 7 involved zero order cor- relation, using the Pearson Product Moment coefficient and simple regression, given the testing of only one dependent and independent variable in the hypothesis. In the correlational analysis, Student's T was applied, utilizing a one-tailed test, as direction has been pre- dicted in the relationship (Blalock, 1972). In the regression, F test procedures were utilized. In each case, significance levels at the .05 level or better were predicted, providing strong justification for rejecting the null hypothesis of no significant dif- ference, especially given the exploratory nature of the study. In addition, the strength of the relationship, as well as significance, was considered with the pre- diction that significant correlation coefficients '76 TABLE 4 INTER-ITEM CORRELATION COEFFICIENTS --...' 3—“; ,_. r' Inter-item Correlation for Variable Exposure to and Involvement in Television Sports £225 1 2 3 4 5 6 7 E32 1--Amount of Sports Viewing .830 2--Like Sports TV .779 .840 3--Like Anncrs. .407 .442 .624 4--Learn from STV .557 .595 .388 .717 5--Relaxing .452 .511 .337 .396 .352 6--Exciting .577 .609 .383 .461 .481 .738 7--Watch if Nothing Better On -.500 -.503 -.271 -.344 -.304 -.433 -.685 Inter-item Correlation for Variable Participation in Sports Item 9 10 Part. 9--Like to Play Sports .821 lO--Amount of Sports Playing .567 .927 Inter-item Correlation for Variable Attendance at Sports Events as a Spectator Item 12 13 Attend. 12--Like To Go To Games .810 l3-—Amount of Attendance .516 .915 Inter-item Correlation for Variable General Television Viewing £535 15 16 17 18 19 Egfl' 15--Like TV .634 16--Amount Watched .507 -.628 l7--Like All in the Family .210 .097 .514 18--Like Violence .270 .217 .152 .602 19--Like News .107 .139 .019 -.051 .258 '77 TABLE 4--Continued i M“:- Inter-item Correlation for Variable Type of Sport Preferred Isa 8--Favorite Sport- TV ll--Favorite Sport (play) 14--Favorite Sport (watch) 8 11 14 Typ. Spt. .055 .489 .166 .428 .343 .063 Inter-item Correlation for Institutional Conservatism 11:2 47--Family Conserv. 48--Friends Liberal 49--School Conserv. 47 48 49 9233!. .644 -.044 -.563 .019 .017 .547 Inter-item Correlation for Institutional Interest in Politics Essa 44--Fam. Interest 45--Friends Int. 46--School Impor- tance 44 45 46 Political O 685 .297 .715 .228 .256 .702 Inter-item Correlation for Institutional Interest in Sports Lea 4l--Fam. Int. 42--Friend Int. 43--Sch. Importance EEe_m 41--Fam. Int. Spts. 44--Fam. Int. Pol. 47--Fam. Conserv. 41 42 43 Sports .759 .463 .781 .295 .414 .751 Inter-item Correlation for Family Influence 41 44 47 Fami y .667 .019 .466 .133 -.031 .590 78 TABLE 4--Continued Inter-item Correlation for Peer Influence Item 42 45 48 Peers 42--Int. in Sports .664 45--Int. in Politics .080 .562 48--Libera1 -.040 .155 -.465 Inter-item Correlation for School Influence 13:2 43 46 49 School 43--Sports Imp. .644 46--Politics Imp. -.012 .540 49--Conserv. .110 .077 .577 Inter-item Correlation for Variable Socioeconomic Status Item 52 54 $85 52--Income .305 54--Occ. Rating .401 .998 79 between the primary independent variable and the dependent variables will average .200 or stronger, a level viewed as indicative of a fairly respectable relationship in the social sciences (Flanigan & Repass, 1968), will be obtained. Hypotheses 8 through 15, involving additional control factors, will be tested with both partial cor- relation and multiple regression. Similar statistical standards will be applied. Hypothesis 16, however, will be tested only via multiple regression, given the large number of variables making a complex order partial cor- relation relatively unworkable. However, in such an analysis, the standardized regression coefficients, or "beta weights" assigned to the predictor variables, may be interpreted as the partial correlation between the predictor and the dependent variable, while controlling for all of the other predictors in the equation. The testing of this hypothesis will represent the most complete testing procedure in that it will attempt to assess the impact of each variable, and all variables Simultaneously, in terms of strength of relationship, variance explained in the dependent/cri- terion measures, and significance. Thus, a socialization model in which scores on political value measures may be predicted by scores on the independent variables may be constructed. 80 In the construction of this model, step-wise multiple regression procedure will be utilized in which the variable contributing the most significant explained variance in the dependent variable is entered first, the variable explaining the next most entered second, and so on. Thus, the relative contribution of each variable in a multiple context can be discerned. This method of regression was chosen primarily because of the lack of previous theory which has readily developed models discussing the relative strength of each variable in interaction with the others (Kraus & Davis, 1976), and the fact that there is little support for the time order- ing of the variables, given their early presence in the life of the adolescent and their constant presence over time suggesting a nonrecursive model (Kim & Kohut, 1975). The choice of this method of regression is further sup- ported by the relative 1ack of colinearity in the pre- dictor variables indicating that the order of entrance into the equation will make little difference in terms of overall variance explained, or that explained by each individual variable, and will also provide, for signifi- cance testing purposes, a more well-constructed model with which to perform step-wise F testing (Finn, 1974). In testing Hypothesis 15, the same procedure for sub-analysis of each individual school district will be followed. 81 While numerous sub-analyses and interaction effects testing are possible, this study will limit itself to general analysis and construction and testing of a general model given the exploratory nature of the research and the limitations on time and available resources. Given these testing procedures, it is generally theorized that certain political values, especially those of a more conservative nature, will be related to exposure to televised sports. CHAPTER IV DATA ANALYSIS Descriptive Information Amount and Preference for Sports Viewing Consistent with conventional wisdom, sports view- ing, as revealed by the questionnaire, seemed to be a popular and frequent activity among the students sampled. When asked to agree with a statement which came closest to the general amount of television viewing they did, the students responded as follows: % N (1) I watch most sports events telecast ‘IO 61 (2) I watch many televised sports events 27 161 (3) I watch sports events occasionally 37 223 §'= 2.9 (4) Rarely do I watch a televised sports event 15 87 (5) I almost never watch a sports event on'rv 2.92. 100 598 Only a little over a fourth (26 percent) of the students reported that they rarely or almost never watched such programming, while almost three-quarters (74 percent) said they watched sports on television at least occasionally. 