TESTING AN INTEREST MODEL OF THE KNOWLEDGE GAP PHENOMENON " ‘ Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY B. K. L. GENOVA 1975 "Er-J"! This is to certify that the thesis entitled TESTING AN INTEREST MODEL OF THE KNOWLEDGE GAP PHENOMENON presented by B. K. L. Genova has been accepted towards fulfillment of the requirements for Ph . D 0 degree in Communication Micl'zigan State University Date June 2L 1975 0-7639 ABSTRACT TESTING AN INTEREST MODEL OF THE KNOWLEDGE GAP PHENOMENON BY B. K. L. Genova The purpose of this dissertation was to examine the notion of differential levels of information acquisition from the mass media in the light of an interest based model. Interest was viewed in terms of perceived information utility to self and to milieu. To the extent that an infor- mation item is seen as having such utility, resulting inter- est determines the kind of attention an individual will give to that information item. A panel survey was used to ex- amine the respondents' knowledge about two events, a foot- ball strike and presidential impeachment deve10pments in the summer of 1974, in terms of their interest with respect to these events. The study's findings indicate that the more interested segments of the audience indeed picked up infor- mation faster and also at any point in time knew more than those less interested in the same event. Furthermore, per- ceived information utility to one's social milieu emerged as the most important component of interest in explaining knowl- edge differences. The study raises some new questions B. K. L. Genova related to the understanding of the knowledge gap phenome- non and suggests some promising routes for further investi- gation. TESTING AN INTEREST MODEL OF THE KNOWLEDGE GAP PHENOMENON A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1975 Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the require- ments for the Doctor of Philosophy degree. 6e COMMIT“ Dinfictfik of Twasis (In II” V‘WLLLEALX , Chairman 76.4.41... Jsé4/ 2/4424“...- jug/mfg a? 94/ @ng @dc erg/MM. F7 Guidance Committee: V) ii ACKNOWLEDGMENTS Much gratitude is due to Dr. Bradley S. Greenberg, my firm yet empathic advisor, who influenced me strongly and did not seem to know it; the members of my committee -- Dr. Randall Harrison and Dr. Gerald Miller, who kept me on my toes while providing valuable support long before this project was started; Dr. Ruth Useem, and her good humored stance on quantification; my friends Muriel Rokeach and Louise McCagg, for their continuous help and patience with my ravings; and Dee Greenberg, for her kindness at strategic points in time. Many thanks go also to my appre- ciated colleague at Syracuse University -- Pauline Atherton who gave crucial emotional support; Wayne Crouch, a very special professional companion; Marta Dosa, my link with the third dimension; Jeffrey Katzer for stimulating and useful discussions; and our fearless leader Bob Taylor, simply for his trust in me. Special thanks also go to Bette Brindle, my crackerjack computer programmer who made life easier; and Mrs. Ruth Langenbacher, who kindly handled the details and production of the final manuscript copy. Finally, to all friends and some foes, for being that all-important ambience —- SALVETE! iii TABLE OF CONTENTS Chapter I RATIONALE AND HYPOTHESES . . . . . INTRODUCTION . . . . . . . Statement of Problem and an Overview . Patterns of Media Use . . . . Role of the Different Media . . . Exposure and Information Gain . . Antecedent Factors Systematized . . An Interest Model . . . . . . Tracing Interest Differentiation . The Role of the Family . . . . The Social Milieu . . . . . Adult Audiences and the Information Environment . . . . . . A SUMMARY . . . . . . . . HYPOTHESES . . . . . . . . II METHODOLOGY . . . . . . . . Questionnaire Development . . . . Sampling . . . . . . . . Questionnaire Administration . . . MEASUREMENT OF VARIABLES . . . . National Football Strike - Measurement of Variables NFL, TIME ONE NFL, TIME ONE NFL, TIME TWO FACTUAL KNOWLEDGE . . STRUCTURAL KNOWLEDGE FACTUAL KNOWLEDGE . . NFL, TIME TWO - STRUCTURAL KNOWLEDGE . Measurement of Variables - Impeachment Impeachment - Dependent Variables . . IMPEACHMENT, TIME ONE - FACTUAL KNOWLEDGE . . . . . . . . IMPEACHMENT, TIME ONE - STRUCTURAL KNOWLEDGE . . . . . IMPEACHMENT, TIME TWO - FACTUAL KNOWLEDGE . . . . . . IMPEACHMENT, TIME TWO - STRUCTURAL KNOWLEDGE . . . . . . . . iv Page H 39 39 41 42 44 45 SS 56 57 58 59 69 69 7O 70 71 Chapter II (cont'd.) Chapter Page STATISTICAL ANALYSIS . . . . . . 74 III TEST RESULTS AND INTERPRETATION . . . . 75 Main Hypotheses . . . . . . . 75 Specific Hypotheses . . . . . . 82 General Hypotheses . . . . . . 112 IV SUMMARY AND DISCUSSION . . . . . . 117 Discussion . . . . . . . . 124 APPENDICES . . . . . . . . . . . 139 A. Sample Characteristics . . . . . 139 B. General Measurement Procedures . . . 146 C. Instrument . . . . . . . . 167 BIBLIOGRAPHY . . . . . . . . . . . 180 Table 1. 2A. 2B. 3A. 3B. 4A. 4B. 5A. LIST OF TABLES Overview of Antecedents . . . . . . NFL, TIME ONE SELF INTEREST - Factor Matrix of the Three Measures of Self Interest . . Pearson Product-Moment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index of Self Interest . . . . . . NFL,TIME ONE MILIEU INTEREST - Factor Matrix of the Five Measures of Milieu Interest . . Pearson Product-Moment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index 0 O O O I O O O O O NFL,TIME ONE POTENTIAL SELF INTEREST - Factor Matrix of the Three Measures of Potential Self Interest . . . . . . . . Pearson Product-MDment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . . NFL, TIME ONE POTENTIAL MILIEU INTEREST - Factor Matrix of the Five Measures of Potential Milieu Interest . . . . . Pearson Product-Moment Correlation among the Component Measures . . . . . . vi Page 15 49 49 49 50 50 50 51 51 51 52 52 Tables 5B. 6A. 68. 7A. 7B. 10. 11. 12. 12A. 12B. 13. Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index 0 O C O O I O O O 0 NFL, TIME TWO SELF INTEREST - Factor Matrix for the Three Measures of Self Interest . Pearson Product-Moment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . . NFL, TIME TWO MILIEU INTEREST - Factor Matrix for the Five Measures of Milieu Interest . Pearson Product-Moment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . . Means and Standard Deviations of Three Measures of Factual Knowledge . . . . Means and Standard Deviations of Three Measures of Structural Knowledge . . . Means and Standard Deviations for Three Measures of Factual Knowledge . . . . Means and Standard Deviations for Three Measures of Structural Knowledge . . . IMPEACHMENT, TIME ONE SELF INTEREST - Factor Matrix for the Four Measures of Self Interest . . . . . . . . . Pearson Product-Moment Correlations among the Component Measures . . . . . . Variable Communalities, Factor Purities, and Factor Score Coefficients for the Overall Index 0 O O O O O I O I O IMPEACHMENT, TIME ONE MILIEU INTEREST - Factor Matrix for the Four Measures of Milieu Interest . . . . . . . . . vii Page 52 53 53 53 54 54 54 55 56 57 58 63 63 63 64 Tables 13A. 13B. 14. 14A. 14B. 15. 15A. 158. 16. 16A. 16B. 17. 17A. 17B. Pearson Product-Moment Correlations among the Component Measures . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . IMPEACHMENT, TIME ONE POTENTIAL SELF INTEREST - Factor Matrix for the Four Measures of a Potential Self Interest . Pearson Product-Moment Correlations among the Component Measures . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . IMPEACHMENT, TIME ONE POTENTIAL MILIEU INTEREST - Factor Matrix for the Four Measures of Potential Milieu Interest . Pearson Product-Moment Correlations among the Component Measures . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . IMPEACHMENT, TIME TWO SELF INTEREST - Factor Matrix for the Four Measures of Self Interest . . . . . . . Pearson Product-Moment Correlations among the Component Measures . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . IMPEACHMENT, TIME TWO MILIEU INTEREST - Factor Matrix for the Four Measures of Milieu Interest . . . . . . . Pearson Product-Moment Correlations among the Component Measures . . . . . Variable Communalities, Factor Purities and Factor Score Coefficients for the Overall Index . . . . . . . . . viii Page 64 64 65 65 65 66 66 66 67 67 67 68 68 68 Table 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Means and Standard Deviations of Three Measures of Factual Knowledge . . . Means and Standard Deviations of Three Measures of Structural Knowledge . . Means and Standard Deviations for Three Measures of Factual Knowledge . . . Means and Standard Deviations for Three Measures of Structural Knowledge . . NFL, Impeachment: Interest Component Intercorrelations . . . . . . NFL, Impeachment: Interest Component Intercorrelations . . . . . . Interest-Education Intercorrelations . Pearson Product-Moment Correlations Between Education and Knowledge . . Pearson Product-Moment Correlations Between Overall Interest and Knowledge . Multiple Regression, Education, Interest and Knowledge - Time One, NFL . . . Multiple Regression, Education, Interest and Knowledge - Time One, Impeachment . Multiple Regression, Education, Interest and Knowledge - Time Two, NFL . . . Multiple Regression, Education, Interest and Knowledge - Time Two, Impeachment . Correlational Data for both Events at Time One and Two: Education, Interest and Knowledge . . . . . Pearson Correlation Coefficients for both Events at Time One: Education, Factual and Structural Knowledge . . . . Pearson Correlation Coefficients for both Events at Time Two: Education, Factual and Structural Knowledge . . . . ix Page 69 70 71 71 72 72 73 76 76 77 78 79 80 81 83 83 Table Page 33A. Pearson Correlation Coefficients for both Events at Time Two: Education, Factual and Structural Knowledge . . . . . 84 34. Correlational Pattern for Education and Interest With Factual and Structural Knowledge . . . . . . . . . 84 35. NFL Factual Knowledge; Multiple Regression of Interest Components - Time One . . . 87 36. NFL Structural Knowledge; Multiple Re- gression of Interest Components - Time One . . . . . . . . . 88 37. Impeachment Factual Knowledge; Multiple Regression of Interest Components - Time One . . . . . . . . . 89 38. Impeachment Structural Knowledge; Multiple Regression of Interest Components - Time One . . . . . . . . . 9O 39. NFL Factual Knowledge; Multiple Regression of Interest Components - Time Two . . . 91 40. NFL Structural Knowledge; Multiple Re- gression of Interest Components - Time Two 0 O O O O O O O O 9 2 41. Impeachment Factual Knowledge; Multiple Regression of Interest Components - Time Two . . . . . . . . . 93 42. Impeachment Structural Knowledge; Multiple Regression of Interest Components - Time Two 0 O O O O O O O O 9 4 43. NFL and Impeachment; Multiple Regression of Interest Components - Time One . . . 95 44. NFL and Impeachment; Multiple Regression of Interest Components - Time Two . . . 96 45. NFL and Impeachment; Education and Self Interest as Predictors of Factual Knowledge - Time One . . . . . . 98 46. NFL and Impeachment; Education and Self Interest as Predictors of Structural Knowledge - Time One . . . . . . 99 X Table 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. NFL and Impeachment; Education and Self Interest as Predictors of Factual Knowledge - Time Two . . . . . NFL and Impeachment; Education and Self Interest as Predictors of Structural Knowledge - Time Two . . . . . NFL and Impeachment; Education and Milieu Interest as Predictors of Factual Knowledge - Time One . . . . . NFL and Impeachment; Education and Milieu Interest as Predictors of Structural Knowledge - Time One . . . . . NFL and Impeachment; Education and Milieu Interest as Predictors of Factual Knowledge - Time Two . . . . . NFL and Impeachment; Education and Milieu Interest as Predictors of Structural Knowledge - Time Two . . . . . NFL; Education and Potential Self Interest as Predictors of Factual and Structural Knowledge - Time One . . . . . Impeachment; Education and Potential Self Interest as Predictors of Factual and Structural Knowledge - Time One . . NFL; Education and Potential Milieu Interest as Predictors of Factual and Structural Knowledge - Time One . . . . . Impeachment; Education and Potential Milieu Interest as Predictors of Factual and Structural Knowledge - Time One . . Comparison on "New Knowledge" at Time Two, for both Events . . . . . . t-test Comparisons of Mean Knowledge for both Events . . . . . . Comparisons of Discrepancy in Knowledge on NFL and Impeachment . . . . x1 Page 100 101 103 104 105 106 108 109 110 111 113 114 115 Table 59A. 59B. 60. 61. 62. 63. 64. 65-65C. 66-66C. 67-67C. 68-68C. 69-69C. 70-70C. Partial Correlation Check for Interest, Knowledge and Education . . . Demographic Profile of the Sample . Use of Separate Media . . . . Overall Mass Media Use Profile . Bartlett's Test Comparisons of the Sample and the Uncooperative Group Comparison of Sample Demographic Char- acteristics and 1970 Census Data . Census Data for Occupational Classifi- cations for Lansing SMSA (breakdown for Lansing Area + Urban Balance unavailable) . . . . . . Varimax Configuration Check, NFL, Time One, Self/Milieu Interest . . Varimax Configuration Check, NFL, Time One, Potential Self/Milieu Interest Varimax Configuration Check, NFL, Time Two, Self/Milieu Interest . . Varimax Configuration Check, Impeach- ment, Time One, Self/Milieu Interest Varimax Configuration Check, Impeach- ment, Time One, Potential Self/ Milieu Interest . . . . . Varimax Configuration Check, Impeach- ment, Time Two, Self/Milieu Interest xii Page 128 140 141 141 142 143 144 152-153 154-155 156-157 161-162 163-164 165-166 Figure 1. 8a. 8b. 8c. 8d. 10. 11. 12. 13. 14. LIST OF FIGURES Three Media Use Models . . . . . Salience and Knowledge Relevance . . Perceived Utility . . . . . . Information Gain Response Functions . . Juxtaposition of Interest and Education . Diagram of Hypothesized Relationships . Summary of Interest Components as Pre- dictors of Structural, Factual and Overall Knowledge . . . . . . Diagram of Hypothesized Relationships . Knowledge Comparisons Over Time, National Football League Strike . . . . . Knowledge Comparisons Over Time, Impeachment Developments . . . . Cross-lagged Analysis of the Causal Flow Revised Diagram of Hypothesized Relationships . . . . . . . Vector Representation . . . . . Vector Representation . . . . . Vector Representation . . . . . Vector Representation . . . . . Vector Representation . . . . . Vector Representation . . . . . xiii Page l3 17 18 22 35 38 86 120 123 123 131 136 152 154 156 160 162 164 CHAPTER I RATIONALE AND HYPOTHESES INTRODUCTION European analysts have found that the average reader retains about 10 percent of the news.... This is probably for- tunate; if he remembered it all, he would go mad. Jacques Ellul, "With a View Toward Assessing the 'Facts.'" N.Y. Times, July 1, 1973. Ellul exaggerated, in the best tradition of specula- tive fecundity, of course. Furthermore, the Eur0pean ana- lysts are not unique for noticing that the audience, be it readers or viewers, often neglects mass media attempts to in- form them on everything at all times. Yet the issue touched upon is not unrelated to the purpose of this dissertation. That is, an exploration of the knowledge gap notion advanced recently by Tichenor et 31. (1970). It has focused attention on a number of formerly disparate findings, which taken to- gether seem to show that various social strata do not share the same propensity to partake in the "information explosion. Statement of Problem and an Overview Recently the Minnesota group put forward the following "knowledge gap" hypothesis: 2 As the infusion of mass media information into a social system increases, segments of the popula- tion with higher socio-economic status tend to acquire this information at a faster rate than lower status segments, so that the gap in knowl- edge between them tends to increase rather than decrease. The authors examined and found partial support in four sets of data from studies originally carried out for entirely different purposes: (a) a 48-hour news diffusion study on political affairs of the day, (b) a study of two communities, one of which suffered a newspaper strike over two weeks, (c) a field study where certain topics received heavy media coverage, while others did not, and (d) time trend data from national surveys dealing with science-related topics. Information infused in the system was Operationalized as highly publicized topics, and stratification was based on education alone. Thus, in terms of the broadly formulated hypothesis, it is expected that knowledge of heavily pub- licized topics will be accrued faster over time by the better educated persons than by those with less education. Proceeding from the general finding that knowledge of public affairs and science issues strongly correlated with education (Wade and Schramm, 1969), Tichenor et 31., utilized education as the key variable in their study. All the while they suggested the potential role of several contributing factors which may reinforce knowledge gap differentiations: (a) communication skills - the better educated are better equipped for reading and comprehending new information. 