82 83 Similar findings were reflected in a question which dealt with feelings toward sports viewing. When asked to agree or disagree with the statement, ”Watching sports on television is something I really like to do" the following responses were obtained. % N (l) Strongly Agree I6 94 (2) Agree 29 173 _ (3) Neutral 37 220 x = 2.6 (4) Disagree 13 76 (5) Strongly Disagree 6 34 Just one-fifth of the sample (19 percent) dis- agreed with the statement to any extent while almost half (45 percent) agreed. Responses seem to indicate overall strong popularity for such programming despite a wide variance in the value teenage viewers place on it and the extent to which they actually watch it. While Real's content analysis indicates that coverage of actual sports action makes up only a small percentage of the program time, there is some indication that this aspect of the programming is more appreciated than the overall television "package" within which it is placed. While Real discussed the potential of analysis to influence audience reaction, it is less enjoyed than the subject of the programming itself. When presented with the statement "I enjoy the announcers and commentary on sports telecasts, because they add a great deal to the program," the following distribution of responses occurred: 84 (l) Strongly Agree 6 35 (2) Agree 26 156 ._ (3) Neutral 39 231 x = 3.0 (4) Disagree 19 115 (5) Strongly Disagree 10 60 While the distribution is Similar to that of sports programming enjoyment in general, it is somewhat less supportive. Types of Sports Preferred When asked to indicate the sport they enjoyed watching the most on television, the following sports and distributions were indicated: % N Football 22 130 Skiing 15 91 Baseball 11 64 Tennis 11 63 Basketball 10 58 Track and Field 7 41 Swimming 7 39 Gymnastics 5 31 Hockey 3 19 Golf 2 12 Other _‘1 _3_9_ 100 597 The survey results are similar to those of the National Nielsen ratings with the exception of the sur- prisingly high popularity of skiing, which is televised relatively infrequently and usually as part of other "multi-Sport' programs such as “Wide WOrld of Sports." Geographical factors making Michigan a popular skiing area could be a possible explanation for this deviation limiting the generalizability of this particular bit of information. The timing of the study (March) could also 85 be a factor. Otherwise, the popularity of team sports such as football and the relatively unpopular nature of sports such as golf, with its older national demographics, and hockey, which has been spectacularly unsuccessful on 0.8. commercial television, are consistent with national survey data. Also interesting is the diversity apparent in the sport considered to be the ”favorite." Seventeen dif- ferent sports were mentioned as being the favorites of one or more of the respondents. Such a variety is incon- sistent with conventional wisdom which tends to concep- tualize sports television as being essentially limited to the more popular team sports. Functions of Sports Telecasts Questionnaire responses also seem to indicate that viewing is purposive rather than haphazard and is felt to serve some function for those who view it. Respondents were asked to indicate if they felt such programming served an educational, escapist, or arousal function for them, and their answers were as follows: ”I watch sports on television because I learn a lot from it.” % N (1) Strongly Agree ‘6 335 (2) Agree 22 134 ‘_ (3) Neutral 37 223 x = 3.1 (4) Disagree 25 151 (5) Strongly Disagree _2 .LEE 99 595 86 "I watch sports on television because it relaxes me and helps me to forget about my everyday problems.“ % N (l) Strongly Agree _6 '38 (2) Agree 24 142 ._ (3) Neutral 30 179 x = 3.2 (4) Disagree 28 165 (5) Strongly Disagree 12 72 100 596 "I watch sports on television because it's excit- ing and gets me emotionallyg involved. ' (1) Strongly Agree I5 92 (2) Agree 38 225 ‘_ (3) Neutral 28 167 x = 2.6 (4) Disagree ll 63 (5) Strongly Disagree 8 48 100 595 "I watch sports on television only if there is nothing better on.“ % N (1) Strongly Agree '—8 19 (2) Agree 21 126 ‘_ (3) Neutral 23 139 x = 3.2 (4) Disagree 34 202 (5) Strongly Disagree 13’ 79 99 595 Great variance is apparent in the function sports programming serves for those questioned. Still, certain tendencies emerge. The fact that almost one-half of the group (47 percent) disagree with a nonpurposive statement about the function of televised sports (watching only when there is nothing better on), and only 29 percent agrees with the statement, is indicative of the active role such programming plays in the behavior of the respondents. Also apparent is the "arousal" function of the programming demonstrated by the fact that over 87 half (53 percent) find such content exciting and emo- tionally involving while only 19 percent do not. Follow- ing Tannenbaum's “arousal hypothesis," (1977) which gen- erally states that such emotional arousal is a necessary precondition for effective impact of the television message upon the viewer, especially for younger viewers, the potential for an effect of sports programming on the viewer becomes more probably given the fact that such arousal is actually sought by the viewers and is one of the reasons why they watch sports on television. Similarly, 40 percent of the viewers disagree with the statement that sports events on television are relaxing. While only 28 percent feel that they learn a lot from televised sports, a similar number, 34 percent, feel that they do not. The tendency of the respondents to view television programming of this sort as purposive and arousing is more interesting in light of the use of television in general by viewers which tends to be nonpurposive and noninvolving. The fact that sports telecasts are counter to this trend makes them somewhat unique. This highly involving aspect of Sports television is further demonstrated by the high inter-item correlations for the items used in the construction of the Sports tele- vision variable (see p. 76). Those who watch a lot of sports events on television tend to enjoy them a great 88 deal (r=.779), learn from them (r=.557), and become emotionally involved in and excited about them (r=.577). The possibility that this viewing is largely incidental is further refuted by the strong negative relationship between amount of viewing and disagreement with the state- ment "I watch television sports only if there is nothing better on” (r=-.500). These correlations support the conceptualization of sports viewing and involvement as a consistent activity and also support the construction of an additive variable in this manner from concepts which could be potentially independent. The Sports Viewer A demographic profile of the sports viewer in the sample indicates similarities to those generalizations drawn from both conventional wisdom and marketing data. While no significant difference exists between race and amount of sports viewing (r=.095, p=.095), viewing by whites is higher than that by nonwhites, and such viewing is related to sex, with more male than female viewers (r=.242, p=.001), although the relationship is much weaker than conventional wisdom would indicate. The sports viewer also tends to come from families with higher socioeconomic status levels (r=.122, p=.002). The finding is not surprising given the fact that this ten- dency has been noted by both advertisers and television 89 programmers in the past, is instrumental in the great advertiser support such programming on commercial tele- vision enjoys, and is consistent with sports development. Also, not surprisingly, the sports viewer tends to be associated with other social groups which are interested in sports with the strongest relationship noted between sports viewing and sports interest within the family (r=.512, p=.001) followed by peer group (r=.495, p=.001) and school (r=.412, p=.001). To some extent, conventional wisdom is also sup- ported by the strong relationShip between sports viewing and participation in sports (r=.513, p=.001) and attendance at spOrts events as an in-person spectator (r=.562, p=.001). However, such relationships also tend to refute other speculations regarding the interaction of these variables. The tendency of those who watch sports on television to also participate in sports activity con- tradicts the idea that such television programming encourages passivity and reduces participation. Simi- larly, the relationship of sports viewing and in-person attendance is consistent with previous research on the long-term effects of television