3 (b) stored information - those already better in- :flarmed are more likely to be aware and responsive to news tepics appearing in the media. (c) relevant social contacts — education increases the "life space," and indicates more reference groups and interpersonal contacts, increasing the chance of discussing public affairs. (d) selective exposure and retention — voluntary ex- posure to the news is often enhanced with education. One would expect the knowledge gap to be most noticeable whenever one or more of these factors is at work. In short, education was taken as an indicator of the trained capacities (or incapacities) of the audience members to respond to in- coming mass media information. Such affirmation of education as an important variable in the study of audience information levels rests on a number of consistent findings. Berelson e; 21. (1954) argues that the better educated have developed cognitive skills which allow them to retain better the information in whatever they see, read or hear. Robinson (1967) studies awareness of political issues in the Far East (e.g., China, Vietnam) among six groups broken down by education, and income level. There were large differences in information scores and his conclu- sion was that education is the most important determinant of information level. Education has been found to increase over- all media use, as in Key's study of campaign news (1961); he found differences between levels of information for people at educational extremes and attributed these to the tendency of the better educated to display higher rates of exposure. At this point it seems appropriate to mention also some of the knowledge gap-like results reported on the Sesame Street program for children during the first and second years of implementation (Ball and Bogatz, 1970, 1971). On the whole, the first year report found that disadvantaged children, as a subgroup compared to the total sample, viewed less, advantaged children watched more. Furthermore, some rough comparisons between the family environment of these children could be made from the available data; it can be seen that some differences existed in the family environment and particularly in parent-child interaction patterns between the advantaged and disadvantaged: (a) the advantaged children who viewed most had mothers talking often about the show, and playing Sesame- based games with them. (b) parent expectation for the child's performance was higher among the advantaged. (c) five out of six parents for the advantaged used the parent-teacher guide provided with the Sesame program, compared to 1 out of 20 for the disadvantaged. (d) mothers of advantaged children read more to them (78% vs. 49%) in general, and also "once a day" (42% vs. 24%). 5 The report on the second year of Sesame Street added comparisons based on encouragement of children to view the program, which turned out to be a significant variable affecting the gains of the children, for all of the total tests. Also, as in the first year, disadvantaged frequent viewers gained as much as advantaged frequent viewers. On the other hand, advantaged infrequent viewers gained more than the infrequent viewers among the disadvantaged; thus, the advantaged children seem to have alternative avenues of gaining knowledge, such as reading and discussion at home, spare time activities conducive to learning, etc. Patterns of Media Use - The most emphasized aspect of the relation between education and mass media use has been the preference for print_media among the better educated. Lower income, less educated adults display less newspaper readership, magazine use (Block, 1970, Dervin and Greenberg, 1972), but watch television almost twice as much as the general population (Dervin and Greenberg, 1972). The breadth of readership is narrower than the general pOpulation (17% versus 39% reading "all" of the newspaper), reading less front page items, and more ads in general. The preference for human interest and confessional type content in magazines is strong. In the same vein, Wade and Schramm (1969) report data collected by the University of Michigan Survey Research Center in the course of four national surveys of political 6 information (issues and candidates) during presidential campaigns (1952, 1956, 1960 and 1964). Percentages repre- sent proportion of respondents making regular use of three mass media during each presidential campaign. The data from the last survey, (1964, n=1570) indicates the following trends: monotonic increase of newspaper reading with edu- cation (38% to 71%), income (39% to 57%), and with occupa- tion (41% to 64%); slight increase with age (47% to 51%), males to females (53% to 47%), whites more than blacks (51% to 43%); for magazines, all monotonic increases with educa- tion (12% to 59%), with income (10% to 35%), occupation (24% to 40%), whites more than blacks (25% to 13%), and no dif- ference by age or sex (24% to 24%). Thus, the familiar pattern describing the print readers as those generally found at a higher SES level clearly emerges. The picture changes with respect to television, where the differences are much less pronounced by education (68% to 72%), income (65% to 70%), occupation (65% to 70%) or sex (69% to 71%). Similarly, Parker and Paisley (1966) found that television remains a chief source of information during campaigns for practically everyone, but respondents from lower SES strata tended to watch television more. In terms of television content preferences, the dif- ferences are not very sharp, either. Apparently the educa— ted viewer is as likely to turn on entertainment programs and shun "heavy" informational content, as are the less 7 privileged (Robinson, 1972, Bower, 1973, Dervin and Greenberg, 1972). The observation that demographic variables have limited power in predicting the use of electronic media was made as early as 1968 by Greenberg and Kumata, on the basis of a national sample study. Similar findings have been re- ported more recently by Bower (1973) and Robinson (1972). In View of the fact that dealing with the mass media is only a part of daily human behavior, after all, factors other than simple demographics may tell us more about people's mass media preoccupations. Some authors have suggested altern- ative variables such as extent of social interaction (Rosengren, 1972, Greenberg and Kumata, 1968), cognitive needs for information (Atkin, 1974, McCombs and Weaver, 1973, Greenberg and Kumata, 1968), or available leisure time (Samuelson 33.31., 1963; Robinson, 1969; Nayman, et 21., 1973). Role of the Different Media - Given such differential tendencies with regard to print media HES! and the relative uniformity with regard to television use, it is important to know their contribution to the audience level of infor- mation. First of all, there is some evidence to suggest that the manner in which a medium is used has a lot to do with information level, apart from which particular medium is being used. Robinson (1967) compared several social groups by eight types of "SOphisticated usage" and under- 8 scored his finding that, aside from education differences, those who are better informed display a "more serious usage of mass media for informational content." Bagdikian (1971) noted that while surveys fail to indicate large scale pre- emption of one mass medium by another, it appears that the media are often used for different purposes. The manner of media usage also emerged as a factor in public affairs knowledge in a study of teen-agers during the 1968 campaign; Chaffee ep’gl., (1970) found that the youths' familiarity with the issues was significantly assoc- iated with their use of television and newspapers for public affairs content, in contrast to use for entertainment. In a similar vein, Schramm and Wade (1967) found that viewers who used television for "campaign purposes" knew more than habitual print media users. Print and broadcast media viewed separately play different roles regarding the information level of audiences. The printed media do indeed emerge as the strongest contribu- tor to political and public affairs information gain (Robinson, 1967, Schramm and Wade, 1967), and it is the better educated who make the most use of them (Robinson, 1970). At the same time, while print media use indeed deline- ates information discrepancies between educationally unequal segments, it is good to find some evidence that television use does not seem to augment such differences much further. 9 There is no evidence of correlation between network tele- vision news viewing and education (Robinson, 1972). Also, Johnson (1973) found that both heavy and light television viewers knew very little about peOple in the news. Similiar report comes from Stern (1971), who found half the audience of a national network unable to recall even one of 19 stories, shortly after they were broadcast. Exposure and Information Gain - Evidence that mere exposure to mass media does not necessarily bring about more news awareness abounds in the public opinion literature (Robinson, 1972). But the fact that the manner in which media are approached by the audience can take precedence over the print-broadcast differentiation as a factor in infor- mation gain strongly suggests that taking a receiver point of view can provide good leads to explanations of audience behavior. The reminder that mass media audiences are not passive receptacles for any and all messages and the informa— tion herein is indeed stating the obvious. There is one saving grace in doing so, however--recognizing that some formulations of what information is may have been unduly normative, by-passing the audience's own discriminative judgment. The point can be made by mentioning the classic Cincinnati project and the Douglas 25.31., (1970) study, with their contrasting outcomes. While Star and Hughes (1950) found that their campaign to educate peOple about the U.N. reached only the more educated and affluent residents 10 of Cincinnati, Douglas 33.31. managed to come across with their six month long information program on mental retarda- tion, particularly so among the lower education strata. The small experimental community Reedsburg (Wisconsin), with an established closely knit system of institutions, indigenous media and interpersonal contacts turned out to be a better setting for an information campaign, compared to the diversified, fragmented locale of Cincinnati. More importantly, the issue of mental retardation, while not con- troversial, had local action implications; that is, the com— munity was trying to establish better facilities for educable retardates, and already had some such facilities in existence. Therefore the campaign, which consisted of 20 news stories, 5 feature stories and a mental Retardation Week ad in the local paper, posters, radio spots, etc., fell on responsive ears since the t0pic was not remote from the reality of the residents. It turns out that the "know-nothings" may know little about politics, but know something about other areas, and ggp_be good information gainers. The question why this is so may be answered through a closer look at audience infor- mation needs and interests, and their role in information gain processes. In order to do so, let us first discuss antece- dent factors which emerge from previous studies in relation to exposure and information gain. Relevance and salience of information were emphasized as better predictors of whether an individual becomes 11 acquainted with a news story at all, as compared to news value (e.g., percentage aware), according to a recent study by Hanneman and Greenberg (1973). Indeed, a number of studies indicate that availability of information may en- hance exposure, but neither will necessarily enhance infor- mation again (Allen and Colfax, 1968; McLeod, 33 gl., 1969; Spitzer and Denzin, 1965; Greenberg, 1964b). It is impor- tant to distinguish the expected consequences from a message- orientation standpoint, from the likely consequences from a receiver-oriented standpoint. The issue of receiver atten- tion can be traced back to the notion of stimulus intensity, i.e., the number of discriminatory stimuli that impinge on a person (De Fleur, Rainboth, 1952). The idea was that the greater the number of messages about an event, the greater the probability of exposure to it, . . . the greater the per- centage of people informed. Not necessarily so. The receiver is likely to act upon a subjectively reckoned in- formation value in giving attention to the story. This implies that the general practice in research, of operation- alizing attention as exposure, needs some refinement with respect to attention variability regarding program components. Thus, interests may direct the information use of the mass media; or given exposure, presence or absence of interest may intervene with the kind of attention given certain content areas and program components. For example, Atkin ep'gl. (1973) studied audience reaction to political ads on TV 12 during two 1970 campaigns for governor and found that: (a) availability of political ads on television was unrelated to the attention given to them, with an average correlation of -.02; (b) interest in the campaign contributed consider- ably to attention patterns; (c) finally, demographic locators such as age, sex, education, or occupation were only slightly related to attention given the TV spots. Several studies have shown a positive relationship between relevance and knowledge (Adams, ep 31., 1969; Levy, 1969). Fitzsimmons, 32 al. (1969) for example, reported that importance of an issue and interest seem to go together and are related to knowledge of public affairs gained from television documentaries. Greenberg, ep'al. (1965) found interest positively and strongLycorrelated to amount of knowledge about a sports event. Funkhouser and McCombs (1971) found interest related to amount of recall of news items. Two recent studies, carried out by Johnson (1973) in Kentucky and Bishop (1973) in Peru, sought to isolate the factors playing a role in political knowledge acquisition. They show similiar findings even though the strength of rela- tionships reported by BishOp was higher than for Johnson. Both authors concur on the role played by interest predictors, suggesting that the model moves from interest, through media use to knowledge in the political realm. On the other hand, Atkin and Greenberg (1974) tentatively posited a different 13 model in their experimental study of a public television series' impact on the political socialization of adolescents in Florida. There, media use seems to be the key variable, while interest and knowledge follow as the consequences. Bishop (1973) Johnson (1973) Political Political .30 Political Knowledge Interest Information Information .6 Seeking .42 .28 Predictors .59 (Interest) .64 Media Use Media Use Atkin/Greenberg (1974) Media Interest Exposure<: Affe\Ct‘O. 24715 Knowledge Figure 1. Three Media Use Models Given the wide variety of events covered by the media, both topically and geographically, there is ample room for manifestation of interest diversities. Yet we know precious little about what the audience, and sub-groups within it, include in their routine "survellance of the environment" in Lasswell's sense. The often dismal picture of public ignorance obtained through some public Opinion studies is based on a narrowly 14 defined set of information which implies disregard for public heterogeneity. Antecedent Factors Systematized - An attempt to pre- sent systematically the manner in which antecedent factors have been studied shows that there is little agreement in the way of defining interest, salience, importance or rele- vance of events. What follows is a summary table and a brief overview of the various perspectives and terms avail- able from previous studies (Table 1). Adams and Mullen (1969) base their distinction on potential social utility in dividing news into "so what?"- type (i.e., minimal interpersonal discussion, event "neither relevant, nor emotional"); "how about that?" events (only a subgroup would talk, news "either relevant or emotional"); finally, "Oh my God" events (everybody talks, news "both emotional and relevant"); Greenberg, EE.El° found interest based on personal utility. Also, Hanneman and Greenberg (1973) utilized composite social, attitudinal and interest measures of relevance (importance) and salience (interest) to predict differential information processing. McCombs (1973) does not distinguish between the two and speaks of relevance in terms of discussion of event and interest, interchangeably; relevance is conceptualized as a receiver variable, antecedent to need for orientation and hence ex- posure. 