coverage on sports attendance. The high correlations tend to indicate that individual involvement in sports is a consistent activity and manifests itself in various ways. 90 Conversely, although sports viewing and general television viewing are related (r=.341, p=.001), indi- cating that those who watch a lot of sports tend to watch a lot of television, and vice-versa, the relationship is not overwhelming and is not as strong as that between sports viewing and other types of sports activities. Thus, it seems that to some extent the sports viewer is different from the general television viewer in terms of amount and types of television viewing preferred. The close relationship between sports viewing and other types of sports involvement is also demonstrated by examining the relationship between the type of sport preferred when watching on television, in person, and participating. There is a fairly strong relationship between type of sport preferred on television and in person (r=.428, p=.001) and in participation (r=.488, p=.001), a relationship even stronger than that between participation and in-person attendance with regard to a particular sport (r=.343, p=.001). Thus, it seems that televised sports is a sig- nificant part of the involvement in sports for those so predisposed, more so than televised sports is a signifi- cant part of the entertainment of the general television viewer. Another demographic factor is that sports view- ing is inversely related to grade level although the 91 relationship is a moderate one, with amount of viewing and involvement in viewing decreasing as age level increases (r=-.l40, p=.001). Value Measures Examining frequency distributions for the depen- dent variables revealed the information presented in Table 5, page 92. By breaking the additive measures into low, medium, and high categories, certain trends are evident. Medium levels of values predominate except in the area of sportsmanship where high levels of this value are reported, and very few are rated as low in sportsman- ship. Otherwise, the respondents tend to be higher in authoritarianism, lower in nationalism, and high in both egalitarianism and individualism with the balance between need determined expression and value restraint almost even. The large number of “medium” respondents suggests either strong agreement balanced by strong disagreement or a large number of respondents answering “neutral” on the Likert scales. Individual item analysis, which reveals approximations of normal response distributions for those items, suggests the latter to be the case. In such a situation, the question of what the "neutral“ response means arises. Does the respondent wish to signify a position half-way between strongly agree or 92 o.ms Loos unsouuumm UsHm> Aav m.H Ammv H.¢ Away e.~ Amway ~.om v.aa 30A Achy o.mhamm¢v Hmuusoz Ahowv m.mm Admev o.mn Abhmv m.Hm Amhmv h.om m.om Enema: Anhmv no.sslooo .oxm .uoo oooz Ahmmvwm.¢m Ahmavwm.mu Aamuvwa.om Aomvma.m «5.5N sown Achav .oamuom .>soos . .sovu .s.uoz .euoa m>\moz season.“ Human maso> mmfldem¢> Bzmazmmmo m0 mZOHBDmHmBmHQ HUZHDOfimh m manta 93 disagree, is he undecided, or does he not hold the value in question to the extent necessary to respond with any strong degree of feeling? While theory regarding value development would suggest the latter, a definitive answer from the data alone is not available. Response bias is not particularly evident given that the higher levels of authoritarianism required agreement, while the lower levels of nationalism and higher levels of sportsmanship required disagreement with the items as listed in the testing instrument. Among the values, however, individualism, equali- tarianism, and sportsmanship were the most consensual. Hypothesis Testing, The initial step of the hypothesis testing involved the bivariate correlation of sports television with the political values previously identified and Operationalized. Specifically, these values were authoritarianism, nationalism, equalitarianism, indi- vidualism, need determined expression/value restraint, and sportsmanship. Bivariate Pearson correlation coef- ficients and simple regression were utilized to test the hypothesis that the possession of these values was associated with, and indeed a function of, exposure to and involvement with sports television. 94 Hypothesis 1: Higher levels of authoritarianism are associated with higher levels of exposure to televised sports. This hypothesis was supported. A Pearson cor- 2=.oa) was derived relation coefficient of .281 (p=.001, r when examining the covariate relationship between the two. The relationship is statistically significant and is in the hypothesized direction, indicating that £2223 more involved with televised sports tend to be more authoritarian. Simple regression produced similar results. Hypothesis 2: Higher levels of nationalism are‘associated with higher levels of exposure to televised sports. This hypothesis was also supported. A Pearson correlation coefficient of .372 (p=.001, r2=.14) was derived from the relationship between the two variables. Once again, the relationship is statistically significant and is in the hypothesized direction indicating that those more involved with television sports tend to be more nationalistic. Simple regression produced similar results. Hypothesis 3: Higher levels of value restraint and lower levels of need-determined expression are associated with higher levels of exposure to televised sports. 95 This hypothesis was ESE supported. While the relationship was significant (r=.110, p=.005, r2=.02), it demonstrated a relationship opposite of the one hypothesized. Simple regression produced similar results leading to the conclusion that those more involved with television sports tend to be more need-determined expressionistic and less value restrained, thus,.1ess traditionbound and more self-actualized. Hypothesis 4: Lower levels of equalitarianism are associated with higher levels of exposure to televised sports. This hypothesis was BEE supported. Pearson cor- relation coefficients were 22E significant (r=.048, p=.126, r2=.00) nor was the relationship at all strong, and, when rounded, no significant variance was explained by the interaction of the variables. Hypothesis 5: Higher levels of individualism are associated with higher levels of exposure to televised sports. This hypothesis was not supported. While the relationship was significant, as revealed by the cor- 2=.01), it is relation coefficients (r=-.O70, p=.047, r extremely weak, is in the reverse direction, and explains little variance. Moreover, simple regression reveals 96 the standardized beta weight for the predictor variable to be not significant (p=.104). Hypothesis 6: Higher levels of sportsmanship are associated with higher levels of exposure to televised sports. This hypothesis was supported. The relationship is (r=.108, p=.005, r2=.02) statistically significant and in the hypothesized direction, indicating that higher levels of sportsmanship are a function of exposure to televised sports, although the low coefficient indicates a somewhat limited relationship. Overall, some support for the general hypothesis of the role of sports television as a developer of con- servative values is supported. Of the six individual values tested, two are related to a fair degree in the hypothesized direction, one to a weak degree in the hypothesized direction, and one to a weak degree in the opposite direction with the other two lacking significance and/or strong direction one way or the other. It cannot be said, at this point, that exposure to such programming acts as a liberalizing factor, nor can it be said that no significant relationship exists between the independent and dependent variables. Intercorrelation of the additive measure of con- servatism with the individual values tended to support 97 the conventional wisdom construct as the following coefficients were obtained: Conservatism with Authoritarianism r = .636, p = .001 Nationalism r = .546, p = .001 Need Determined Expression/Value Restraint r = -.361, p = .001 Equalitarianism r = -.316, p = .001 Individualism r = .267, p = .001 Sportsmanship r = .442, p a .001 Hypothesis 7: Higher levels of conservatism are associated with higher levels of exposure to televised sports. This hypothesis was supported. Pearson cor- relation coefficients of .241 were derived when comparing the two variables (p=.001, r2=.06). The correlations revealed a relationship in the hypothesized direction and of statistically significant proportions explaining significant amounts of variance (r2-.06). Simple regression statements reveal similar results as they have consistently throughout this testing procedure. Support for the overall theoretical perspective of the study is thus in evidence with exposure to and involvement with televised sports events related to more conservative values. 