15 Awaco omcfimoc maacnumwocoov R. «mGOHDHmommHo loam Hospfl>flccw ou huflamuucoo mecca mmmE on» CH moammflo Amuoesuv ummumucfi "scammsomwc .mmmum>oo capo: ucm>o CH umoumDCH commmnmxm .ucm>o mo conmsomHQ mofiufl>wuom cousco “wuHmOfimHH Ion “ucm>m CH ummumucw commoumxm muHHHud HMGOmHom ow>flooumm m3oc mo coflumoHMMHo> .ucm>m mo conmsomHo ecannon coumucofiuo Hmumcmm CH uco>o mo coflumsam>m «Hmscw>wc ICH ou Hummcflcmme MHHMCOADMDDHm .mmoao no aco>m mo cofluwcmoowm muonmoflecfl scanneaflmmm msoemflamm mocmflamm mocm>mamm mocw>maom oocm>maom mocc>maomnoocmuuomaH mocmflammnumououcH ummumucH ucoEm>Ho>cH ommuumououcH mocm>oammuoocmcfiuuom mocowammumocm>maom mocm>mamm Ameaav muom AmmmHv cwmumq one «fiance: Amsch mnaoooz AMhmHv cmEoccmm can mumncomuo Ammmav .mm.mm muwncmmno mnmav Haumm Ammmav umuumo Ammmav QOHHDE can mEMUd cowuwcfimmo HMCOfiumnomo 6mm: manmflum> sesum mucmowomucd mo 3mfi>no>o .H manna 16 Other authors talk of interest as a receiver vari- able antecedent to media content consumption (Medalia and Larsen, 1958) in a manner quite identical to McCombs' notion of relevance. Fathi (1973) talks about ego-involve- ment in an event which is "central to the self" in the Sherif and Sherif (1956) sense. Finally, Carter (1965) theoretically subsumed three notions under relevance: situational relevance, salience to individual and perceived pertinence of event, in terms of receiver goals and evalu- ation of the environment. All in all, the definition of importance is often based on post-event inferences, i.e., the news which spread fast and widely must have been important (Budd, ep 31., 1966, Rosengren, 1972, Adams 35 31., 1966). Rota (1973) comes from a different standpoint, where relevance is rela- ted to the comparative emphasis placed in display of given news events by the mass media at the same time; salience is related to news' centrality to a person's cognitive, affec- tive and behavioral predispositions. He proposes use of factorial design based on these definitions, with diffusion of information as the dependent variable. Rota predicts highest diffusion of information about events of both high relevance and high salience, with second best for low rele- vance and high salience. For Rota, the determining factor appears to be salience, the receiver variable, while rele- vance as defined by him remains in the background. Thus, for 17 both the cases of low relevance and low salience, or high relevance and low salience, low salience would determine limited diffusion. It is intriguing to transfer this scheme to the Tichenor ep 31. study and see what explanation would result (Figure 2). Hi Lo Lo Low Minimum 3 Knowledge 4 Knowledge Salience Maximum Second Hi 1 Knowledge Highest 2 Knowledge Figure 2. Salience and Knowledge Relevance Now according to Tichenor, the news events studies had been given wide and pronounced mass media coverage, and these events were assumed to be of general appeal; this outlines a case of high relevance/high salience (Box 1), whereby the burden of differentiation falls on educational level alone. We suggest that, rather than assuming a uniform level of salience, we would take into account its variations, obtain- ing needed contrast, besides educational difference. In other words, the explanation of the knowledge gap can be facilitated when it is viewed as resulting from a case of High relevance/High salience vis a vis High relevance/Low salience, with the ensuing audience behaviors and differen- tial knowledge gain (e.g. Box l/Box 3). 18 An Interest Model As stated before, our aim here is to integrate ele— ments of all the previously mentioned standpoints, since resulting definitions have already been applied with various success in the corresponding studies. Despite the varied nomenclature, they all share an implication of certain selec- tion criteria applied by members of the media audience. In short, people do not engage their attention indiscriminately but rather according to some choice hierarchy which has meaning to them. So in seeking parsimony we may attempt sub- suming already mentioned considerations under the notion of interest and degrees of it. The conceptualization has its roots in a functional approach, which emphasizes utility, from the receiver's perspective (Figure 3). Se 1f \/ Milieu Potential Figure 3. Perceived Utility 19 Interest is a function of the interplay of the fol- lowing components: (a) perceived information utility to gelf; (b) perceived information utility to milieu; (c) per- ceived potential utility to either self or milieu. Utility to self is seen in terms of relatively immediate, daily coping behaviors related to the functioning of individuals and their home and family. Utility to milieu is seen in terms of communicative utility and facilitation related to an individual's social environoment, the various membership and reference groups he is associated with (e.g., friends, relatives, fellow workers, neighbors, etc.). Potential utility refers to routine scanning of the information en- vironment, focused on a relatively consistent set of infor- mation areas kept under surveillance by the individual. To the extent to which an information item is per- ceived to have one or more of the above attributes, result- ing interest will determine the kind and amount of attention an individual will give to that information item. The com- parative emphasis on each attribute commanding attention will depend on the individual's short-term and long-term priorities, habits, pressures or changes in the environment. We believe that defined thus, interest now subsumes the im- portance, relevance, salience, or pertinence notions from a receiver's perspective. In an overview of the literature dealing with man's methods of attending to information, Sears and Freedman 20 (1967) reemphasized the point that people expose themselves and absorb information that is useful or functional in some way. Atkin (1974) in discussing political information and mass media use emphasized adaptive requirements to the indi- vidual's satisfaction. Greenberg ep 31. asked respondents if a major event has been of personal interest (self-con- sumption), or for social uses; they found that of those who attended the event, highly interested respondents found it useful both personally and socially, less interested indi- viduals found it mainly of social utility. In their discus- sion of information seeking behavior, Westley and Barrow (1959) emphasized "the persistent tendency to place a positive value on information that is potentially relevant to the in- dividual's orientation to his surroundings." Atkin (1972) described two modes of exposure, information receptivity and information search, and brought up communicatory utility as an explanation for some media use patterns. In the constant flow of information, few events are attended to by some, and disregarded by many; the bulk of news are noted for intermediate functional importance, scanned by 30-50-70% of the audience, and not necessarily disseminated further (Greenberg, 1964b). But in the way of example, note the changes that took place in the routine scanning for potential utility, when a group of workers were switched to a four-day work week; Nayman ep.gl. (1973) reported that not only mass media exposure increased, but 21 specific types of content were sought out for utilitarian application. A substantial number started reading hobby magazines, news weeklies, increased exposure to outdoor sports reports, gourmet cooking programs, pets, homemaking, sewing, public television, and general interest magazines. As changes in daily c0ping and social behavior took place, new areas of potential utility were included in the routine scanning of the information environment. The newly emerging functional relevance to novel content surveillance was sub— stantiated by the correlation found between an index of gain in viewing various outdoor recreation programs and the index of separate increases in camping, fishing, boating, skiing, etc., after the change to a four-day work week. Similar evidence of the influence daily routines may have on the perceived relevance of incoming information, as an indirect form of functional selectivity, can be found in Hill and Bonjean (1964) and O'Keefe (1969). Perhaps our Figure 4, using a modified version of Lazarsfeld's "different functions," would best convey our idea on the role of interest in information gain discrep- ancies, and illustrate the knowledge gap model underlying this paper's rationale. The response function for information gain is depen- dent on the stimulus situation. A given information item or items are seen as distributions varying in their perceived interest for the receiver, and the information gain is a mcofloocsm omcommmm cflmo coflumEH0mcH .v munmflm . . . med» Hm>o mommmmoe mo Hones: an Ilia». II Ho>oa III et mfioma3ocx \\ HOHHQ 22 ~s mom M \ ureb uorqemzogur — esuodse: go Kurtrqeqoxd v 9» 23 function of interest. If the message was publicized for an indefinite time, theoretical expectations dictate that the knowledge gap regarding the specific information carried by this message will eventually close; actually this can rarely be the case, given the finite nature of mass media publicity on any given tOpic. So whether the resulting knowledge gap closes may depend partly on whether stimulus intensity in the mass media publicity is maintained, or is reduced, or eliminated at a point* (see point in Figure 4--designatedly *) when only the more alert persons have gained that knowl— edge. Previous knowledge level is an indication of prior interest, which may be rooted in relatively constant behavior— al, interpersonal and environmental processes associated with the given type of information. The positive relationship be— tween prior knowledge and information gain has been shown in research (McNelly ep‘gl. 1967, 1968; Fitzsimmons and Osburn, 1969; McLeod, §£.El' 1969). Also, keeping track of informa- tion areas scanned routinely will facilitate information gain in that it implies possible foreknowledge of specific infor- mation items. Adams 33 31. (1969) have demonstrated that attention to media content can be a function of foreknowledge, which is a function of interest. Tracing Interest Differentiation - In view of the central role posited for interest in directing audience attention, and in order to trace the subsequent differen- tials in levels and areas of knowledge manifested in mass 24 media users, it is necessary to look carefully into interest formation; what are the factors that can help explain the shaping, maintenance and changes in the three components of interest postulated above? Some pertinent evidence can be brought to bear in understanding the processes which con- tribute to audience differentiation with respect to some areas of interest and commonality with respect to other areas. Aside from native ability and biological maturation, there are two classical factors called upon to account for intellectual development: environmental influence (family, life experience) and social transmission (education). The family environment in all likelihood sets the track for lasting assumptions regarding EBEE is worthwhile to know (e.g., art) and E92 wgyg to go about it (e.g., reading print sources). The Role of the Family - A number of studies indeed illustrate the shaping influence that family environment exercises on youth's cognitive development and interest direc— tion. Children in politicized families are more likely to be interested in political matters than children from politically apathetic families (Milbrath, 1965). Johnson (1973) also found that interests rest strongly on socialization factors; mother's education and family conversation were best predic— tors of foreign country knowledge. On the other hand, father was found to be a stronger influence by Clarke (1965), who reported that where parents were interested in public affairs, children also manifested more interest in informational 25 content and valued reading as a leisure activity. McLeod and Chaffee (1971) indicate that parent-child interaction partly determines the way a child learns to structure daily situations and to relate information to them. Socialization explanations in their entirety would go beyond the scope of this discussion; the social interaction View within this tradition, however, does seem pertinent to our concerns. It draws on elements of modeling and reinforcement in proposing that early norms involved in a child's interactions with relevant others (e.g., parents and parental circle) will shape behavior, including communication style and preferences. Parental guidance in terms of structuring the child's spare time, emphasis on school performance, expression of certain topics, reinforcing concentration on other tOpics, can all be seen as having potential effect in placing certain areas of interest on the youth's developing "cognitive map" in Tolman's (1932) sense. While most mothers seem to restrict viewing of certain television programs (Lyle and Hoffman, 1971a), it looks like real encouragement and reinforcement in watching specific pypeg of programs comes more so from middle-class parents (Greenberg and Dominick, 1969a). Other authors have found that when "concept-orientation" in family discussions is evi- dent, or where a "pluralistic" atmosphere is part of the habitual pattern, family members display heavy attention to news content. There is evidence that these influences persist 26 beyond childhood and become part of an individual's be- havior in new social situations (McLeod ep 31. 1967; Chaffee, ep_gl. 1971; Chaffee, 1970). Work done by achieve- ment motivation researchers has also linked factors of family interaction and environment to the child's subsequent behavior patterns. Most of the evidence for analysis of social origins and consequences of achievement in children comes from studies on parent-child interaction, parent reports on child rearing practices, and experimental work. Of particular interest to us would be the distinction made between early emphasis on problem solving (e.g., "mastery" school related) versus emphasis on "caretaking" (e.g., tasks around the house), where the former is associated with achievement orientation, and intellectual task independence. Winterbottom (1958) found that parental encouragement was the main factor in middle class boys' achievement orientation; the activities encouraged relatively early in life were: doing well in school, choice of books to read, having own interests and hobbies, doing well in competition with others, etc. The tendency to achieve determines interest, impetus at undertaking an activity with the intention of doing well. In addition, a number of studies have traced the behavior of high achievers later on in life; the evidence suggests that they are displaying better memory for incomplete tasks, are more active in college, community activities, etc. - behaviors that may be linked to a wider range of information needs 27 (Atkinson, 1958; McClelland, 1961). It is interesting enough to note that high achievers do not necessarily per- form compulsively on any task and in any circumstances; rather, they do better when the performance has some sig- nificance to them,i.e., when they see some meaning or utility to the task (French, 1955). In the light of the preceding section, future studies should devote more attention to background factors which determine initial levels of information and interest build- up, upon which there are differential effects later on in life. Also, further attention to the transitional years of adolescence should provide continuity to tracing these processes. The Social Milieu is at work with the adolescent push toward real or imagined independence from parental constraints and more toward peer-defined activities. The influences that bear on the communication behavior cannot be understood prop- erly without clarifying the extent to which the new overlay of peer-defined tastes and notions of "relevance" would depart from the basic directions along which the youth is already in motion. If the milieu is comprised of peers coming from families very much like one's own, there may be no deep change in the fundamental set of criteria already absorbed via family setting. The more diverse the peer milieu then, the more likely are influences to depart from early behavior formats. Coleman ep‘gl. (1966) reports that 28 the scholastic achievement of minority children is often strongly related to intellectual proficiency Of their school- mates. They suggested that socioeconomic mix may have a lot to do with the intellectual responsiveness of youth, regardless of school facilities and resources. Troldahl and Costello (1966) found that media use among teen-agers was bound to its potential for discussion with friends. At the same time, there is evidence that social class continues to play a differentiation role in the communication behaviors Of adolescents; Lyle and Hoffman (1971a) found that youngsters Of white-collar households tend to use newspapers more than their blue-collar counter- parts. Similarly, Greenberg and Dominick (1969a) found lower-income and working class teen-agers spending more time with television. Most importantly, Lyle and Hoffman (1971a) and Clarke (1969a) data seem to contradict Kline's (1970) Opinion that social class will diminish in predictive power as the child grows up. Adult Audiences and the Information Environment - The most definite interest diversification can be expected to occur in response to the pressures, responsibilities, or Opportunities that arise with adulthood. Yet it is more likely to be a matter of increasing specificity in informa- tion interests and needs, rather than a discontinuous shift from previous patterns. A number Of attempts to classify information has been 29 made in a search for effective typologies and audience re- sponse to them. Rosengren (1972, 1973) distinguishes between hard and soft news on the basis of their diffusion rate and proposes subjective interest as a criterion for analyzing the differences; Robinson (1972) talks of school knowledge tOpics (e.g., academic knowledge, public affairs, hard news) vis a vis life knowledge (e.g., health, consumer, human interest) similarly to Chaffee's (1973) news tOpics and consumer topics; finally, Rogers and Shoemaker (1971) bring up awareness knowledge, "how-to" knowledge and prin— ciples knowledge in their description of the innovation diffusion processes. In view of the Observation that almost any mass media content can perform any function, depending on the audience (Rosengren, 1972), it becomes possible to talk about the effectiveness of typologies only to the extent to which they reflect priorities of various audience segments. It becomes necessary then, to accumulate a better under- standing Of consistent variations in functional priorities between and across different social groups; such priorities are tied to utility perception and hence interest distribu— tion among content categories. One such study has been directed at identifying certain attributes of poverty life styles, viewed as mani- festations Of a functional response to the demands of the immediate environmental conditions. Greenberg and Dervin (1972) underscore the fact that a middle-class oriented 30 society such as ours assumes skill, literacy, motivation education and information seeking abilities, all largely fashioned after middle class desiderate, and without taking into account that the poor may not be prepared, nor inclin- ed to deal with society on these terms. The poverty sub- culture emphasized family, friendship and kin relations as a functional response to the realities of poor life. This implies homogeneity of interpersonal contacts within a largely closed system, which leaves them unprepared for role flexibility and social skills expected "outside." It is easy to see the implication for the danger Of defining the knowledge gap on the basis of unrealistic ex- pectations regarding areas of interest and ways of deriving information from the mass media. It would behardly sur- prising tO find that the poors' interests direct them to- wards routine monitoring of media content judged low in information utility according to a middle class yardstick, yet holding personal and social utility promise tO lower strata. By the same token, we need to know what attributes of non-poverty and middle-class life styles would help determine the areas of information that are routinely scanned due to their functional utility to daily demands. Areas of infor- mation close to professional and economic patterns of coping behavior can be expected to take priority, and therefore, to direct interest with regard to media content. 31 Recognizing that the social milieu plays a signifi- cant part in the life of people means including interpersonal communication processes among the factors influencing infor- mation handling and intake. McCombs (1973) notes that inter- personal exchange Of information often may function as an independent variable preceding any selection of messages from the mass media. Depending on the variety Of social contact available to them, peOple of similar demographic groups may also vary in their information selection patterns. .The communicatory utility of incoming information then de- pends largely on the kinds of interpersonal networks that individuals are involved with. Tipton (1970) reports that among respondents discussing election events, the greatest information seeking occurred for people whose friends were equally interested in a given issue. Larsen and Hill (1953) proposed that interpersonal communication about an issue be used as a measure Of interest. Chaffee and McLeod (1967) found that voters who anticipated conversations on 1966 campaign issues were more likely to request information pamphlets than those not planning to discuss the topics in the near future. Berelson ep 31. (1954) found that one's primary groups tend to be politically homogeneous, and Katz and Lazarsfeld (1955) have demonstrated that "like talks to like" based on the Opportunity for interaction between people placed within similar social loci. Chaffee and McLeod (1973) report that ongoing communication between respondents and 32 their social contacts accounted better for information seeking as compared to individual differences alone. Since interpersonal communication tends to require social and in- formational equivalence, we can expect that perceived utility is subject to the influence of what may be termed an agenda set by the milieu. Relatively permanent social networks will encourage the persistence of certain selec- tion criteria, while changes in the social milieu should result in new perceptions of information utility, and par- tial re-direction Of interests. A SUMMARY We have explored the knowledge gap phenomenon in the light Of available findings, and also proposed an interest model of analysis. Some of the main points we sought to emphasize can be recapitulated as follows: (1) the initial presentation of the knowledge gap hypothesis used education as the main locator variable in detecting information level discrepancies, and focused on the print media. (a) while demographic variables have been help- ful with respect to the print media, they have little power in predicting the use of electronic media. (b) exposure to print media contributes well to audience's information acquisition, yet there is indication that the manner of media 33 use may take precedence over the print broadcast dichotomy. (c) evidence that exposure to the mass media does not necessarily enhance information gain dictates a need to study other de- terminants of the information acquisition processes; thus, we gave consideration to interest as the underlying factor of receiver attention given certain content areas and program components in the mass media. (2) the interest model presented is rooted in a functional approach which emphasized information utility from the receiver's perspective. Some considerations flowing from such a model will be further examined in the following section. HYPOTHESE S The preceding discussion in its entirety has been directed at clarifying the interplay of factors deemed rele- vant to the study Of the knowledge gap phenomenon. Our position has been that its fruitful examination should go beyond education-bound stratification of audiences, include the electronic media and allow for a more detailed look into interest diversities. We do recognize that education develops cognitive skills which facilitate handling Of mass media in- formation; that it expands receiver horizons for events of 34 significance (Buss, 1969) and is likely to widen an indi- vidual's overall scope Of interests (Wade and Schramm, 1969). Yet at the same time it must be noted that this does not require the assumption of interest homogeneity among similarly educated audiences. As discussed earlier, people do not engage their attention indiscriminately; rather, they are likely to apply certain choice criteria in attending to media content. Thus, we suggested than an interest-based model would allow a more sensitive examina- tion Of information gain processes and help trace the patterns of knowledge differentiation among mass media audiences. In explicating our notion Of the independent variable, interest, we proposed a treatment in terms of the underlying components: perceived information utility to self, perceived information utility tO milieu, and perceived potential utility to self or milieu, To the extent that an information item is perceived as having one or more of these attributes, resulting interest will determine the kind of attention an individual will give to that information item. In the absence Of empirical evidence on the relative weight of each component and their relationship in determining interest, we can nevertheless posit a certain priority ordering for them. Thus it can be expected that an individ- ual's vested interests (self, mate, immediate family) will take precedence over the social milieu; also, relatively immediate concerns would prevail over delayed ones. 35 Given such a framework, we expect that interest, com- pared tO education, will turn out to be a better predictor of knowledge acquired from media content. A tentative tuxtaposition Of these two variables may be useful in the way of clarification, where level of information is the dependent variable (1 = highest); it also suggests the possibility of interaction between the two (Figure 5). Interest Hi LO Education H1 1 3 Lo 2 4 Figure 5. Juxtaposition of Interest and Education In terms of the dependent variable, knowledge, we adopted Atkin and Greenberg's (1974) component measure differentiat- irig between factual and structural knowledge, treating them separately in hypothesis testing. Factual knowledge refers to the respondent's knowledge of specific items, names, dates, places, facts and figures, relatedtoflépecific itemsfl names, dates, places, facts and figures, related to; \Zpecific news occurrences. Structural knowledge is taken as the respondent's understanding Of the relationships 36 manifested in an event, how or why it took place, and the event's place in the broader framework of related phenomena. The model we have proposed incorporates a time di- mension and the independent variable of overall interest. The following three general hypotheses are addressed to these aspects: HA: As the infusion Of mass media information into a social system continues, those with a higher level of interest will acquire new information faster than those less interested, so that the knowledge gap be- tween them will tend to increase. H : At any point in time, then, more interested members of the media audience will display a higher level Of knowledge than those less interested in a publicized event. H : As the publicity on a topic continues over a long period Of time, the knowledge gap between those more and less interested will begin to decrease. Furthermore, we have postulated certain types of interest stemming from various perceptions of utility. We have indicated the use of multiple dependent variables, e.g., factual and structuralknowledge treated as replicates for hypothesis testing; and we have already discussed in a com- parative fashion education and interest as predictors of information level. The following main and derived specific hypotheses have been directed accordingly: H1: There is a positive correlation between education and knowledge. Hla: Education will correlate positively with both factual and structural knowledge. 37 There is a positive correlation between overall interest and knowledge. H2a: Interest stemming from perceived utility to self will be a stronger predictor for both factual and structural knowledge, than interest stemming from perceived utility to milieu. Interest stemming from perceived potential utility to self will be a stronger predictor for both factual and structural knowledge, than interest stemming from perceived potential utility to milieu. 2b: Interest stemming from immediate utilities will be a stronger predic- tor Of knowledge than interest stemming from potential utilities. 2c: H : Education and overall interest combined will correlate more strongly with knowledge, than either one taken alone. H : The correlation between overall interest and knowledge will be higher than the correla- tion between education and knowledge. H4a: Interest stemming from utility to self will correlate higher than will educa- tion with both structural and factual knowledge. Interest stemming from utility to mi- lieu will correlate higher than will education with both structural and factual knowledge. 4b: Interest stemming from potential util- ity to self or milieu will correlate higher than will education with both structural and factual knowledge. 4c: The following diagram of the hypothesized relation- ships may help in making the synchronic part of the model more visually apparent (Figure 6). 38 K N O W L E D G E (Factual and Structural) O V E R A L L mmflnmcoflumamm ooNHmmsuommm mo Emummflo .m ousmflm A OA EV T A B FIJCJDUflf-IHOZ A swede: mammhmwuv am A DOHHHS V. m mamm . mufiaeuo ©o>fimonmm N smflaflz V .H. maow . A A l\ Omwafizlva H mawm [\ _____-...-1 ------I--I----—-----Ib--I ——-——4I—— ———_———T————4 CHAPTER II METHODOLOGY The data collection for this dissertation was done in a panel survey from August 6-16, 1974 in the greater Lansing area, Central Michigan. This chapter outlines the procedures employed, namely, questionnaire development, sampling, questionnaire administration and data coding; also, in this chapter we report on the measurement and in- dexing Of variables specific to this study. Questionnaire Development The questionnaire was developed in July 1974, with the objective of tapping, at two points in time, respondent interest and knowledge regarding two different kinds of events publicized in the media: the National Football League strike and the impeachment developments. The selec- tion of these topics was guided by the following criteria consistent with study goals: First, the need was for at least two news topics contrasting in the likely interest they hold for the mass media audience. Thus we made the assump- tion that the football strike would have less general appeal compared to impeachment events. 39 40 Second, the need was for news topics having different duration of display in the mass media. Again, the football strike was recent, while the impeachment events had a relatively longer standing. And finally, once selected, both events had to remain in the news, at least throughout the period Of study. The questionnaire was pretested with adult residents of East Lansing in order to check on item wording and vari- ances and improve the phone administration style. Inter- viewers for the pretest were graduate students at Michigan State University. They were trained in a two hour session prior to the pretest and debriefed afterwards. The entire process was carried out under the supervision of the project director herself. In all, twenty-five respondents selected randomly from the East Lansing telephone listings were inter- viewed for pretest in one evening. The final questionnaire, designed for the first administration and put in a code- book form, was seven pages long and included predominantly closed-ended items. The average time for interview comple- tion with a respondent was ten minutes. The subsequently developed questionnaire designed for the second administra- tion was five pages long and took an average Of seven minutes to complete with a respondent. The two versions Of the questionnaire will be discussed in a later section of this 41 chapter. Sampling The survey site for this study was the greater Lansing area in Central Michigan. This included Lansing, East Lansing, Bath, De Witt, Diamondale, Eaton Rapids, Grand Ledge, Holt, Laingsburg, Mason, Okemos, Perry, Potter- ville, Shaftsburg and Williamston. The choice corresponded to the goal of reaching varied population strata and the practical need for accessibility by phone from the project headquarters on the MSU campus. Thus, three communities, Charlotte, Dansville, and Onondaga, were excluded since they fell outside the local call area. The estimated pOpulation was 115,482 with 73% residing in Lansing, East Lansing and Okemos and 27% in the surrounding area outlined above. We also checked to ascertain availability and reception of the three main TV station signals throughout the area. The telephone directory was used as the sampling frame to draw a systematic probability sample (n=400) for this study, com- pleted in late July 1974 with a check for overlaps with the pretest respondents. Actual respondent selection within each household followed the procedure recommended by Troldahl and Carter (1964). Eligible respondents were all adults, ages 20-80. At least two call attempts were made for each interview obtained. In all, 253 usable interviews were com- pleted during the