98 mo. w as soo. v i.m.zo~os. s.m.zv~o~. sso. soo. v soo. v mso. o «mm .H «mm .H «mm .H vmm .H emm .H vmm .H vmm .H at om.~m wm.~ w~.H mw.w mm.mm Ho.mv o~.m m sew. soo. moo. aos. «hm. Ham. mos. u was: mo. so. so. No. «a. mo. so. mm sauce Hem. ono.I avo. mos. «pm. How. mos. sowns>osoe i O ¥ O Q mquOnmm n s >umso osH doom \Mxmmowov H.unz nuns momsanumm oaossum> nusmwowmmooo cownnwumom HouUonHm omusosooooum one mosoosuo> oosuoosso mnm>mn NDA<> UZHBUHDfimm ZOHBdDOH ZOHmmmmUmm flAmSHm m flqmdfi 99 Hglationships Among the Dopendent Variables The values discussed and tested are complex in that at least theoretically they are all independent yet, to some extent, they are related to and a function of the overall conservative value system. Thus, to be appro- priately tested in this context, measures and operation- alizations of these values should produce relationships which are strong and significant but which are not so strong as to be colinear and, thus, essentially measure the same thing. An analysis of this potential problem at the bivariate level, via correlation matrix of the dependent variables, reveals the possibility of problems in this regard to be relatively slight. ConfoundingVVariables Numerous other variables, as previously discussed, have been ascertained to be a factor in the political socialization process and are involved in the relation- ship between televised sports and political values. While basic bivariate comparison supports the thedry being tested, these various additional variables must be controlled if the relationship between the primary dependent and independent variables is to be correctly assessed and not artificially inflated or reduced. Thus, much of the rest of the data analysis concerns itself with controlling for one or more, or all, of the 100 A a oss.I .oso. soo.- ooo.I moo. ~44. assoooSousoom ooo. oos. «moo. .oso.u pom. Sassooos>soos sms. soo.: .mso. osm.u amsoossmussosom oms. ooo. soo.- uososuumm moso> « \sowmnmumxm owswsuoumo ommz vmv. ovm. _Emsamsowusz mam. Emwdswusuwuoouod Enwum>ummsoo saw was sooom \wwmmowwm sumz room >uomooo mmqdedew ZOHmmBHmU\Bzmn—memo .mO xHMHEZ ZOHngOU 203mm MB¢§>HQ h flqmdn. 101 other identified possible confounding variables via both partial correlation and multiple regression procedure. Hypothesis 8: These values will be associated with higher levels of exposure to televised sports when controlling for participation in competitive athletic events. Results of hypothesis testing provide mixed results fairly similar to those produced by the zero-order correlations. Controlling for participation in athletic events changes the relationship between sports television involvement and political values very little, although sports viewing and sports participation are related; and both are related to the dependent variables. Participation and Sportsmanship .130, p = .001 Authoritarianism. .151, p = .001 Nationalism .138, p = .001 Need Det. Exp./ Value Res. .018, p = .330 (N.S.) Equalitarianism .060, p = .070 (N.S.) Individualism -.056, p = .084 Conservatism .132, p = .001 The relationships between the dependent variables and participation in sports are similar to those found with involvement in sports via television. First order partial correlations controlling for the effect of par- ticipation when considering sports television and politi- cal values derived the following coefficients and sig- nificance levels: 102 Sports Television with Sportsmanship .048, p = .133 (N.S.) with Authoritarianism .239, p = .001 with Nationalism .354, p = .001 with Need Det. Exp./ Value Res. .117, p = .004 with Equalitarianism .020, p = .319 (N.S.) with Individualism -.O49, p = .130 (N.S.) with Conservatism .204, p = .001 With the exception of participation acting as a confounding variable with respect to the interaction of sports television and sportsmanship, the strength, direction, and significance of the relationships remain relatively stable, providing more evidence for the relationship between conservative values, especially authoritarianism and nationalism, and exposure to and involvement with sports television. Similar results are obtained when developing a step-wise regression model utilizing the two predictor variables of participation and sports television in an effort to predict levels of each of the political values. The following equations are obtained (see Table 8, p. 103). The models which allow for the effect of each predictor point out the relative strength of the sports television variable, as compared to the participation variable, which has received more attention in previous literature. Only in one instance does participation explain significant amounts of variance in the criterion variables greater than that explained by sports television 103 oonoosoosmum osm mUHQMon> soonwuwso 2. m o. soo. v i.m.zoo-. i.m.zoomm. soo. soo. soo. moo. o .mmm .m mmm .~ mom .~ mom .m mmm .~ mom .~ mom .~ so mo.ms oo.s oo.s oo.m os.oo oo.- o~.m a soo. moo. soo. oss. oom. soo. oms. s Ass: mo. so. so. so. 4s. oo. so. No soooe sso. o~o.I oso. ~mo.- soo.- oso. .sos. oosuaososssoo .mmm. omo.I «mo. mos. soo. mom. mmo. >9 .msom R i a. umom Hm> >Hmso ooH Hmsom \mxm oomz Humz muse mandamumm moabtass> musmmoommmoo sowmmosmmm Houoooosm m mflmda mBmOmm NMOBdeUHBmdm 02¢ ZOHmH>mAmB mBmOmm "ZOHadbom ZOHmmmmwmm 104 while the latter variable does so four times. Thus, the theoretical assumption is supported. Hypothesis 9: These values will be associated with higher levels of exposure to televised sports when controlling for in-person attendance as a spectator at sports events. Overall, the hypothesis was supported. While evidence of the confounding effect of in-person attendance on sportsmanship was noted, a mild suppressor effect on equalitarianism and individualism was also noted with no change in the other variable relationships in terms of direction or Significance. The zero order correlations of in-person sports attendance and political values are consistent with those of the previously discussed independent variables as noted below: In-person spectator sports attendance with Sportsmanship .129, p = .001 with Authoritarianism .168, p = .001 with Nationalism .195, p = .001 with Need Determined Expression/ Value Restraint .020, p = .319(N.S.) with Equalitarianism -.Ol9, p = .320(N.S.) with Individualism -.005, p = .452(N.S.) with Conservatism .204, p = .001 Thus, such attendance is related to higher levels of sportsmanship, authoritarianism, nationalism, and general conservatism. The fact that it is also highly related to participatory and television sports, which are also related to dependent variables, makes it 105 imperative that its effect be controlled for when examin- ing the relationship between televised sports and con- servative political values. When performing first-order partial correlations to exert control, the following results were obtained: Sports Television with Sportsmanship .043, p = .161 (N.S.) Authoritarianism .228, p = .001 Nationalism .323, p = .001 Need Determined Expression/Value Restraint .119, p = .003 Equalitarianism .072, p = .049 Individualism -.082, p = .029 Conservatism .156, p = .001 The results indicate that when controlling for in-person attendance at spectator sports events, exposure to and involvement in televised sports is associated with higher levels of authoritarianism, nationalism, conservatism, and need determined expression, and, to a lesser extent, with higher amounts of equalitarianism and lower amounts of individualism. Thus, in the aggre- gate, the hypothesis is supported. Similar results are obtained by entering the variables in a predictive regression equation so that they may exert control on each other (Table 9). The results are consistent with those previously obtained with other equations. Involvement with sports on television again seems to be, overall, not only a significant, but a stronger,gpredictor than its control variable, in this case in-person Spectator attendance. 