first wave, with 28% refusals, disconnects 42 or not-at-homes, 2% not eligible and 7% "don't-call-next- week." The second wave completed 243 interviews, with 10 respondents unwilling or unable to cooperate again. FOr further details on sample characteristics see Appendix A. Questionnaire Administration This author was the project director supervising all phases in the administration of telephone survey question- naires. Twenty-two students at Michigan State University, eleven males and eleven females, were hired as interviewers. All interviewers spent four evenings on the job. Prior to actual interviewing, all interviewers went through a thorough training session which consisted of the following: (a) Review and discussion of the questionnaire taken item by item, with an emphasis on Optimum familiariza- tion with its contents. (b) Discussion and practice of introduction to respon- dents, handling Of problems, maintaining rapport, with appro- priate utilization of experiences from pretesting. (c) Review and discussion of the interviewer's role, and caveats with respect to potential introduction Of bias. (d) Practice Of interviewing by role-playing and then with two outside respondents not included in the sample. Interviewing days were Wednesdays and Thursdays for botfil waves, beginning around seveh P.M., until approximately 8::3C) P.M. On the average, interviewers completed 6.4 inter- Vlews each evening. Validation of the interviews was done 43 by the project director on 38 randomly drawn interviews, or 15% Of completed interviews. Validation took place within two days from the actual interview date of the second survey wave. It was aimed at ascertaining that the same designated respondent had been interviewed both times, with a check on age, sex, and replies on three randomly selected closed- ended items from the questionnaire. Validation results con- firmed the age, sex and identity of respondents in 34 out of 38 cases; three respondents were not available, but another household member confirmed that interviewing had taken place on the evening in question; one household was reluctant to provide information for validation purposes. With respect to closed-ended items, we compared interviewer- recorded quantitative codes with the validation check results, using Stempel's (1955) Percentage Agreement Index procedure. From a total of 102 items validated, (e.g., 34 interviews x three items each), 97 items checked out on exact code agreement, or approximately 95%. Since some Of the items were knowledge questions, it is possible that some of the discrepancies were due to information picked up by respondents after the actual interview. The questionnaires were in code-book form and in- cluded four Open-ended items out of sixty for phase one, and two Open-ended out of thirty-four in phase two. These open-ended questions were exploratory and thus not included in the present analysis, but coded and kept for later 44 examination. The remaining data derived from closed-ended items were transferred to IBM cards for machine analysis. There were differences in items between code—books: phase one contained demographic questions which were not repeated in phase two; also, phase one contained items on perception of potential utility not repeated in phase two; finally, in terms Of knowledge questions, two items out Of six remained constant across phases; four items varied, e.g. "current" questions during phase two replaced "Old" items from phase one . MEASUREMENT OF VARIABLES Overall, eight variables were created for the pur- poses Of this study: five independent and three dependent variables. The basic independent variables were: interest stemming from perceived utility to self, utility to milieu, potential utility to self, potential utility to milieu, and in addition, overall interest. The basic dependent vari- ables were: factual knowledge and structural knowledge and in addition, overall knowledge. These variables were used at two points in time with regard to two topics (NFL strike and impeachment deve10pments), treated as replicates within our study. Each of the basic independent variables was created as a single index consisting of factor scores obtained from a factor analysis Of the component items done separately 45 for each topic, at two points in time. Similarly, each basic dependent variable was created as a single index Obtained by summing standardized response scores across knowledge items. The variables overall interest and overall knowledge were created by summing the single index scores across the above basic four independent and two independent variables, respectively. Education was measured by Obtain- ing the last grade in school completed by the respondents. These data were collapsed into six categories: less than to sixth grade, junior high to some high school, finished high school, some college, finished college, and graduate work (see Appendix A). We shall now proceed with the indexing procedures specific to each event. See Appendix B for a detailed de- scription of generalized measurement procedures preceding final index creation. National Football Strike - Measurement Of Variables 1. Independent Variables, National Football Strike (NFL) (a) The variable SELF INTEREST was a single index built from the factor scores of three measures tapping perceived utility to self. (b) The variable MILIEU INTEREST was a single index built from the factor scores Of five measures tapping perceived utility to self. (c) The variable POTENTIAL SELF INTEREST was a single index built from the factor scores of three variables tapping perceived utility to self. (d) The variable POTENTIAL MILIEU INTEREST was a single index built from the factor stores Of 46 five measures tapping perceived potential utility to one's milieu. (e) OVERALL INTEREST was a composite index obtained by summing the single index scores on the above four variables. As mentioned before, Appendix B describes in full detail the preliminary work involved in establishing the final set of component measures for each predictor class. A brief recapitulation here should, therefore, suffice. In terms of self interest and potential self interest, we started with five questionnaire items; 1. NFL Effect (NFL Potential Effect): DO you think the NFL strike has an effect on your life in any way? (DO you think the NFL strike could affect you in any way in the near future?) 2. NFL Cost (NFL Potential Cost): DO you think the strike has an effect on the cost of living, or prices for you? (Would there be an effect on the cost of living or prices for you?) 3. NFL Job (NFL Potential Job): Do you think the strike has an effect on your job, or the job of someone close to you? (Could it af- fect your job, or the job of someone close to you?) 4. NFL Enjoyment (NFL Potential Enjoyment): Does the NFL strike have an effect on the enjoyment you get out of watching the game? (Will the strike have an effect on your enjoyment watching the games?) 5. NFL Keeping Up (NFL Potential Keeping Up): Is the NFL strike the kind of thing you just want to keep up with? (Is the NFL strike the kind of thing you will want to keep up with?) All response categories to the above items were dichotomous. 47 The factor analytic procedures described in Appendix B led to abandoning Of items four and five; thus the final set of three items comprising the category self interest (potential self interest) was: NFL effect, NFL job and NFL cost. In terms of milieu interest and potential milieu interest, we started with four questionnaire items; 1. NFL Talk to Friends (NFL Potential Talk to Friends): Have you discussed it with friends? (DO you think you will talk about it with friends?) 2. With relatives? 3. With peOple at work? 4. Anybody else? All response categories to the above items were dichotomous. The factor analytic procedures described in Appendix B led to the addition Of one item previously viewed as part of the self interest group; thus the final set Of five items comprising the category milieu interest (potential milieu interest) was: NFL enjoyment, NFL talk friends, NFL talk relatives, NFL talk at work, and NFL talk others. Here, we shall present the last phase in measurement procedures - building the independent variable indices. Independent variable indexing entailed the following steps: (a) Factor analyses (Quartermax) of each group Of component variables, in order to arrive at a factor score 48 coefficient for each component. (b) Next, factor scores were created for each respondent on the chosen factor, i.e. multiplying a respon- dent's standardized score on the component variable by the factor score coefficient for that variable. (c) The final index score was obtained by summing a respondent's standardized scores on 3 component measures, each multiplied by the appropriate factor score coefficients. Accordingly, we shall now report the appropriate tables for each group of component measures involved in arriving at the independent variable indices for NFL, time one and two. Following each group Of tables, we shall note the resulting independent variable index range, mean and standard deviation. 49 NFL, TIME ONE SELF INTEREST Table 2. Factor Matrix of the Three Measures of Self Interest Factor la Factor 2 Factor 3 NFL Effect 0.81417 0.03685 0.07201 NFL Cost 0.73507 0.18163 -0.05769 NFL Job 0.57266 -0.21739 —0.01318 Proportion of variance accounted for by factor 94.5% 5% 0.5% athe factor chosen Table 2A. Pearson Product-{lament Correlations among the Component Measuresa 'NFL EffeCt 'NFL COSt NFL JOb NFL Effect 1 . 00000 NFL C081: 0 . 60101 1 . 00000 NFL Job 0.45728 0.38222 1.00000 acorrelations of i .13 or greater are significant with n=253 at p<.05. Table ZB. Variable Camunalities , Factor Purities and Factor Score Coefficients for the Overall Index of Self Interest ' 2 a Factor Scoreb Variable Ccmnunality (h ) Factor Purity Coefficient NFL Effect 0.66942 0.99021 0.50965 NFL Cost 0.57665 0.93699 0.35011 NFL Job 0.37537 0.87361 0.20579 h2 is interpreted as the percent of variance in each variable explained by the factor solution, including all its factors. a‘factor purity is obtained by dividingzthg squared factor loading by the variable's carmunality, i.e. (FL) /h . It is interpreted as the prOportion of variance accmmted for in a variable by the chosen factor . bthe factor score coefficient can be interpreted as the beta-weight for the variable's regression on the hypothetically constructed factor. The resulting index of self interest ranged from -.357 to +3.317, with a mean zero and a standard deviation .889. 50 NFL,TD’IECNE MILIHJINTEREST Table 3. Factor Matrix of the Five Measures of Milieu Interest Factor 1a Factor 2 Factor 3 NFL Enjoyment 0.43112 -0.04802 -0.04867 NFL Talk Friends 0.90607 0.11012 0.24014 NFL Talk Relatives 0.74278 -0.24814 0.03849 NFL Talk at Work 0.73808 0.35059 0.03197 NFL Talk Others 0.54490 -0.00117 -0.10585 PrOportion of variance accounted for by factor 90.8% 7.1% 2.1% athe factor chosen Table 3A. Pearson Product-Manent Correlations anong the Component Measuresa NFLTalk NFLTalk NFLTalk NFLTalk NFL Enjoy Friends Relatives at Work Others NFL Enjoyment 1.00000 NFL Talk Friends 0.37294 1.00000 NFL Talk Relatives 0.33110 0.65553 1.00000 NFL Talk at Work 0.30063 0.71548 0.46178 NFL Talk Others 0.23902 0.46796 0.40142 1.00000 acorrelations of i .13 or greater are significant with n=253 at p<.05. Table 3B. Variable Camunalities , Factor Purities and Factor Score Coefficients for the Overall Index 2 a Factor Score Variable Camunality (h ) Factor Purity Coefficient NFL Enjoyment 0.19054 0.97543 0.07003 NFL Talk Friends 0.89075 0.92165 0.55542 NFL Talk Relatives 0.61478 0.89742 0.23476 NFL Talk at Work 0.66869 0.81466 0.16866 NFL Talk Others 0.30812 0.96361 0.10678 2 h,a,b-seeTab1e 2B The resulting index of milieu interest ranged from -.445 to +3.023, with mean zero and standard deviation .944. .— r." rr t.1 r l 3 v "i h t '1) 109 am W n: rfl 51 NFL, TIME ONE POTENTIAL SELF INTEREST Table 4. Factor Matrix of the Three bbasures of Potential Self Interest Factor la Factor 2 Factor 3 NFL Potential Effect 0.67834 -0.20812 -0.01028 NFL Potential Prices 0.85394 0.00292 0.02274 NFL Potential Job 0.69632 0.20696 -0.01149 Proportion of variance accounted for by factor 95.1% 4.9% 0.0% athe factor chosen Table 4A. Pearson groduct-Fbment Correlations among the Component NEasures NFL Poten- NFL Poten- NFL Poten- tial Effect tial Prices tial Job NFL Potential Effect 1.00000 NFL Potential Prices 0.57842 1.00000 NFL Potential Job 0.42939 0.59496 1.00000 acorrelations of i .13 or greater are significant with n=253 at p<.05. Table 43. Variable Commmalities , Factor Purities and Factor Score Coefficients for the Overall Index b . . 2 . a Factor score Variable Communality (h ) Factor Purity CoeffiCient NFL Potential Effect 0.50357 0.91375 0.24391 NFL Potential Prices 0.72973 0.99928 0.55862 NFL Potential Job 0.52783 0.91859 0.25923 hz, a, b — see Table 213 rIhe resulting index of potential self interest ranged from -.408 to + 2.993, with mean zero and standard deviation .907. 552 .mam.~ cofiumfi>oe nuance». can ouon came nufi3 .vvm.m+ ou OH5.HI Eoum oomcmu oco mafia um Amz MOM umououcw Hamuo>o mo xOOCw can .0ma< .voa. coflumfi>oe Oumpcmum can ones some cuw3 .mmm.~+ on nom.| Eoum Ommcmu ummuoucw SOHst Hmwucouom no xopcfi mcwuasnmu one wanna mom I a .u . n oomeo.o mmvmm.o veam~.o xame mumnuo Haflucmuom qmz oemmm.o asqmm.o hoomh.o xame xuoz aneucwuoa aez ~mmm~.o mflhvm.o owmam.o game mo>flueamm Hehucmuoa qnz msvsm.o «momm.o Hmmmm.o xdme mecwfium Hanucouoa qmz Hoqmo.o vama.o oevwm.o ucwssoflcm Hmnucmuom qmz ucmnofimmmoo exhausm uouomn .~nc suede::EEoo wanaflum> thoum HOUUMW xoccH Hamuo>o on» now mucoaoamwooo ouoom uouomh can mowueunm uouomm .mowuflamcssEoo manmwum> .mm manna .mo.vm um mmmuc cua3 annuauficmwm cum noumoum Ho ma. H No mcowuoamuuoo fl ooooo.H mommm.o “momv.o wmoam.o mvvm~.o xame mumeuo Hmeucouom nmz ooooo.a Hmooo.o nmome.o oaomm.o xflme xuoz Huflucmuom amz ooooo.H pmvmm.o maamv.o xams mm>fiumflmm amwuamuom Amz ooooo.~ mmdam.c xdea mecoflum Hmflucouoa amz ooooo.d ucmsaomcm Hmflucouom qmz xama unmnuo game xuoz game mm>fiumamm game mecowum ucmsmoflam Hmwucouoa qmz Hmflucmuom qmz Hmfiucmuoa qmz HMMUCOUOA Ahz Hmflucouom nmz M mousmmoz ucocomsoo on» Ocoam COADOHOHHOU acoEOZIuoscoum comumom .«m manna comoco uouoou ocum w>.m wo.m wm.mw HOOOOM xo mom coucsooom mocmwum> mo coHuuomoum emo-.o- maneo.c ovom¢.o ease mumsuo Hmfiucmuoa qmz mmdmo.o vemm~.ou shovm.o xaae xuoz Hmflucmuom qmz vmmao.o- mammm.o mamem.o game mm>fiuuawm Hmfiucouom amz omom~.o endefl.o omeom.o game mecmflum Hawucouoa qmz fiasco.ol ~o~oa.os mnmmm.o unmasoflcm Hufiucmuoa qmz m HOUOMh N HOUOMh OH HOHOME umUHUUCH DGfiHflz HMHUGOUOQ NO mmhflmmwz m>flh 03“ MO Xflhuflz HOfloflh om @HQMB BmQKMBZH DNHAHZ AdHBszOm HZO mZHB .Ahz 53 NFL,TD’IETDD SELFINTEREST Table 6 . Factor Matrix for the Three Measures Of Self Interest Factor la Factor 2 Factor 3 NFL Effect 0.76854 -0.15306 -0.09111 NFL Cost 0.81789 -0.06155 0.10427 NFL Job 0.61106 0.21484 -0.01143 Proportion Of variance accounted for by factor 94.7% 4.2% 1.1% athe factor chosen Table 6A. Pearson Product-Nbrent Correlations among the Carponent Measuresa NFL Effect NFL Cost NFL Job NFL Effect 1.00000 NFL Cost 0.62851 1.00000 NFL Job 0.43778 0.48536 1.00000 acorrelations of i .13 or greater are significant with n=243, at p<.05. Table GB. Variable Camunalities , Factor Purities and Factor Score Coefficients for the Overall Index 2 a Factor Scoreb Variable Carmunality (h ) Factor Purity Coefficient NFL Effect 0.62238 0.94901 0.37344 NFL Cost 0.68360 0.97855 0.47871 NFL Job 0.41968 0.88970 0.21522 2 h,a,b-seeTable 28 The resulting index Of self interest ranged from -.278 to +4.305, with mean zero and standard deviation .900. 54 NFL, TIMETMD MILIEUINTEREST Table 7. Factor Matrix for the Five Measures of Milieu Interest Factor 1a Factor 2 Factor 3 NFL Enjoyment 0.50612 -0.18377 -0.00486 NFL Talk Friends 0.90882 0.23896 -0.00649 NFL Talk Relatives 0.81484 0.15276 0.15589 NFL Talk at Work 0.86766 0.11007 -0.14647 NFL Talk Others 0.47121 -0.23570 0.00261 Proportion of variance accounted for by factor 93.0% 5.5% 1.5% athe factor chosen Table 7A. Pearson Product-Moment Correlations among the Component Measuresa NFL Talk NFL Talk NFL Talk NFL Talk NFL Enjoy Friends Relatives at Work Others NFL Enjoyment 1.00000 NFL Talk Friends 0.41886 1.00000 NFL Talk Relatives 0.38191 0.77606 1.00000 NFL Talk at Work 0.41771 0.81571 0.70112 1.00000 NFL Talk Others 0.28204 0.36948 0.34978 0.38411 1.00000 acorrelations of i .13 or greater are significant with n=243 at p<.05. Table 7B. Variable Carmmalities , Factor Purities and Factor Score Coefficients for the Overall Index 2 a Factor Scoreb Variable Communality (h ) Factor Purity Coefficient NFL Enjoyment 0.28995 0.88342 0.09630 NFL Talk Friends 0.88309 0.93529 0.44197 NFL Talk Relatives 0.71160 0.93305 0.19593 NFL Talk at Work 0.78641 0.95729 0.29097 NFL Talk Others 0.27760 0.79981 0.10045 2 The resulting index Of milieu interest ranged from -.437 to +3.123, with mean zero and standard deviation .954. The index of overall interest for NFL at time two ranged from -.715 to +7.427, with mean zero and standard deviation 1.473. 