106 mo. w a. doc. v A.m.zvmwa. A.m.zvm-. owe. Hoo. v Hoe. v ooo. o mmm .N mmm .N mmm .N mmm .N mmm .~ mmm .m mmm .N no mv.ms Hm.a we.a mm.m mm.~v Hm.- Ho.m m mmw. «mo. «so. mad. mom. Haw. emu. H pass we. Ho. Ho. no. es. mo. «o. «m sauce IIII. IIII IIII IIII. IIII. Idll. «cos. omo. ooo.I soo.: cmo.I mac. «ooH. .osuud .uoomm «ems. mmo.I moo. «ova. «mmm. aaom. Hmo. >9 .muom mm s >umso osH Hmsom \Mxnmowwm Humz zoom momsEmumm mmaomwus> muomwowmmmoo cowmmoummm uouomomum oonsouooooum ooo mosomsso> oossousuo mfimomm mOBdavmmm QZd ZOHMH>mAmB mamomm ”ZOHfidbom ZOHmmmMUmm a mam >9 «mow. moo.I moo. ammo. «mam. emu. SNHH. >9 .muom >umso osH Hmsom \Mmmmowwm Humz nuns mandamuom moaosmuc> mucmsommmmoo sommmoumom Houooooum monsoumocsum out mmaomwum> soououwso UZH3MH> ZOHmH>mAMB Admmzmw 92¢ ZOHwH>HQmB mamomm OH wands ”ZOHfidbOH ZOHmmmmwmm 110 on average, than the control variable and is, thus, in each case, the best predictor in the equation, further supporting the hypothesis. Hypothesis 11: These values will be associated with higher levels of exposure to televised sports when controlling for the type of sport (individual v. team) preferred. The limited, delineation of sports into individual and team categories is the only evidence in the literature that different types of sports serve different types of individuals and thus may have differing effects on them. Therefore, the preference for individual or team sports of the individual was also utilized as a control factor in the correlation between sports television and political values, despite the lack of significant relationship with the other independent and most of the dependent variables as indicated below: Type of sport with sports television participation in sports sports spectator attendance television viewing sportsmanship authoritarianism nationalism need determined expression/value restraint individualism equalitarianism conservatism .001, -0049, -.006, .035, .040, -.039, .002, -.093, -.043, -.009, .051, 'O'U'O'U'O '0 '0 "O’U'O'O .488(N.S.) .121(N.S.) .443(N.S.) .206(N.S.) .172(N.S.) .176(N.S.) .477(N.S.) .013 .418(N.S.) .152(N.S.) .114(N.S.) 111 Type of sport shows little relation, by itself, to the other variables even to such values as indi- vidualism, which conventional wisdom would hold to be related. Sports television with sportsmanship .108, p = .007 authoritarianism .280, p = .001 nationalism .372, p = .001 need determined expression/value restraint .109, p = .006 equalitarianism .048, p = .132 (N.S.) individualism -.071, p = .052 conservatism, .241, p = .001 Overall, however, the trends established in previous testing hold, and the hypothesis is supported. This is further verified by placing the two variables into a regression analysis, reported in Table 11. Once again, the relative strength of exposure to sports television as a predictor of political values is demonstrated. Hypothesis 12: These values will be associated with higher levels of exposure to television sports when controlling for the effect of other established socialization agents; such as family, school, and peer group. Due to the importance of these institutions in the socialization process, several methods of controlling for their influence were utilized. Initially, the degree to which all of these institutions were perceived to be conservative politically was considered as a control 112 v « soo. v A.m.zvaoa. A.m.szmm. ooo. Hoo. soo. v omo. o mmm .N mmm .N mmm .N mmm .N mmm .N mmm .N mmm .N no s~.os mm.s «o. oo.m mo.~v H~.m~ hm.m m own. moo. moo. med. mom. mom. mas. H van: oo. Ho. Ho. v «o. «H. mo. so. am Hence «we. moo. moo.I moo.I woo. omo.| ooo. ASST» .> .osmv a usomm mo Game How. ooo.I moo. mos. Hon. omm. mos. >9 mumm « t k. .s a. sham Hn> >smoo oos sooom \axm oooz suoz suns mom:Smuom moaomssm> musmsosmmmoo nowmmosomm Houosomsm oousosoooosm one mosoosum> oossousuo Qmmmmmmmm BmOmm m0 mmMB Q24 ZOHmH>MAmB mfimomm "ZOHB osoosoo .somumsoo so ooooaocm no: mass .muoumemnmm Tonsuomou mmmm mooooxm manomus> AIL .mo. w as soo. v A.m.szHH. omo. woo. Hoo. v Hoo. v Hoo. v m «mm .o «mm .o «mm .v «mm .o «mm .v «mm .e vmm .v «o mo.m om.a om.~ Hm.m Hm.m~ mm.~a H~.o m mom. mas. oNH. ohm. mom. oom. mam. u ass: oo. so. no. mo. ma. mo. mo. mm Hmpoe vmo. soo.: mHo. moo. mao. moo. soo. psE>H>GH 4 « umaom umsH mmo. soo. I Hmo.r omo.I ovo. «ova. uss>a>sH ouom boos ovo. mmo. mHH.I «omo. mNH. moo. omH.I Emmum 4 a « I>uomsoo umsH mom. mmo.I soo. baa. mom. ova. mmo. >9 mumm « s a s >umso ocH swoon umom Ho> Humz ovum momsfimumm \mxm omoz mmaoossn> Houomoosm muamsommmooo sowmmmumom ooumoumosmum was mesonmum> somuoumuo Na quda mmoeodm AGZOHBDBHBmZH 02¢ ZOHmH>mAmB mamomm ”ZOHH¢DOH ZOHmmfimUmm 115 The perceived conservatism of family, friends, and school was considered to be a possible confounding variable given the hypothesized conservative values expressed in the television content. The measure was constructed additively given the rationale that consistent conservative influence from all institutional areas would be more of a factor than more inconsistent influences. Also considered, therefore, were the interests of these groups in sports due to the hypothesized associ- ated values and the interest of these groups in politics. This latter measure was utilized with the rationale that an environment in which politics were consistently viewed as important would be more prone to exert influence on the individual since political tapics would be discussed more, perceived to be more important, etc. As can be seen from the data, stronger effects are noted from television sports, which is the best pre- dictor of values than from these institutional factors. However, institutional conservatism is a significant pre- dictor (the best of the three control factors) on several of the values being associated with lower levels of sportsmanship, higher levels of nationalism, higher levels of need determined expression, and lower levels of equalitarianism. Institutional sports involvement is only significantly related to its most salient value, that being higher levels of sportsmanship. Institutional 116 political involvement is a factor only with significantly lower levels of individualism. In addition, each particular institution, its impact, and possible confounding effect, was examined. To assess the role of family, peer group, and school, both separately and together, the perceived involvement of each group in sports, politics, and political con- servatism was considered. Using first-order partial correlation to control for the role of socialization institutions in the relationship between sports television and political values, the following coefficients were obtained (Table 13). TABLE 13 PARTIAL CORRELATIONS BETWEEN SPORTS VIEWING AND VALUES Controlling for: Sports television with: Family Peers School All Sportsmanship .071: .082: .088: .053* Authoritarianism .243* .254* .253 .221* Nationalism .329 .336 .365* .314 Need Determined Expression/ * * * * Value Restraint .110 .124* .109 .122* Equalitarianism .052 .101 .026 .077 Individualism -.082: -.054 -.057* -.061* Conservatism .185 .191* .229 .160 *p 5 .05 The relative contribution of each socialization agent and sports television was then considered in a step- wise multiple regression analysis (Table 14). 