55 2. Dependent Variables, National Football Strike (NFL) (a) The variable FACTUAL KNOWLEDGE was a single index obtained by summing the standardized scores on responses to three factual questions regarding the football strike. (b) The variable STRUCTURAL KNOWLEDGE was a single index obtained by summing the standardized scores on responses to three questions of issue understanding. (c) OVERALL KNOWLEDGE was an index built by summing the single index scores on the above two vari- ables. NFL, TIME ONE - FACTUAL KNOWLEDGE The three measures of factual knowledge were obtained from responses to the following questionnaire items: (1) What are some of the demands Of the National Football League players? (NFL 1) (2) How long has the strike been on? (NFL 2) (3) What star quarterbacks have crossed the picketlines? (NFL 3) Table 8. Means and Standard Deviations of Three Measures of Factual Knowledge Variable Mean Standard Deviation Cases NFL 1 0.4071 0.4923 253 NFL 2 0.1423 0.3500 253 NFL 3 0.3399 0.4746 253 After summing the standardized scores on these measures, the resulting index Of factual knowledge ranged from -l.950 to +5.046, with mean zero and standard deviation 2.336. 56 NFL, TIME ONE - STRUCTURAL KNOWLEDGE The three measures of structural knowledge were obtained from responses to the following questionnaire items: (1) What is Ed Garvey's role in the NFL strike? (NFL 4) (2) Have exhibition games been successful with rookies and free agents playing? (NFL 5) (3) Do you think veterans lose money by remaining on strike? (NFL 6) Table 9. Means and Standard Deviations Of Three Measures of Structural Knowledge Variable Mean Standard Deviation Cases NFL 4 0.1660 0.3728 253 NFL 5 0.3557 0.4797 253 NFL 6 0.5217 0.5005 253 After summing the standardized scores on these measures, the resulting index of structural knowledge ranged from -2.229 to +4.536, with a mean zero and standard deviation 2.122. The index Of overall knowledge about NFL at time one ranged from -4.l79 to +9.582, with mean zero and standard deviation 4.034. 57 NFL, TIME TWO - FACTUAL KNOWLEDGE The three measures of factual knowledge were ob- tained from responses to the following questionnaire items: (1) How long has the strike been going on? (NFL 3) (2) How long is the cooling-Off period supposed to last (NFL 4) (3) What is the decision of the Minnesota Vikings regarding the cooling-off period? (NFL 5) Table 10. Means and Standard Deviations for Three Measures of Factual Knowledge Variable Mean Standard Deviation Cases NFL 3 0.3990 0.4910 243 NFL 4 0.3130 0.4650 243 NFL 5 0.0780 0.2690 243 After summing the standardized scores on three measures, the resulting index of factual knowledge ranged from -1.776, to +6.129, with mean zero and standard deviation 2.283. 58 NFL, TIME TWO - STRUCTURAL KNOWLEDGE The three measures Of structural knowledge were Obtained from responses to the following questionnaire items: (1) Are veterans going to play in exhibition games in the coming weeks? (NFL 1) (2) What is Ed Garvey's role in the NFL strike? (NFL 2) (3) Can veterans walk out again if agreement is not reached in two weeks? (NFL 6) Table 11. Means and Standard Deviations for Three Measures of Structural Knowledge Variable Mean Standard Deviation Cases NFL 3 0.3790 0.4860 243 NFL 2 0.2590 0.4390 243 NFL 6 0.3990 0.4910 243 After summing the standardized scores on these measures, the resulting index Of structural knowledge ranged from -2.182 to +4.190, with mean zero and standard deviation 2.365. The index of overall knowledge about NFL at time two ranged from -3.958 to +10.319, with mean zero and standard devia- tion 4.183. 59 Measurement of Variables - Impeachment 1. Independent Variables (a) (b) (C) (d) (e) The variable SELF INTEREST was a single index built from the factor scores of four measures tapping perceived utility to self. The variable MILIEU INTEREST was a single index built from the factor scores of four measures tapping perceived utility to one's social milieu. The variable POTENTIAL SELF INTEREST was a single index buiIt from the factor scores of four measures tapping perceived potential utility to self. The variable POTENTIAL MILIEU INTEREST was a single index built from the factor scores of four measures tapping perceived potential utility to one's milieu. The variable OVERALL INTEREST was a composite index obtained by summing the single index scores of the above four variables. As mentioned before, Appendix B describes in full detail the preliminary work involved in establishing the final set of component measures for each predictor class. A brief recapitulation here should, therefore, suffice. In terms of self interest and potential self interest, we started with five questionnaire items; 1. Impeachment Effect (IMP Potential Effect): DO you think the impeachment events have an effect on you in any way? (Do you think impeachment deve10pment could have an effect on your life in the near future?) 2. Impeachment Cost (IMP Potential Cost): DO impeachment events have an effect on the cost of living or prices for you? (Could there be an effect on the cost Of living or prices for you?) 60 Impeachment Job (IMP Potential Job): Do you think these events have an effect on your job, or the job of someone close to you? (Could there be an effect on your job, or the job of someone close to you?) Impeachment Satisfaction (IMP Potential Satisfaction): DO impeachment events have an effect on your general satisfaction with things around you? (Will impeachment events have an effect on your general satisfaction with things around you?) Impeachment Keeping Up (IMP Potential Keeping Up): Are the impeachment events the kind of thing you just want to keep up with? (Are impeachment events the kind of thing you will want to keep up with?) All response categories to the above items were dichotomous. The factor analytic procedures described in Appendix B led to abandoning Of item five; thus the final set of four items comprising the category self interest (potential self interest) was: IMP effect, IMP job, IMP cost, and IMP satisfaction. In terms of milieu interest and potential milieu interest, we started with four items; 1. Impeachment Talk with Friends (IMP Potential Talk Friends): Have you discussed the impeachment events with friends? (DO you think you will be talking about it with friends?) With relatives? With people at work? Anybody else? 61 All response categories to the above items were dichotomous. The factor analytic procedures described in Appendix B suggested no changes, so the final set of four items com- prising the category milieu interest (potential milieu interest) was the same as above; i.e., Impeachment talk friends, IMP talk relatives, IMP talk at work, IMP talk others. Here, we shall present the last phase in measurement procedures - building the independent variable indices. Independent variable indexing entailed the following steps: (a) Factor analyses (Quartermax) of each group Of component variables, in order to arrive at a factor score coefficient for each component. (b) Next, factor scores were created for each respondent on the chosen factor, i.e. multi- plying a respondent's standardized score on the component variable by the factor score coefficient for that variable. (c) The final index score was Obtained by summing a respondent's standardized scores on 3 component measures, each multiplied by the appropriate factor score coefficients. Accordingly, we shall now report the apprOpriate tables for each group of component measures involved in arriving at 62 the independent variable indices for Impeachment, time one and two. Following each group of tables, we shall note the resulting independent variable range, mean and standard deviation. 63 IMPEACHMENT, TIMEONE SEIFIN'I‘EREST Table 12. Factor.Matrix for the Four.Measures of Self Interest Factor la Factor 2 Factor 3 IMP Effect 0.59069 -0.00569 0.21591 IMP Oost 0.64777 -0.05759 -0.09900 IMP Job 0.63652 -0.10274 -0.07l78 IMP Satisfaction 0.43458 0.22425 -0.00200 Proportion of variance accounted for by factor 91.6% 5.0% 3.4% athe factor chosen Table 12A. Pearson groduct-Fbment Correlations among the Carponent Measures IMP Satis- IMP Effect IMP Cost IMP Job faction IMP Effect 1.00000 IMP Cost 0.36147 1.00000 IMP Job 0.36119 0.42627 1.00000 IMP Satisfaction 0.25502 0.26897 0.25354 1.00000 8correlations of i .13 or greater are significant with n=253, p<.05. Table 12B . Variable Commmalities , Factor Purities , and Factor Score Coefficients for the Overall Index 2 a Factor Scoreb Variable Oommnality (h ) Factor Purity Coefficient IMP Effect 0.39556 .88206 0.29778 IMP Cost 0.43273 .96965 0.34915 IMP Job 0.42087 .96264 0.33452 IMP Satisfaction 0.23915 .78967 0.17991 2 is interpreted as the percent of variance in each variable explained by the factor solution, including all its factors. afactor purity is obtained by dividing the squared factor loading by the variable's communality, i.e. (FL)2 . It is interpreted as the proportion of variance accounted for in a variable by the chosen factor. b the factor score coefficient can be interpreted as the beta-weight for the variable '3 regression on the hypothetically constructed factor. The resulting index on self interest ranges from -l.492 to +.982 with a mean zero and standard deviation .833. 64 IMPEACHMENT, TIME ONE MILIEU INTEREST Table 13. Factor Matrix for the Four Measures of Milieu Interest Factor la Factor 2 Factor 3 IMP Talk Friends 0.80314 0.09148 0.10294 IMP Talk Relatives 0.73261 0.20998 0.06513 1MP Talk at Work 0.70288 -0.06301 -0.l7006 IMP Talk Others 0.44708 -0.l9465 0.00446 Proportion of variance accounted for by factor 93.4% 5.3% 1.3% athe factor chosen Table 13A. Pearson Product-Marent Correlations among the Ccmponent Measuresa IMP Talk 1MP Talk 1MP Talk IMP Talk Friends Relatives at Work Others IMP Talk Friends 1.00000 IMP Talk Relatives 0.61505 1.00000 IMP Talk at Work 0.54106 0.49080 1.00000 IMP Talk Others 0.34221 0.28649 0.32586 1.00000 acorrelation of i .13 or greater are significant with n=253, p<.05. Table 13B. Variable Cmmunalities , Factor Purities and Factor Score Coefficients for the Overall Index 2 a Factor Scoreb Variable Caummelity (h ) Factor Purity Coefficient 1MP Talk Friends 0.66400 .97143 0.42293 1MP Talk Relatives 0.58506 .91735 0.29532 IMP Talk at Work 0.52693 .93758 0.28882 IMP Talk Others 0.23779 .84057 0.12363 112, a, b- seeTable 1213 The resulting index of milieu interest ranged from -1.766 to +.787, with mean zero and standard deviation .902. 65 IMPEACHMENT, TIMEONE POTENTIAL SELFJNI'EREST Table 14 . Factor Matrix for the Four Measures of Potential Self Interest Factor 1a Factor 2 Factor 3 IMP Potential Effect 0.74382 0.10928 0.17110 IMP Potential Prices 0.77436 0.17459 0.04398 IMP Potential Job 0.67242 -0.05353 -0.18478 IMP Potential Satisfaction 0.47864 —0.20084 -0.01499 Pr0portion of variance accmmted for by factor 92.6% 5.6% 1.7% athe chosen factor Table 14A. Pearson Product-Moment Correlations among the Component Measuresa IMP Poten- IMP Poten- IMP Poten- IMP Potential tial Effect tial Prices tial Job Satisfaction IMP Potential Effect 1.00000 IMP Potential Prices 0.60331 1.00000 IMP Potential Job 0.46226 0.50373 1.00000 IMP Potential Satisfaction 0 . 33191 0 . 33445 0 . 33566 1 . 00000 acorrelations of i .13 or greater are significant with n=253, p<.05. Table 14B. Variable Cammmalities , Factor Purities and Factor Score Coefficients for the Overall Index 2 a Factor Scoreb Variable Commnality (h ) Factor Putity Coefficient 1MP Potential Effect 0.59448 .93066 0.33675 IMP Potential Prices 0.63205 .94870 0.38416 IMP Potential Job 0.48916 .92431 0.27411 1MP Potential Satisfaction 0 . 26966 . 84955 0 . 14638 2 h,a,b-seeTablelZB The resulting index Of potential self interest ranged from -l.739 to +.745, with mean zero and standard deviation .896. 6(5 .Omw.~ cowumfi>oo ouOOSOUm one who.| some nufi3 mm~.m+ ou Hem.mi Eouw ommcmu .oco oefiu .ucoecomoefia uom xmocw awakened Hauuo>o och .oma. coHuMfl>0m ouoocmum ocm oumu come ouMS .mvn.+ o» mew.ai Eouu oomcmu ummumucw OOMHME Hewucmuom mo xoocw mcfiuazmou one mmw manna men n n .e . n N ommmo.o hammm. memN.o meB muwnuo Hawucouom mzH mmmma.o mmowm. vhmnm.o xaca xuoz HMflucwuom mzH aomwm.o maomm. hNth.o xHMB mm>fiumawm Hmwucmuom mZH mammm.o mnmom. mmwwm.c xama mocowuh Hmwucwuom ASH nucoaowuwmoo ouoom uouomm maueusm uouomm Luce hoaaocssfioo cancaums xoocH Hamuo>o ecu now mucowowuwoou ouoom uouomm can moauwusm uouomm .mofiuwHMSOEEoo manmwum> .mmH manna .mo.vm .mmmuc zuw3 ucmowwwcofin mum kumoum Ho ma. H mo mGOwumHouuoo m ooooo.~ mommm.o mwmmm.o aom~v.o same mumnuo Huflucmuoa mzH ooooo.a aemmm.o mqaom.o xame xuoz Hofiucmuoa mzH ooooo.~ mmoam.o game mm>numamm Hmfluamuoa mzH ooooo.H xama «oceans Hmeucmuoa mzH xflme umnuo xama xmoz xame mm>fiumHmm game mecmfium Huflucmuom mzH Hmaucmuoa mzH Hmflucmuoa azH Huflucmuoa mzH mousmmoz ucocomEoo on» macaw mcofiumaouuoo ucoeozluosooum comumom .cma mHnma comoco uouooM ocum wh.o av.¢ am.ve uouomw xn u0w poundooom oocmwum> mo coauuoeonm mmaao.o moth.ot mmoom.o XHOB muocuo HOflucwuom mZH omvHH.OI mommo.ot hNNVb.o xamfi xuo3 Hafiucwuom ASH HmmNH.o momma.o oomvm.o xHMB mw>fiumamm Hmwucmuom ASH mmvoo.OI vmmha.o mmmaa.o xHMB nocwwuh Hmwucwuom QZH n uouomh N uouomm OH Houomm umouwucH smeawz Hmaucouom mo mousmmoz usom on» you xwuumz uouome .ma wanna EmmmmazH DmHAHE A mo managed 83% 93 OH mediums .Ez .80 mass .. NEE Ea bgfi .8388 .Saammummm mamas .8 £an. mommm . o mmoea . o coflmosom umououcH EH umwuoucH Hmz 96 THE. £8,395 3890 cam coflcucom gem n Page 2&3 3nt sammm.mn Aueaumeaov mma.aa ammma.o ~mmm~.o ~mmom.o coeuausem ham.a~ memeo.o mmmmm.o omaom.o ummumucH qmz m m gonna new aumm m mannebm> eammm.mu Auqmumgbov Nmmmm.o ~e~om.o memmm.o mmmao.o mmeoa.o ammoa.o coaumosem ammmm.o omamm.o oomam.o momHH.o momaa.o oomam.o ummumueH EEH aumm m m maesem mmqmeu.omm ,muasem_m Am macmufigz l7 mange Numeesm cowumosom ummumucH mavea.m ammmo.m uonum bnaeamum ezH memoa.m~ eammm.omm mmema.o mumswm mm. b 635m :82 w _ mamoad m medias _ bananas maomm.mm- .omm amueammm ammme.oee .m aoammmummm mmmwazoeMIEZH wabaenm> unmecmmmc mmHmswm mo 85m mo contender mo Enhance 83% m5 5 moanmflms bemeeommaeH .meo mega I mmeoazoax ace ummumneH .eoauauaem .eoamambmam maaauaazr .mm mange 79 monmflwm mo Sm mo woqmfium> no man? memee.~u Aneaumeouc amm.~a aHHo~.o mmeo~.o ameoe.o scenausem mma.~m mmeea.o memmm.o meaba.o ummumueH umz a m bonus cum 8.8 m 633.3, memee.mt luqaumeoov mmeo~.o emeoe.o eaamm.o ~e~ao.o mmmea.o Hmoaa.o eoeuausem memmm.o meaea.o mammm.o maema.o maema.o mammm.o ummumueH qmz Dom m m 395m 69MB 0mm magnum m m made? A, magma g eoeuaoenm ummumueH mmmee.aa Hmamm.m Henna enaecaum amz maeem.am maama.emm mmmma.o mumsvm m e muasem game. v. _aaoaa.o m macabaaz _ mabaenm> aeaao.aamm .oam Haaeammm mamoa.aae .m eoemmwnmmm mmeoazoex oez manmmnm>_uemeemamo some; 93 5 Evacuate“ .Ez . 05. T59 I mmooazosm one ummuoucH £0quch . coflmmohmmm waged: . mm magma. 80 mmth.o Hmmho.o cowumocom umonoucH mzH ummumucH ESH 039 mafia .ummnmch HHmum>o_ocm coeumosom comsuom H m.comnmom .ccm oHnmE mommH.~I Aucmumcoov omo.HH mehH.o HommH.o mHmhm.o cOHumosom mmm.mm hemeH.o wommm.o mHmmm.o unencuCH mzH m m .65 Em Dom m wannabes mmmmH.NI Hucmumcoov HommH.o NHmnm.o woomm.o mHhmo.o mmomH.o mmmmv.o cOHumosom vommm.o mHmmm.o amHmm.o HHmmH.o HHmmH.o mmHmm.o amonmucH EEH 3mm m m 395m mmcwfi 0mm 883 m m 336.32 A. mHnt mnmeesm COHumocom DmmumucH mmmmo.OH mmoem.m “chum oncocmum mzH OHHom.mm mhmHh.Hom luv mmomH.o mumccm_m m ONE 59 w _ mmemad m mamfldsn 633.5, «Home.nmmm .Omm Hmcowmmm SEEMS .N 83% mammaeblnm ea 30.35,. 2338 mmumcmm HOissm mo mocmHHm> mo mawwamcm cOHumcmm ecu OH mOHQMHHm> eggs .95. 9&9 .. 683399 Ea papaya .coflaosem .83me «Heidi . om Esme 81 Hypothesis 4 - The fourth main hypothesis was stated as follows: The correlation between overall interest and knowledge will be higher than the cor- relation between education and knowledge. The results Obtained (Table 31) supported this hy- ' * potheSls (rint > redu throughout). Table 31. Correlational Data for both Events at Time One and Two: Education, Interest and Knowledge NFL Time One Beta Standard Error B F r. 0.317 0.282 0.090 23.267 lnt r 0.287 0.247 0.190 17.915 edu NFL Time Two 0.355 0.961 0.167 32.923 r. lnt r 0.234 0.706 0.201 12.334 edu IMP Time One r. 0.345 0.292 0.073 24.317 int r edu 0.293 0.225 0.158 14.485 IMP Time Two r. 0.391 0.356 0.146 36.359 lnt r 0.260 0.196 0.174 11.020 edu * r. = Pearson's r between overall interest and int knowledge. r = Pearson's r between education and knowledge. edu 82 II. Specific Hypotheses Hypothesis la - The first specific hypothesis was stated as follows: Education will correlate with both factual and structural knowledge. The results obtained for both events at time one (Table 32) and time two (Table 33) supported the hypothesis. Furthermore, for the purposes Of comparison we ran comparable correlations, using overall interest (Table 32A, 33A). Also a pattern which seems to emerge deserves note here (Table 34). While both education and interest correlated with factual and structural knowledge, there was a definite contrast in the correlational pattern Of these variables; education seems to correlate a bit better with factual knowl- edge, while interest goes with structural knowledge. As can be seen in Table 34, the r's between interest and str. knowl— edge are in each case significantly higher than those between education and str. knowledge, whereas the interest/factual knowledge and education/factual knowledge differences are never very large. This should not present a great surprise if we assume that educational attributes would facilitate recall, while interest in an event may enhance the effort to understand it. Future work should definitely incorporate elements which allow a closer look at this interesting possibility. 