117 .sowmoaosw coflumsvm now mumuoemumm mmmm umofi uos moon Any .mo. w as Hoo. v x.m.zvmo~. soo. soo. soo. v soo. v moo. m Hmm .o Hmm .v Hmm .v Hmm .w Hmm .v Hmm .v Hmm .v up mm.oa m¢.H om.o hm.~ H~.- mv.~H eh.~ m mum. moH. mad. mud. mum. mam. Ned. u was: mo. so. so. No. «a. no. mo. mm Hosea mac.l Hmo.l «mad. I Avo.l who. Noe. Hoonom aomo. ovo. «mma.u moo.| coo. mmo. mmo. muomm «soo. moo. u oso.- omo. omo. moo. assess send. moo.| omo. tuna. noon. tvvm. mmo. >9 mumm >umso ocH Hmsvm \Mmmmowww Humz spud mnmsfimumm moanmwum> musoflowmmoou sowmmmummm Houowpoum oonmouoosmom oom noaomsus> soasouauu mfizmwd ZOHaflqudHoom MHMBO QZfl ZOHmH>mAmB mamomm wa flamda "ZOHadbom ZOHmmmmwmm 118 In controlling for other institutional socializers in this manner, it was felt that the most comprehensive assessment of their influence could be derived. That is, it was felt that the institution which was active in sports interest, conservative politically, and thus con- sistent with the values hypothesized to be inherent in sports and sports television and, at the same time, interested in politics and thus more likely to discuss the subject and otherwise exert influence with an indi- vidual, would be more likely to be influential than one less involved along one or more of the dimensions. When examining the coefficients, however, it can be noted that sports television is a better predictor of values than are the other institutions of socialization. With each coefficient representing the independent impact of the predictor variable in question, controlling for each other variable, the independent effect of each variable can be assessed. As can be seen in the table, the coefficients are significantly higher for sports television than for institutional influences. Therefore, the relationships are stronger. It can then be said that exposure to sports television is more strongly related to these political values, when controlling for institutional factors, than those institutional factors are when con- trolling for the impact of television. 119 The data support the hypothesis. Even more interesting, however, is the fact that sports television plays such a relatively large role in the process, usually serving as the best predictor of some of the values under study. The relationship is still significant when controlling for other socializing institutions, and these institutions seem to have less impact than does television in this area, failing to disprove the hypothe- sis. Hypothesis 13: These values will be associated with higher levels of exposure to televised sports when controlling for the attribute variables of sex, socioeconomic status, and race. Demographic factors, and their possible relation to sports television viewing and political values, were examined in this test given the relationship between socioeconomic status and sex with sports television and similar correlations with political values and demographic characteristics (Table 15). Controlling for these factors failed to alter the established trends in the relationships, however, as the following third-order coefficients indicate: 120 Sports television with sportsmanship .100, p = .011 ' authoritarianism .301, p = .001 nationalism .369, p = .001 need determined expression/ value restraint .112, p = .005 equalitarianism .059, p = .086 (N.S.) individualism -.047, p = .141 (N.S.) conservatism .247, p = .001 TABLE 15 CORRELATIONS BETWEEN DEPENDENT VARIABLES AND DEMOGRAPHIC FACTORS Sex S.E.S. Race Age Sportsmanship -.032 .187: .131* .066: Authoritarianism .033 -.128* -.062 -.156* Nationalism .143 -.104 -.006 -.105 Need Determined Expression/ * * Value Restraint .054 -.090 -.041 -.068 Equalitarianism -.041* .010* -.019* -.040* Individualism -.065 -.092 -.071 .065 Conservatism .027 -.030 .013 -.026 * p 5 .05 Results are consistent with those obtained with other controlling factors with exposure to sports tele- vision significantly related to higher levels of sports- manship, authoritarianism, nationalism, need determined expression, and conservatism, and no significant relation- ship between such exposure and individualism or equali- tarianism. In addition, as in other such equations, the variable of sports television accounts consistently for 121 oo. w o. Hoo. v A.o.zoooo. A.m.zo~oo. oso. Hoo. v soo. soo. v a Hmo .o Hmo .o Hmo .o Hmo .o Hno .o Hmo .o Moo .o up oo.o o~.~ oh. om.m oo.o~ oH.oH oo.m m oom. oma. oso. mod. «no. omm. omu. u pass no. mo. Ho. mo. oH. Ha. no. as House «Ho. oso.- aao.u mmo.a soo.: ooo.u «ooa. moms soo.: oso.: moo.u ono. ooo. owo.n oso.- mom moo.u oso.: moo. «ooH.u «Noa.u aooa.u «boa. .m.m.m oom. oso.: moo. oaa. ohm. mom. Hod. >9 mono a t a a a poem Ho> >noso osH Assam Huoz spam osmosouom \mxm omoz moanownm> musofioammmou sowmmmummm Houowcmum oonoouoosoum mom ooaooouo> soonououo mmOBU¢m UHmmmAmB mamomm mH fiflmdfi «ZOHBflDOm ZOHmmmmwmm 122 more of the significant explained variance than the others, and demonstrates, on the average, the strongest relationship when controlling for each variable. 0f the three demographic control variables entered, only one, Socioeconomic status, displays a con- sistently strong relationship and acts consistently as a significant predictor of political value levels. Sex and Race are not significant predictors except for the role of race with regard to sportsmanship, with the coefficient indicating higher sportsmanship levels among nonwhites than whites. With regard to sportsmanship, S.E.S. is a stronger predictor than the others with higher levels of sportsmanship associated with higher socioeconomic status to a limited, but significant, extent. Higher socio- economic status is also a significant predictor of several other political values, including authoritar- ianism, nationalism, and need determined expression. In this relationship, lower levels of S.E.S. are associated with higher levels of authoritarianism, nationalism, and need determined expression and lower levels of value restraint. gypothesis 14: These values will be associated with higher levels of exposure to televised sports when controlling for age. 123 Given the changes in television viewing, as described by previously cited audience studies, and the changes in levels of political socialization, described in previously cited socialization studies, and the fact that the sample was skewed toward younger respondents, age was also considered as a possible confounding variable. When entering age as a control factor, how- ever, overall trends are not changed, as the following coefficients reveal: Sports television with sportsmanship .118, p = .003 authoritarianism .264, p = .001 nationalism. .363, p = .001 need determined expression/ value restraint .101, p = .101 equalitarianism .043, p = .155 (N.S.) individualism. -.062, p = .072 (N.S.) conservatism .240, p = .001 Once again, sports television is significantly correlated with higher levels of sportsmanship, authori- tarianism, nationalism, need determined expression, and conservatism, despite the utilization of age as a control factor. Thus, qualified support for the overall hypothesis again exists. Hypothesis 15: These values will be associated with higher levels of exposure to televised sports when controlling for geographical area. 124 v a soo. v A.o.zooaa. i.o.zvooo. omo. Hoo. v Hoo. v ooo. o mmm .N on .N on .m omm .N mum .N mum .N wmm .N no o~.ma ¢H.N mm. mm.m ma.m¢ ww.w~ om.o b How. omo. mmo. NNH. ohm. eon. mma. H van: mo. Ho. Ho. v Ho. «a. mo. No. mm Hmuoa soo. ooo. soo.. «soo.