83 Hoo.onm Hoo.onm Hoo.oum Noo.oum Hoo.onm Amvm v Amem v Amvm v Amen v Ho V hmmm.o mvH~.o mmmm.o hmmH.o oooo.H cOHumocom omomHsocM mmOOHsocm mmomHsocM mmOOHzocm coHumusom Houduocuum EEH Hmcuomm EZH Housuoouum qmz Hmsuomm_qmz 88H; 383m one Hospomm .COHumusom "036 OEHB um mucm>m anon MOM mucwwofimmmoo coauOHouuoo comumom .mm oHnma Hoo.onm Hoo.onm Hoo.oum Ammm v Ammm v Ammm V Hmmm.o omm~.o Hmmm.o ummnoucH EZH Hoo.oum Hoo.onm MHo.Oum Ammm v Ammm v Ammm V ommm.o h~e~.O moeH.o ummHoucH qmz mmOOHzocM mmOOHsocM mmOOHsocM mmOOHsocM QOHumocom Honsuocnum EZH HmsuommemzH Hmucuosuum qmz Hmcuomm omz $838M Haguosfim ocm Hmsuomm .cOHumosom "moo OEHB um mucm>m coon mom wucmHOHmmooo coepmHmHHoo commune ecmm manB Hoo.onm Hoo.onm Hoo.oum Hoo.onm Hoo.oum 8mm v 8mm v Ammm V 8mm v 8 V mmmm.o omm~.o memH.o mmHm.o OOOO.H coHumosom omOOHsocM mmOOHsocM mmOOHzocM mmooasocx coeumosom Hmucuocuum EZH Hmsuomm EZH Hmucuocnum qmz Hmsuomm.Hez 88Eo§ Eagm 98 838m . 838:3 65 95. um 85% 8.8. nob flfiaoflmwoo 833880 commune . mm wanes 84 m Haoq. ,ll/‘xxx mmmm. Hmmm. ///<\\\.ommm. mu yuan hmmm. momm. mmmm. m¢ma. Hum swmu mmnm. ,///\\\\ vomm. ommm. _\\\)/// nmvm. 0mm yuan swam. hmma. ommm. «mam. emu scmu E HE S E m mafia H_m2fia mmvmdzaqm amusuusuum cam Hmsuomm nUfiz.ummHmu:H cam GOHumoncm Mom cumuumm HMCOHuMHwHHOO .vm mant AmocmoHMHcmHm\AmmmmuV\ucmfloflmmmouV Hoo.onm Hoo.onw ~oo.onm Amvm V Amvm V Amvm V Haov.o mmnm.o mmha.o ummHQuCH mzH Hoo.onm Hoo.onm moa.onm Amvm V Am¢m V Amwm V mmmm.o . vmmm.o mmho.o ummumch gmz mmomazocm mmwwazocx mmuoazocx mwowazocx gowumoscm amusuosuum mZH Hmsuomm mzH amuspusuum qmz Hmsuomm qmz mmuwazocx amnsuusuum new Hmsuomm .QOHumosnm "038 mEHB um mucm>m anon Hem mucwaoammmoo COHuMHmHHOO comummm g¢mm manna 85 Hypotheses 2a - c - The next specific hypotheses were stated as follows: H Interest stemming from perceived utility to self will be a stronger predictor for both factual and structural knowledge than interest stemming from perceived utility to milieu. 2a: 2b: Interest stemming from perceived potential utility to self will cor- relate higher with both structural and factual knowledge than will potential utility to milieu. The correlation between interest stemming from immediate utilities and knowledge will be higher than the cor- relation between potential utilities and knowledge. 2c: The results we obtained (Tables 35-42) did not lend support for the above three hypotheses. Specifically, milieu interest emerged as the stronger predictor of both factual and structural knowledge, rather than self interest, as expected according to Hypotheses 2a and 2b. Also, while immediate milieu utility did emerge as the strongest pre- dictor, the overall pattern expected according to Hypothesis 2c failed to emerge. We ran an additional analysis using overall knowledge instead of factual and structural knowledge separately (Tables 43-44), and found that the dominance of milieu interest persisted. The following summary Figure 7 may facilitate a quick check of the results mentioned above: Time One Time Time One Time Two Figure 7. 86 National Football Strike Knowledge Factual Structural Milieu Milieu Potential Self Potential Milieu Self Potential Milieu Potential Self Self Milieu Milieu Self Self Impeachment Knowledge Factual Structural Milieu Potential Milieu Potential Milieu Potential Self Self Milieu Self Milieu Potential Self Self Milieu Self Overall Milieu Potential Milieu Potential Self Self Milieu Self Overall Potential Milieu Milieu Potential Self Self Milieu Self Summary of Interest Components as Pre- dictors of Structural, Factual and Overall Knowledge. One of the possible implications stemming from these findings regards the nature of information made available through the media; e.g. presently it appears that the interest component stemming from perceived utility to self was not acti- vated as a predictor of knowledge. Further work should incor- porate even more diverse publicized tOpics in an effort to un- cover kinds of information that would activate self interest. 87 oaooo.o voo.o wwmma.o mmwoo.o Hmmao.o mmm.o qmmma.o nnmmo.o hhwva.o mo. v mmm.m «Homa.o moama.ot mmHHm.on Hoo. y mmh.av mmama.o Hommv.o vavo.a a 93 em m uoflm 3m 38 m cowumqmm mcu ca mmHQMHHm> oaooo.o movoo.o HmmHo.o moomo.o Hoooo.o moaom.o mmmv¢.o wnmmo.o nnvva.o womam.o mmmoo.o voHom.o mmmvv.o moama.ot mmHHM.OI mmaeo.on mmoao.o nmmma.o mmmvv.o Ho~m¢.o mvvvo.a mmmmv.o mvnma.o mvhma.o mmmm¢.o 3mm m m 393 magma 0mm 88:3 m mama»? manna mwmeanm 8835 gnome 9.2 35988 $835 .38 9% $335 832 83:38 amz ummuBfi 3% H3558 HE $885 852 .Ez 388$ 353209 3335 «me .Ez ummnmufi 332 Hmflcmuom .Ez ummumufi 3mm Hmficwuom HE 3835 93:: dz flange, 338.5 ucwecmmmo m8 mfie u $88080 ”6835 no Snmmmummm 3.532 “mmeflzoé 330mm dz .mm manna 88 Hmooo.o hmth.o mnmmo.o oHNmH.o NmHhH.o mmmmo.0I ommam.ot mmmva.o Hmaha.o mmnnm.o oomva.o mmmvm.o vommh.o m HOHHm gm 6&8 m QOHuMJmm may CH mmHQMflHm> mbmoo.o mHNmH.o mvaH.o mvmom.o amnmw.o mmmmo.o: owmam.ot mmm¢o.o manom.o Hammv.o Hmaba.o mmnnm.o mmmam.o hmvom.o mmamv.o mmmvm.o wommh.o mmmmv.o mmvma.o mmmmv.o m 893 85.6 08. 888 m m 2833: manmeemwmsasm 353808 3.0.835 3mm 82 8835 38 33838 .82 8838 5832 33:38 RE 388.5 833 dz 3838 383.808 38.838 3mm .82 388.8 38 H3838 82 3838 5332 H3538 .32 $838 832 .32 3838 888299 33383...“ 82 2835, 3:888 95 83 I 388980 888.5 m0 203888 38332 683399 333m .32 . mm flame 89 mmmao.o mmo.o mvmma.o meaao.o moeeo.o nam.a mmmma.o moamo.o HmHmH.o mmm.o ooama.o ovveo.o nmema.o mo. v mmm.~ smmma.o mmoma.o mmmm~.o m mam «m m uouum cum mumm m coaumnmm we» c8 mmHnMflum> mmmao.o mhmao.o monvo.o moama.o. mmooo.o ommmo.o ommm~.o mmamo.o Hmama.o nomma.o mmmoo.o eammo.o mmqm~.o oweso.o «mama.o mmmma.o smmoo.o memqo.o m¢-~.o ~mo~a.o mmmm~.o omoo~.o omowo.o omoqo.o omoo~.o mumm m m magaflm mmcmgu 0mm mumsem_m m mamfluagz mange mumaasm mmwmazocx 33mm mZH AucmumcpoV ummumucH «How 82H 3888 :83: 3388 “EH ummumch mamm Hafiucmuom 82H ummnmucH smflafiz_82H 8838 iucmumgpuV ummumch mama 82H 8888 58:: 33:88 82H ummumucH mama Hmflucmuom 82H 3885 :83: ea magmaum> wHQMHHg 8898me 86 839 I 38:00:00 pmmumufi wo gflmmmuwmm manna; N 096392 Hannah ugmnwfi .hm manna. 90 mmmao.o mam.o mmoaa.o amaao.o mnemo.o ao. ama.m mamma.o aaama.o oaaa~.o am~.~ mmmaa.o aasaa.o mmmmm.o aoo. ma>.a mamaa.o samaa.o avasm.o m mam an m uoaam apa mama m coaumnma may ca amanmaum> mmmao.o mmavo.o mssmo.o aaeom.o moaoo.o aamma.o amaam.o avama.o oaaa~.o aamam.o movao.o naama.o vamam.o aasaa.o mmma~.o amav~.o mvamo.o manaa.o mmmam.o smaaa.o avasm.o maaom.o mmmmo.o mammo.o mamom.o mama m _m mammaa macmzu_oam mumsma_m Hm maaauanz magma gm mmcmazocm amasuoafium mzH AacmuacaoV uamampca mama 82a 8885 smaaaz mza uamamuca mama amaacmpom mza uamamuca smaaaz.amaacmuom 82a manmaum> aanmaacooV uamumuca mama mza uamumuca smaaazwmza uamumuca mama amaucmuom 82a uamnmuca smaaaz amaucmuom mza magmaum> magmaamaw #598me m8 manna. I 35:09.00 ummamuca mo gammmummm may; “mmcmazocm amaafioshum 38.53% . mm magma. 91 mmooo.o vma.o mmama.o mvmmo.oI «ammo.oI aoo. v ma¢.vm mmmma.o mmvam.o vmmmh.o 8 88mm m 88.8 8a 88 a coaumswm mcu ca mmanmaum> mmooo.o mvmmo.OI vmmmo.OI moomo.o swooo.o ~mmmo.o mmmom.o mmvam.o vmmmn.o nwmom.o mammo.o mammo.o bemom.o 88 m 8 898a 885 08 888 m m 8833: magma MHMBan 88308 888 82 3888V 8888 mama 82 8888 838 82 8888 38888 8888 mama .amz ummenEH DOHHHE .HmZ manm3m> mammaam> ucmccmmmo 039 mEHB I mucmcomeoo ummamuca mo coaammnmmm mamauagz “mmcmazoqx Hmsuomm amz. .mm manna 92 omooo.o a am . o «mama . o 8va .0 33a .o aoo .v mom . am «$36 Ram: mammad m oa~.~a m uouum qua mama m scammnma mam ca amanmaum> omooo.o vnmao.o vmvaa.o ammaa.o anaoo.o mmuam.o oaaa¢.o oumaa.o ammma.a aava¢.o namam.o namam.o aavav.o mmmm a m mamaaa mmcmzuvoam mnmsma_m m mamaudaa manna gm 88.392 888m :22 AHCBQHOUV ummHOuGH mamw 1E2 ”5835” SOHHE .Hmz 888.5 AugmpacooV 3888 mama 82 mamumuca smaaaz_amz magmaum> magmaum> acmucmmma 039 meae I aucmcomapo mamamuca «0 coaaamuamm mamamasz.“mmmmazocm amusuusuma amz .ov magma 93 aaaoo.OI mmm.o wmhma.o o~¢mo.o moama.o aoo. v www.ma mmama.o muchm.o onwaw.o a 8.0.8.8 m 888 8a 8mm m coaumswm may ca amanmaam> aaaoo.OI omvmo.o mmama.o mmbma.o ~mmoo.o mmomo.o maaom.o mnonm.o onvam.o mmmmm.o hammo.o hammo.o mmmmm.o 30m m m 88a 885 08 888a m m 8838 manB.MHmmmww 88308 888 82a AuamumgQuV umwhmucH mawm g 8888 83a: .8 8838 38888 8888 mama 82a 8888 838 8a 8838 8838 38888 03H. Q59 I 38:09:00 8835 no couammmnmmm $3.332 “ggag 8.5on ugmmnwfi .Hv wanna 94 mmooo.on Hoe. v nmm.n onmma.o Hmmna.o maomv.o Hoe. v www.mm mmoma.o mmnom.o hmo~5.o m owm.~m m Hanna cum mumm m 83% 0:.» 5n mmangbg mmooo.ou Hmmna.o mam~¢.o moam~.o mummo.o mmmma.o monov.o mmnom.o hmomh.o mavnm.o mmmma.o mmmma.o mawnm.o muwm m .m mamamm mmnmgm 0mm mumswm_m m manquasz manna. gm mmwmazocx amusuusuum mZH Auamumgpuv ummumucH mama mzH “magnum DmHHE mEH magmaum> Aucmumcpuv ummumch mamm mZH ummumucH :mflaflz_mzH manmflgm> manmflum> ucmccwmma 03H. 95. .. $5898 3335 no scammmummm 3&3: 683g amusuufium #553? . mv magma 95 Amoco.o .m.c omm.o mmmmm.o mmmmo.o nm¢oa.o mo. v omn.~ mmmam.o Hmmaa.ou mvomm.ou Ho. v mwm.m mmmm~.o mm¢ma.o mammm.o Hoe. v mqm.m¢ mman~.o nmnmv.o mwnmm.a m mqm «m m woman cum mumm m c0393 man 2.." mmanmflumxw Hmoo0.o mmmmo.o nmwma.o wnmoa.o mnooo.o mpwvm.o whvm«.o HmmHH.ou mwomm.on mmaoo.ou mmmoo.o ooqv~.o nmmmv.o mm¢ma.o mammm.o H~¢m~.o mnnoo.o namm~.o ¢m¢m¢.o nmnmv.o mvnmm.a mmmbv.o H¢nmm.o an-.o mmmhe.o mumm m m mamaflm mmcmgu 0mm mgmsvm_m m mamauasz 3nt gm hmamo.o .m.c mH~.o mooam.o mwmmo.o mpvva.o .m.c sho.~ maomm.o mmmaa.o mamav.o Ho. v hm~.¢ omnm~.o haama.o mmmmm.o Hoe. v m-.m mmmwm.o mumoa.o moohm.o m mvm Mm m ~85. Bm 38 m caaumswm mcu ca mmanmflum> smamo.o mvmmo.o mnwwa.o mmoom.o Buooo.o ¢vmma.o navmm.o mmmaa.o mamav.o ommmm.o mmmao.o nmvma.o mommm.o naama.o mmmmm.o mmnmm.o ¢mq~o.o momoa.o mmomm.o mumma.o moonm.o maomm.o omwmo.o omqmo.o mHom~.o mamm m .m mamaflm mmcmgu 0mm mumswm.m _m mamfluasz 3nt gm mmcmazoam mzH Aucmumgouv $835 38 .Ez ummnmucH mamm Hmapcmuom gmz $335 8:2 dmflcwuom gmz ummumucH smflaflz_qmz manmflg Aucmumgpuv ummumuqH «Ham qmz pmmnmuCH uamm Hmflucwuom qmz ummnmch smHHH2.HMHucmuom gmz ammumuaH :mqaflz_qmz mHnMflum> mmomdsocx qmz manmflum> ucmwcwmmn 3.5202 ammumch mamm mZH ummnmucH mamm Hmaucwuom mZH $885 832 .5 ummnmch smflauz Hmaucmpom QZH mflfiflfiw Aucmumccov ammumch mamm mZH ummumch mamm Hmaucmpom mZH ummumucH :mHHHz_mzH ummumucu smaaflz.HMflucmuom QZH m~nmanm> magmanm> gcmocwmmn mco mafia. I 35.8950 umwumucH mo scammmnmg «Hang 323mg cam _Hmz .mv manna. 96 whooo.o .m.c mvo.o mmom~.o mvmao.o oommo.o Hoe. mna.om nmem~.o nmsmw.o mnwwm.a m oqm.~m m.uougm cum mumm m c033 may 5 mmanmwumxr whooo.o mvmao.o oommo.o wnwma.o «Hooo.o mumma.o moflmw.o pmnmv.o ~n¢nm.a mmomv.o mmmma.o mmmma.o mmomv.o 38 m a 03:3 mag 0mm 98:3 m m «3332 0.309 gm mmcwazocm qmz Homoo.ou Ho. mmm.v mhmo~.o mmhma.o mmnbm.o Hoo. mnm.mm mmmm~.o whamm.o mma¢m.a m o¢m.~m m 000mm cum mumm m coflummwm 0:0 ca mmanmflum> Homoo.ou mmnma.o mmnnm.o Ho¢m~.o HHmHo.o mmmma.o mm¢o¢.o mnamm.o mmavm.a mawmm.o mmuwa.c mmhea.o ma¢mm.o 38 m m mammarm mag 0mm 895m m m 3033 0.33. gm mmumazocm mZH Aucmumcoov uwmumucH mamm mzH ummnmucH smfladz mZH magmaum> Aucmumgpuv ummumucH mama 02H $835 93$: EH gangs manmflum> ucmucmmma Aucmumccov ummumch mamm mZH 08038 53:: g 2035, “unnumgpuv uwmumch mamm mZH ummnmucH smflaflz.mzH flamflg manmflnm> unaccmmmo 05. mag. I 30000960 #0835 m0 033% mamfladz 3g; cam. Hmz . iv magma 97 Hypothesis 4a — Specific Hypothesis 4a was stated as follows: H4a: Interest stemming from utility to self Will correlate higher than will education with both structural and factual knowledge. The results obtained for both events at time one (Tables 45-46) and time two (Tables 47-48) did not provide support for the above hypothesis. Taken alone, the self interest component did not match the predictive power of education for both factual and structural knowledge. This result is consistent with the outcome shown for specific Hypotheses 2a-c discussed before. It is interesting to note, however, that for both events the contrast in favor of edu- cation was more strongly marked with respect to factual knowledge, less so with respect to structural knowledge. This pattern appears to be consistent with the observations made in discussing specific Hypothesis la (see Table 34). th.~ Hom.mH m Nwm.o mao.mN 0mm mm 98 mmmne.au Aucmumcoov mmmva.o mmaoa.o m-«~.o moama.o mmoao.o mmnno.o momn~.o ummumucH mamm mzH mmhmo.o Ho~v~.o mommm.o vommm.o oanmo.o oaamo.o vomm~.o :Onumosom m uonum bum mumm m m mamaam mwcmnu 0mm mumswm_m _m mamfluagz mHQMfium> magma g $0030ch Quorum % manmfifig ucmpcwmwo emmmm.mu Aucmumcoov evama.o vnmvo.o oaoma.o moomo.o momoo.o Hosea.o ommmm.o ummumch mamm gmz mmNHH.o manam.o hmnmm.o mmmam.o ~moao.o maaoa.o mmmam.o coflumunum m uohm 3m 38 m m 305m 985 0mm 890m m m mflfldfl 0335/ magma gm magmazacm Hangman qmz manmnum> ucmnammmo mac mane I mmuflfié 336mm no muBonum mm ammumufi 3mm 05.. coflmBB 38.50% 05 dz . ms magma meam.o NmN¢H.o mwmmm.o Hm¢ma.o m HOHHH Gum 339988 mmmmN.Hl mmmm¢.o mvhom.o ommmm.o ommmm.o m HOHHm.UHm mmbao.o Hammo.o mvmmm.o ummumch mamm.Haz mmnmo.o mmhmo.o Haema.o coaumusom m 39am mango 0mm gnaw m m 03332 3035/ magma Nam—55m magmazoé fifigm 9% 20.3.3, 0808me 308.208 Hummo.o mmmmo.o mmhm~.o ummumucH “How mZH mommo.o mmmmo.o ommm~.o coaumozom m 0.35m magma 0mm madam m m 3&3? QBMHHQ/ 99 manna mHmEEDm mwcmazocm Handpusnum QEH mHQMHHm> acmwcwmmo 08 95. I mmomazosa Hmnsuusfim no $80603 mm 3835. 30m 05 838:8 3ng can .22 .3 magma 100 wmm.m mom.m mmm.a oam.v ovN.Nm omgm . HI 3:00.909 mmoma.o mamma.o mmmm~.o mmnma.o mmmao.o mmmmo.o ¢m~m~.o ummumucH mamm mZH ammoa.o ommmfl.o ma:~m.o omom~.o mmmwo.o mmmvo.o omomm.o coflumosum m uoHHm cum mumm m _m magawm wmamgwomm mumsvm.m _m mHmHquz. manmnum> magma. gm 0903305” Hmzuomm g 3 amum> #:096me nmwmm.cu Aucmumcpov Hemma.o mm:no.o Hmmma.o moomo.o mamoo.o mammo.o nmmma.o ummnmuaH mama qmz mamaa.o avova.o mmaw~.o ovema.o momao.o momao.o oeemfl.o coaumusnm m “Bum 3m 8.8 m m 293 mango ow: 889$ m m mafia; mammflg magma mums-5m monoazoam Hmsuomm gmz mHHMflHm> unaccmmmo ggfil mmcwazacx Hmsuomm :0 mucuOHumum mm ummumuaH mamm new aoflpmosnm “unmanummmaH cam gmz .mv magma 101 Nva.m vam.m omv.h mmm.wa ovm mm Amcmmmcoov ammummaH mamm qmz :OUMQDB thmo.o mammo.o mmmwc.o mammo.o mmmmm.o bmmha.o maamnu 0mm mnm:am.m m mamflmanz manmflnm> mmma¢.HI Haema.o mmmna.o mmmov.o mmmoa.o mmoma.o mmvma.o mnenm.o nmmna.o m Hoaum cum mumm m .m mamsflm vmmoHIHI manoa.o «Hosa.o mammm.o mmmm~.o ommma.o mmmm~.o mmmmm.o moamm.c mugfiufi 38 m .m3&@ manmE_MHmbsaw mmmmazocm amusuusumm gmz mammanm> mammcmmmo Auammmcoov monmo.o mmmaa.o mmmmm.o aoaumusmm mnemo.o mavwo.o mon~.o ammumch mama mZH 00:90 0mm gm m m mwmfldfi manmflum> magma mwmfigsm mmomazoax amusuusuum mZH manmflnm> ucmocmmmo Hmuduunfium mo mnouoflowum mm 380ch 30m mam 8.30058 33:62 EB .52 039 mega I mmcmagoaz .me magma 102 Hypothesis 4b - Specific Hypothesis 4b was stated as follows: H Interest stemming from utility to milieu will correlate higher than will education with both structural and factual knowledge. 4b: The results obtained for both events at time one (Tables 49—50) provided only partial support for the above hypothesis. That is, in the case of factual knowledge about impeachment, education did better as a predictor compared to milieu interest. For the other two comparisons, interest was more highly correlated than education, e.g. in the case of structural knowledge, the obtained pattern was as could be expected, emphasizing milieu interest. Again, this out- come is consistent with previous note on the tendency of interest to enhance structural knowledge (e.g. understanding) and for education to enhance recall of factual information. The results obtained for both events at time two (Tables 51-52) did provide support for the above hypothesis. 103 emamm . .T. 35809 8mg“. 393.0 mmofld gamma omoomé 2385 823.0 magma umflfifi 832 ea mama 2805 mommmd «.285 «8&5 02.85 Samoa gamma 83% m m “ohm Bm 3mm m m 395m mango 0mm wasam m m mafia: 30.2.5, manna. gm mmcmHEcm 38mm as $335, ”regime «.88.? 