- ooo.- .oHH.- moo. mom How. Nmo.l owe. «ca. own. new. mad. >9 muom « a a s s >Hmso osH Hmoom \Mmmmowwm Humz sus¢ msmsfimumm moanmfium> musoflowmmooo sawmmoumom nouowooum powwoumosmum was mmHQMHHo> coauwuwuo NEG QZd ZOHmH>mAWB mamomm ”ZOHB¢DOH ZOHmmmmwmm ha MAM9 mumm « t s s a poem Hm> >Hmcu osH Anson Husz nusd manssmumm \oxm vooz moanoflum> nouowooum musowowmmoou cowmmoumom oonoonoooouo mom noaoooum> soououono 20H8400A 02¢ ZOHmH>mAmB mfimomm ma mqmfla “ZOHB¢DOM ZOHmmmmUmm 127 relationships previously evident between sports television and political values. Therefore, the hypothesis is sup- ported (see Table 19). Hypothesis l6: These values will be associated with higher levels of exposure to televised sports when controlling for all of the previously mentioned possible confounding variables. Testing of this complex hypothesis will provide the best overall analysis of the particular role of tele- vised sports in the socialization process with regard to the political values under study. Also, it will approxi- mate the construction of a socialization model enabling one to predict levels of these particular values, given information on degree of exposure to and involvement in television sports and other significant socialization factors. In testing this particular hypothesis and develop- ing the socialization models, step-wise multiple regression procedure, as utilized by the Statistical Package for the Social Sciences computer program system, was applied for reasons as discussed in the previous chapter. In the analysis, regression procedure utilizing each of the previously tested variables and the television sports variable was performed on each of the dependent political value variables. The results appear in Table 20. 1J28 TABLE 19 PARTIAL CORRELATION SUMMARY TABLE Correlations between Sports Television and Dependent Variables when Controlling for: gigg; Participation Spectator Gagsral Sportsmanship .108 .048 .043 .105 Authoritarianism .281 .239 .228 .227 Nationalism .372 .354 .323 .308 Need Determined Expression/ Value Restraint .110 .117 .119 .091 Equalitarianism .048 .020 .072 .040 Individualism .047 -.049 -.082 -.074 Conservatism .241 .204 .156 .194 Type Family, Sch., Inst. Sport Peers Conservatism Sportsmanship .108 .053 .121 Authoritarianism .280 .221 .271 Nationalism .372 .314 .360 Need Determined Expression/ Value Restraint .109 .122 .099 Equalitarianism .048 .077 .062 Individualism -.071 -.061 -.074 Conservatism .241 .160 .234 32:2.ngiY:' Segécgzs, Location Age All Pol Int. Sportsmanship .032 .100 .102 .118 .016 Authoritarianism .244 .301 .291 .264 .190 Nationalism .378 .369 .383 .363 .302 Need Determined Expression/ Value Restraint .117 .112. .118 .101 .121 Equalitarianism .061 .059 .052 .043 .073 Individualism -.093 -.047 -.069 -.062 -.106 Conservatism .179 .247 .243 .240 .099 .nuouosouoo mmmm mooooxo oaamwuo> AIL .mo. I o 129 v # soo. v i.o.zooma. soo. ooo. Hoo. v ooo. v Hoo. v a mom .oH «so .oa ~Ho .oH «so .oH «Ho .oH Nam .oH «so .oH mo oo.m no.9 . oo.~ o~.~ oo.o om.o 9~.o m son. «as. oou. sou. ooo. soo. omm. a pas: oH. oo. so. so. am. no. 99. mm Houo9 ooo. soo.- ao3. ooo. «ooo. sows. I swam oooo oNo. soo. ooo.- ooo.- m~o.- .ooH.n «no. son ooo. ooo.- -o.- ooo.- Nao.n soo.: «soo. oosm suo.u oso.- ooo.- ooo. soo. ooo.- ooo.- xoo ooo.- ooo.- . ooo.- asoo.. .om9.u «NNH. .m.m.o ooo.- .ooo.u a~m~.u ooo.- a~9.1 sso.: o-. Hoorom «ooo. ooo. «ooo.- «om9.u s99o.- «ooo.- ”ooo. osooa «99o. sooH. «ooo. Hoo.u soo.- ao~9.u «Hom. >Hosoo ooo. . oso.- .ooa. «ova. amam. .mou.u uoH pom oosH ooo. smma.u ooo.- ooo. sons. .mHN. ano~.- .ooH Hos onsH soo. oso.: soo~.u aoma. «mos. soda. aoa~.u >unso oosH ooo. ooo. ooo.: «~o9.u soo.- ooo.- soo. unooo o>9 «oso. omo. oso.- ooo.- ooo. n «no. musoom oooo «so. u soo. oso.- soo.: ooo. ooo. nuuoom sumo soda. ooo. ooo. sso. «and. «mos. duo. >9 Honosoo ooo. ooo.- oso. ans. mom. ooo. oso. >9 ouuoom « a s s « >umsu osH Anson \wmmmowww Huoz nape mandamumm ooaooouo> uouowomum musowowmmmoo sowmnmumom nonaouoosmum pom moanmwum> sowuouwuo mHmNA¢Z€ ZOHmmmmwmm WAmHBADZ mamnmzoo om qudfi 130 As can be seen in the table, complex relationships among the variables are revealed. Many variables have an impact. Yet, several general tendencies emerge. Most importantly, in every case but one, the model developed is significant, re-enforcing the value of the tested variables as predictors of political values. Also, i2 five of seven instances, sports television explains sig- nificant amount of variance in those political values, even when controlling for all possible confoundipg effects which were tested. Moreover, this variable proved to be a relatively powerful one. In the five equations in which it was sigpificant, it was the best predictor in three and relatively a good predictor in the others. Its impact can be seen in Table 21. TABLE 21 VARIANCE ACCOUNTED FOR BY SPORTS TELEVISION To al Spor 8 Value Model RE TV RE % Authoritarianism .17 .08 .47 Nationalism .21 .14 .67 Need Determined Expression/ Value Restraint .07 .01 .14 Individualism .04 .01 .25 Conservatism .10 .06 .60 As a result, exposure to and involvement in sports television seems to be one of the strongest pre- dictors in the models in which it is significant. Other 131 traditional socialization agents, family especially, plus the peer group and school appear to be consistently sig- nificant in their predictive power as does socioeconomic status, although to a lesser extent. Surprisingly limited in its significant predictability, given its attention in previous research, is the variable of direct partici- pation in athletic competition. When all other related factors are controlled, it is not a significant predictor in any of the models. Also interesting is the complexity of the indi- vidual values and their determinants which make up the overall value system. For instance, sports television is an important variable in the areas of authoritarianism and nationalism but relatively unimportant in the areas of equalitarianism and individualism. The data tend to provide qualified support to the hypothesis. The role of television is even more evident when the two variables, sports television and general tele- vision viewing, are considered together. Table 22 indi- cates this to be the case. Further Analysis Not only did the role of sports television vary with the variable being tested but so did the overall strength and significance of the model. Therefore, in an attempt to resolve some of the ambiguity of the results, further testing was conducted. The consistent 132 significance of the model predicting levels of sports- manship, authoritarianism, nationalism, and need determined expression/value restraint, and the consistent lack of significance obtained with individualism and equalitar- ianism, tended to indicate a natural subgrouping of the variables. TABLE 22 VARIANCE ACCOUNTED FOR BY ALL TELEVISION VIEWING To al STV General Total Value Model R3 R2 TV R2 TV R2 % Authoritarianism .17 .08 .03 .ll 65 Nationalism .21 .14 .02 .16 76 Need Determined Expression/ Value Restraint .07 .01 - .01 14 Individualism .04 .01 - .01 25 Conservatism .10 .06 .01 .07 70 The former four variables can be considered, to some extent, as more system oriented. They deal with values of behaviors and attitudes which apply to normative concepts and relations with institutions and are thus more general. They deal less with specific individuals or people in general. Rather, they deal with approaches to behavior patterns, groups, and organizations. The latter two, however, deal with evaluations of people, especially on an individual level and with how one should evaluate, approach, and deal with other people. This focus on evaluations of people, as opposed to more 133 general and abstract evaluations of normative and insti- tutional relationships, separates these two values from the first four. The valuing of "individualism“ or "equalitarianism," in terms of other people, is thus different than evaluating norms such as "sportsmanship," "traditional values," or evaluating institutional inter- actions with "authority" and "the nation.” Such a breakdown would be consistent with Rokeach's (1968) concept of central vs. peripheral values and beliefs in which individualism and equalitarianism could be viewed as being more central since they are directly experienced by the individual and more consensual within the individual's cultural environment. Peripheral values are more remote, less consensual, and less directly exper- ienced, being derived indirectly, and thus being more susceptible to change. Such values as individualism and equalitarianism, calling as they do for evaluations of other people, would be directly experienced more often than the others. Also, they seem to be more consensual than others, being accepted as part of the American political culture. The other values are more peripheral. One does not become centrally involved in the concept of "sports- manship," for instance, unless one participates in sports, and the evaluation of traditional loyalties to norms and values vs. a situational ethics approach is also less 134 consensual. Still more peripheral is a value such as "authoritarianism" since definitions of and specific interactions with "authority" as personified or repre- sented via institutionalized political power varies greatly among individuals. The most peripheral would be the concept of "nationalism," especially as operational- ized here, as it deals with the more symbolic and indirect components of this value. Most relationship with this value would be indirectly derived, rather than directly experienced, and its rationale probably least consensual. Therefore, for further analysis, the variables of individualism and equalitarianism.were considered to be "person-oriented" central values since they dealt with evaluations of individual status and behavior and were combined additively. The other variables can be con- sidered "system-oriented” peripheral values since they deal with evaluations of overall standards of conduct. These were also combined additively. The development of a model for these two concepts of "person-oriented” and "system-oriented” values produced the following results (Table 23) . By treating only "system-oriented” values in the conceptualization of the value system, a sounder model is constructed than that which dealt with conservatism in general, accounting for almost a quarter of the variance. Sports television, by itself, accounts for 135 TABLE 23 MULTIPLE REGRESSION ANALYSIS FOR PERSON AND SYSTEM ORIENTATIONS Criterion Variables and Standardized Regression Predictors CoeffiCients Person- System- Orienteda Orientedb Sports Television -.044 .273* General Television -.036 .100* Participatory Sports .033 -.046 Spectator Sports -.016 -.028 Type of Sport .012 -.085* Institutional Conservatism -.115* .302* Institutional Political Interest .127* .252* Institutional Sports Interest -.032 .360* Family .130* -.220* Peers -.047* -.268* * * School .113 -.207 * * S.E.S. -0065 -0177 Sex -.040 .036 * * Race -.054 -.078 Age .026 -.109* * Geographical Area .053* .132 * 5 .05 aaz = .04, mult r = .191, F = 1.21, df = 16, 512, p = .254 (N.S.). bR2 = .24, mult r = .488, F = 10.02, df = 16, 512, p < .001. 136 over one-third of the explained variance. The consistency of the previous models is again evident with traditional socialization agents contributing to relatively large amounts of significant variance, and other factors, such as participatory sports, contributing to little if any of the significant explained variance. An additional additive dimension was constructed given the relative differences between the predictive power of the models for authoritarianism and nationalism, which were strong, especially as far as sports television was concerned and sportsmanship and need determined expression/value restraint, which were relatively weak and at times ambiguous. Once again, however, these factors could be developed into natural groups given that the latter two represent relatively abstract, socially oriented concepts, whereas the former two rep- resent more concrete concepts related to more tangible institutions such as national loyalty and respect for and obedience of authority, which represent more of an organizational approach to political value. Again, Rokeach's central-peripheral continuum can be applied with values relating to institutions such as "authority" and “the nation" considered more peripheral than more generalized normative values relating to social concepts such as "sportsmanship" and I'traditional values“ which would be relatively central. 137 Therefore, for further analysis, the variables of sportsmanship and need determined expression/value restraint were considered to be "social" values, since they dealt with general normative factors with regard to the individual and those around him. The variables of authoritarianism and nationalism, then, were considered "institutional" values since they dealt with the indi- vidual's relationship with more tangible and specific social organizations such as "authority” and the "nation." Here again, the predictive power of televised sports, with respect to particular political values, especially those dealing with institutional, systems- oriented support such as authoritarianism and nationalism, is evident. Consistent with other similar equations, sports television accounts for the most explained variance (15 percent out of 25 percent, or 60 percent of the explained variance). The role of television in the socialization model is especially powerful when sports television and other types of television content are combined. In this case, while sports television is the most important variable, general television ranks third, the two together accounting for 17 percent of the total of 25 percent explained variance. The figures are similar when considering equations for authoritarianism and nationalism separately. 77 pp- 138 TABLE 24 MULTIPLE REGRESSION ANALYSIS FOR SOCIAL AND INSTITUTIONAL ORIENTATION Standardized Regression Coefficients and Cri- terion Variables Predictors Social Value Institutional Supporta Supportb Sports Television .070 .291: General Television -.006 .141 Participatory Sports -.057 -.010 Spectator Sports -.052* .008 Type of Sport -.091* -.036* Institutional Conservatism .280 .160 Institutional Political * * Interest .139 .203 Institutional Sports * * Interest .256* .253* Family -.219* -.099* Peers -.244* -.l4l* School -.l74* -.ll8* S.E.S. -.125 -.131 Sex .063* - Race -.084 -.036 Age -.074 -.080* Geographical Area .051 .132* * p 5 .05, (-) exceeds SPSS parameters. 2 an = .11, mult r = .327, F = 3.83, df = 16, 512, p < .001. sz = .25, mult r = .495, F = 11.04, df = 15, 512, p < .001. TABLE 25 VARIANCE ACCOUNTED FOR BY TELEVISION VIEWING To al STV Gen. Tota Value Model R5 R2 TV R2 TV R % System-Oriented .24 .08 .02 .10 42 Inst. Support .24 .15 .02 .17 71 Soc. Value Support .11 - - - - 139 Hmo.| mod. who. omo. omo. hwo. mod. Hmo.| moo.l oso.: cousowHOIcoouom moo. won. woo. ooo.: ovo. hmo. mmo. How. moo.| woodm> Hus09usuwumsa mmo. oam.| hma. who. own. mma. ooo.: mmo.u mooam> Howoom mom. Hoo. HHo. on. mas. one. mmm.n oousofiuousounhm mow. oam.1 Hom.| mom. omo. Nov. Euwuu>uonsoo ooo. ova. moo. oHo.I OHH.I Emwamsvw>flosH Hma. ooo.: mao. nHo. Emwsowuouwamsom oma. ooo. m-.u vowsuunom osau> \cowmmoumxm oosweuouoo oomz vmv. ooo.: emwflusoflumz mno. Endsm«umuwuonus¢ owsnsoamuuoom p.89 35 com summm Coco o5 Home m>xmoz 3oz 59¢ oomsnomm mmqmdfimdxw ZOHmMBHmU\Bzmn—zmmma A42?“ .mO van—”mad: ZOH9