3:38.08 soa.ma Hamoa.o Hmmm~.o mmamv.o mmmam.o mmvmo.o mm~¢~.o amma¢.o coaumuqum $18 082.0 Same 239° mammsd 92:5 92:5 mmNQd ummHBfi 53H: 9.2 m m “ohm 3m 38 m m mam—5m mmqmfiomm gm mmamflufifi manmwums manna g 363g Hmsubmm .Hmz manmwnflw u§cwmmo 96%| magma; gown m8 flflbfiflm mm 5835 53:2 can 83% 355% can .32 . 3 Same 104 Ham.m mma.ma m wam.m ham.mv m Hmomo.Hn Aucmumgpov mmmmo.o vamH.o Homm~.o ommm~.o mmamo.o mmmHH.o mawmm.c coaumosnm maama.o hmm¢~.o omnmm.o mmmm~.o msflmo.o mnamo.o mmmm~.o ummumucH smaaflz_mzH m Hone Rm 8.8 m m 29mm 8&6 0mm 8% m m mafia; 303.5,. 3nt gm mmaoazocx amusuusuum mzH mannaum> ucmoawmma nmmou.on Auamumgpov nammo.o maoaa.o Humma.o Haema.o HmHHo.o omoma.o mmmv¢.o aoaumoaum ~moma.o vomo¢.o mamam.o mmm~¢.o mmvma.o mmqma.o mmmm¢.o ummnmucH smaaaz.gmz m uoaam_cum mumm m _m maaamm mommao 0mm mumsmm_m‘ .m meHuazz manuaum> magma g mmnmazocx amusuusuum amz mannaum> ucmcammmo Huggm mo mnouoflcmum mm. ummumufi 503.“: can saunas 355% cam .32 mac mess I mmuwazocx .om magma 105 mmm.m mmwdm m oma.m mam .mm .m $5.3” . .7 Augmcouv e836 «dado Emamd omommd H386 mmmmad Hdmmd 838.3 mommaé SE3. smmmod mmmmmd 259° 259° mmmmmé 6835 55: as m .85 Em 8.8 m m 395% magma 0mm mg m m maids 39.2.5 magma Mamba—4m mmcflg 38mm .5 manage, €86me 2ST? 35809 mmmdd 28:. mmmomd 21.2.0 8:06 88:. 852. Eflmnfim 32:5 ommmmd $55 283. mammod mammod Swomd 3335 :33: .52 m “chm 3m 3mm m m 395m mango 0mm @3an m m mafia": manage, magma gm @8392 153mm emz 30.3.3, usmciwn 03H. $5.8 I $6330ch Hmouomm mo mnouoflmfim mm ummHmuCH Donn": new cog—mung figmmnmfi cam .32 . Hm manna 106 Nvm . m mvo . mm .m wamé mmndw .m 933.7 fiqfimcoov 3.810 mommaé Hmommd mmmmmd .3386 5.336 «p.36 GOUMUDB mmmmad mmwmmé vammmd mavhmd mmmmaé mmmmad N356 ummumufi 53.32 g m HOPE 3m 3mm m m 39am momma 0mm 38$ m m 39.33: manmflg manna. gm mmpwazbcx Honouoaum mzH manmflfig panama Elmo .T Aufiumcoov mmmoa.o amava.o mm~s~.o hmmha.o Hmmao.o hmmmm.c mommv.o aoaumosum 3535 RS"; mmafié mgmvd Smamé Sflmd mugs ummumpfi 82H: .22 m Bum 3m 38 m m 295m magma 0mm 83$ m m mafia: $03.5, manna gm mommazoé amusgm dz 3035.5 ucmocwmmo 0%. mfie .. wacflg nguushum mo muouofiumum mm ummnmufi swung cam sojounfim “ugmmgfi can 1.52 . mm 3..nt 107 Hypothesis 4c - Specific Hypothesis 4c was stated as follows: H40: Interest stemming from potential utility to self or milieu will correlate higher than will educa— tion with both structural and factual knowledge. The results obtained with regard to potential self interest (Tables 53-54) did not provide support for the above hypothesis, with an exception of structural knowledge about impeachment (Table 54). The results obtained with regard to potential milieu interest (Tables 55-56) provided only partial support for the hypothesis; as could be expected, potential milieu interest was a better predictor in the case of structural knowledge about both events. 108 vhHmN.HI Aucmumcoov mm.o mmeva.o vmmqo.o momoa.o mmmwo.o momoo.o moowo.o HHoo~.o ummumucH mamm Hmfluampom.aaz mm.m vamoa.o ommma.o mmwmm.o Hmva.o mmnmo.o mmnmo.o Hmva.o aoaumuzom m m “ohm 9% 38 m m 395m mound 08 88$ m m maids @333, manna gm mmumazoqm amusuosuum qmz mannaum> unaccmmmo oomvm . ml Augmgov mm.o mavma.o sasmo.ou mommo.ou mmavo.ou mmaoo.o ommoa.o ovamm.o ummumucH mama Hmauawpom gmz mm.m~ mmNHH.o samam.o mmoom.o mmmam.o mmaoa.o mmaoa.o mmmam.o coaumuscm m m Hahn Em 38 m m 29%. magma omm 83mm m m mamflasz m3mflm> 3nt g mocmazonm Hmspomm qmz manmanm> ucuocmamo «so mafia . mmcmagocz Hmnsuusuum can sunbeam mo muouoflwmnm mm ummnmch mama Hmaucmpom can aoaumoacm “qmz .mm manna 109 mmmmH.Hu Auamumcpov «mm.aa qnvmo.o mmmom.o wmmam.o ommmm.o opovo.o mmmoa.o mmmam.o coaumoscm noo.ma moama.o mamam.o nmmnw.o Hmmv~.o mmawo.o mmamo.o Hmmvm.o ummumch mamm dQfiflhHKXHmSHs m m uouum 3m 38 m m $9me mango 0mm 383 m m maniac: m3mflm> wanna mnmhsaw mmcmazocs amusuusuum mZH manmflum> unmocmmmo momm¢.HI Aucmpmcouv mmm.m mamma.o Haema.o mmsmm.o mmmma.o HmvH0.o avamo.o mmmm~.o ummumch mama Hmflucmpom mZH mam.mH mammo.o wmav~.o ommmm.o womm~.o oahmo.o oanmo.o womm~.o coaumusum m m uoflm Em 38 m m 39% mag 0% 33$ m m 03332 30.3.5, magma anabgsm mmomazocx amsuomm mZH magmanm> unmocwmmo mco mafia u mmcwazocm amusuosuum can Hmsuomm mo muopoflomum mm ummnmpaH mamm Hmaucmuom can ccflumosum “unassummmaH .vm manna llO Hmmmm.0I mmm.w mmmoa.o mmoma.o Nmmmm.o hm¢.¢m whama.o mmmmm.o Hmamm.o m m Honum Cgm mumm m vmhoo.ml hmm.m mmwva.o NHmbH.o mavmv.o HNN.¢N vmmaa.o mmmmm.o bmmmm.o m m .HOHHM Qua 30m m 3quan Rafi . o mmmmo . 0 gm NH . o 51.3 . o 83893 93,395 53.32 83.2. p.336 mmaoaé 83nd 33:38 .32 m 394m mmcmfi 0mm 893 m m mflflfifi flags, 3nt gm mmvflg amusfishm .Ez flamflms uficfimma 35382 «Emma mmmmod H326 magma $835 832 Hmflfiuom Hmz mmmdm .0 82:6 «335 mum: .o Sigma m 395m mmqmfi 0mm 833 m m mamflaz manmflg 3nt 55m wmcflgé 38mm gmz manage, €85me Hmnsuoshum cam 390$ mo muouoflcwnm mm ummnmufi 53.32 HMHfiflpom USN 20.30033 “Hm—z m8 mfia .. wmomag . mm magma lll mmmmo.HI Auamumgouv hma.m mmvmo.o NmNmH.o mommm.o ommmm.o hammo.o mmmma.o wmvmm.o COHUMQDGE mmm.mH momma.o mmmom.o mmomm.o mmmom.o mmmmo.o mmmmo.c mmmom.o ummHmUCH DQHHHZ HMflucwuom mzH magma; E3835 ea 3035/ £88me mahmm . HI 33989 :46. 2535 9835 8:36 58.36 mBSd mmmmo.o mmmmmd $885 5332 Hmflcmuom m5 moa.¢a mnmmo.o NmHmm.o mmohm.o woam~.o oanmo.o cahmo.o vomm~.o coaumoscm m m “Bum Em 8.8 m m 39% magma 0mm 939$ m m mflfldfi magmflg magma gm mmwmagcm 38mm 9m magmflg €85me m8 9:9 .. $8399 dfiBosfim cam 350mm mo flBoHcmum mm ummumufi 332 aflucwuom can 8333 “ ufigmmgfi . mm magma 112 General Hypotheses The first general hypothesis was stated as follows: HA: As the infusion of mass media infor- mation into a social system continues, those with a higher level of interest will acquire new information faster than those less interested, so that the knowledge gap between them will tend to increase. After obtaining scores on "new" knowledge for each respon- dent at time two, we compared the mean knowledge differences between the high and low interest groups* for each event. "New" knowledge was measured on the four new knowledge items asked at time two, which reflected event deve10pments after the first contact with respondents. The results are shown in Table 57. General Hypothesis A found support in the data. The second general hypothesis was stated as follows: HB: At any point in time, then, more interested members of the media audience will display a higher level of knowledge than those less interested in a publicized event. The results for both events at time one and time two (Table 58) gave support to the above hypothesis. *Included in these groups were only those respondents who consistently remained interested, or disinterested, in the event over time one and time two. Respondents who switched, e.g. became interested or lost interest at time two, were excluded from analysis. 113 me can 3. pm Emmumufimflc u N macaw... ma Em 2. pm Emmumufi n a macaw... — u . mama SQN 3%.? an «m 985 ooo.o nmfl Hm.m . . Ema Sod 8:5 om L 980 . mmomazocx 52 m5 . - . ommd mam.m «~34- m: «m macaw ooo.o mmH mm.» 03.0 $3 384 a. L 98o wmoflzocm zwz .Ez 538m Eng 83> . Hg coflmfimo 58: 896 mo manage, Hana mo 858 a . Emccmum 385$. uBBz mug 50m mom .039 mafia um =mmomd5cm 3oz: co comflumnwgoo .3 03mm. 114 $ch #835 5H can EH: cfi £38 8. ccmc mm... x83 umcuccfi HHSc>c cfi mo uHHcm 9mg .1 _ came 98.». 335.. RH m 955 ooo.o Hem Hm.m . . mm~.o mcH.m aowm.o omH H cacao . coccHzccx mZH . . mnm.o mmm.m mhmm.ou omH m cacao ooo.o Ham cm.c . . mmm.o vmm.m tho.m mm H cacao . coccHzccx.Ham . cze msHe . . mama mama gamma- 8H m 9.80 ooo.o Hmm mc.v . . cc~.o ~oo.m mmmm.o sNH H cacao . . moccHzccm mzH . . mm~.o mcH.m ooah.H- mmH m cacao ooo.o Hmm sm.~ . . mmm.o mmc.a hmvc.H omH H cache . ccccHzccm Hmz . . cco csHa — Ir 5.303 Eoommhm 83> H .Houum coflgmo Com: mommu NO 39.35“ HHmaum «0 mcccmcc a . cucccccm cnmccmum cmcacz .8835 30H u m macaw Sconcufi EH n H 988 mucc>m nucm com cmccHzccx Emma no 203.3960 33..» .mm mHome \ I I‘ll I!" I'! In 115 Hypothesis C HC: As the publicity on a topic extends over a long period of time, the knowl- edge gap between those more and less interested will begin to decrease. Since impeachment was the longer standing event, we tested whether indeed the discrepancy in knowledge level between those more and less interested would be smaller for impeachment than for NFL; e.g., IMP d(§Hi - 2L0) < NFL d(§Hi - 3:0). The results are shown in Table 57. Hypothesis C found support in the data. Table 59. Comparisons of Discrepancy in Knowledge on NFL and Impeachment NFLd (—> IMPd t df p Time One 3.3867 1.8519 2.51 502 < .05 A V Time Two 3.9346 1.7234 2.20 482 < .05 Despite the fact that we found support for the above hy- pothesis, a few remarks are in order here. Admittedly, the test performed and reported above, was not the best way to test Hypothesis C; that is, we compared the corresponding gaps for two events, NFL strike and impeachment, which are not equivalent. We settled for this solution, though in- adequate, because the time span between our two measures was one week and deemed insufficient to allow a manifestation of the processes implied by the hypothesis. The best test should be one whereby knowledge discrepancies are compared within one and the same event, along a time continuum. 116 Yet let it be noted that the indications which the data on Table 59 give us, are encouraging. That is, the comparison of discrepancies between time one and time two for each event, while not reaching significance, seem to "move in the right direction." Thus, for the NFL strike, which is the short term event, the knowledge gap seems to be increas- ing at time two in accordance with general Hypothesis A; at the same time, the gap for impeachment developments, which were the long standing event, seems to be diminishing at time two, according to the presently discussed Hypothesis C. CHAPTER IV SUMMARY AND DISCUSSION Our main argument in this dissertation emphasized the receiver viewpoint as follows; in attending to media content, people do not engage their attention indiscrimi- nately, but rather according to some choice hierarchy which has meaning to them. This implied that the frequent prac- tice in research, of operationalizing attention as exposure, needs some refinement with respect to attention variability regarding program components. Thus, interest may direct the way mass media are used for information; given exposure, presence or absence of interest may intervene with the kind of attention given certain content areas and program components. We suggested that an interest-based model would allow sensitive examination of information gain processes and help trace the patterns of knowledge differentiation among mass media audiences. For the purposes of this study we chose a relatively narrow functional perspective. Thus, in expli- cating our notion of the independent variable, interest, we prOposed a treatment in terms of these components: 117 118 perceived information utility to self, perceived informa- tion utility to milieu and perceived potential utility to self or milieu. To the extent that an information item is seen as having one or more of these attributes, resulting interest would determine the kind of attention an individual gives to that information item. Utility to self was seen in terms of relatively im— mediate, daily c0ping behaviors related to the functioning of individuals and their home and family. Utility to milieu was seen in terms of communicative utility and facilitation related to an individual's social environment, the various membership and reference groups he is associated with (e.g., friends, relatives, fellow workers, neighbors, etc.) Poten- tial utility referred to the routine scanning of the infor- mation areas kept under survellance by the individual. The comparative emphasis on each utility attribute commanding attention will depend on the individual's short— term and long-term priorities, habits, pressures and changes in the environment. For the purposes of testing in this study we postulated an ordering whereby interest stemming from perceived utility to self took precedence over the social milieu; also relatively immediate concerns were ex- pected to prevail over delayed ones. In terms of the dependent variable, knowledge, we used a component measure differentiating between factual and structural knowledge, treating them separately in hypothesis 119 testing. Factual knowledge refers to the respondent's knowledge of specific items, names, dates, places, facts and figures, related to specific news occurrences. Struc— tural knowledge is taken as the respondent's understanding of the relationships manifested in the broader framework of related phenomena. Furthermore, we were interested in comparing the role of interest and that of education in tracing differential knowledge levels over time. The following diagram of the hypothesized relation- ships represented the synchronic part of the model (Figure 8). Thus, we formulated fourteen hypotheses which dealt with various aspects of the model. First, the model incor- porated a time dimension and the independent variable of overall interest. The following three general hypotheses were addressed to these aspects: HA: As the infusion of mass media informa- tion into a social system continues, those with a higher level of interest will acquire new information faster than those less interested, so that the knowledge gap between them will tend to increase. H : At any point in time, then, more interested members of the media audi— ence will display a higher level of knowledge than those less interested in a publicized event. H : As the publicity on a tOpic continues over a long period of time, the knowl- edge gap between those more and less interested will begin to decrease. 120 Hm K N O W L E D G E O V E R A L L (Factual and Structural) mmflnmcoflumamm omnammcuommm mo EMHmMHQ Wk A\ \ I /\ A /\ A 1‘ /\ z. 0 CA H 8 ¢, Hm U D Q- N mawm . m mamm em m m smeeez s437::_ 9 \Hfla\ mamm Z H /\ ¢~ l I I I I l I I DmflHHZ V. mamm .m .m mHDmHm. 0A suede»: ©w>flmoumm no nape Hm. 121 Furthermore, we postulated certain types of interest stemming from various perceptions of utility. We indicated the use of factual and structural knowledge as dependent measures; and we wanted to discuss in a comparative fashion education and interest as predictors of information level. The following main and derived specific hypotheses were directed accordingly: H1: There is a positive correlation between education and knowledge. Hla: Education will correlate with both factual and structural knowledge. There is a positive correlation between overall interest and knowledge. H2a: Interest stemming from perceived utility to self will be a stronger predictor for both factual and structural knowledge, than interest stemming from perceived utility to milieu. H2b: Interest stemming from perceived potential utility to self will be a stronger predictor for both fac- tual and structural knowledge, than interest stemming from perceived potential utility to milieu. Interest stemming from immediate utilities will be a stronger pre- dictor of knowledge than interest stemming from potential utilities. 2c: Education and overall interest combined will correlate more strongly with knowledge than either one taken alone. The correlation between overall interest and knowledge will be higher than the cor~ relation between education and knowledge. 122 H4a: Interest stemming from utility to self will correlate higher than will education with both structural and factual knowledge. H4b: Interest stemming from utility to milieu will correlate higher than will education with both structural and factual knowledge. H40: Interest stemming from potential utility to self or milieu will cor- relate higher than will education with both structural and factual knowledge. * 'k * Overall the findings of this study indicated that the proposed model stands on several sound foundations. At the same time, the work is not finished in terms of clarify- ing certain conceptual and methodological issues. In summary, the outcomes obtained on the 14 originally stated hypotheses were as follows: General: The more interested segments of the audience indeed picked up incoming information faster, and also at any point in time knew more than those less interested in the same event (Figure 8a). Furthermore, the resulting knowledge gap seemed tempered in the case of a longer lasting event (e.g. the impeachment developments (Figure 8b). gain: Both education and overall interest were related to information gain, and interest was a better predictor throughout. At the same time, the relationship between edu— cation and interest seems not entirely topic-independent, being consistently lower for NFL and higher for impeachment. 123 4/ 1:39? 3 \) hi interest group 2~J—_____3:L€ K,§fl' .mean 1, /, 1 knowledge I, { l a / . ‘1' , I I / ” t [I i 333 j /” I] L [I 1w. L, I 1 / -1 4. 9* 7'10 interest group ,AN 1" ---- &_:,¢ _2 _ 4 I’,/’ w/,’ T1 T2 Figure 8a (Hypothesis A). Knowledge Comparisons Over Time, National Football League Strike. 1 1 1"”"-—---_ . 0 mean iéflfh————‘ffjfi+n~h1 interest group knowledge A- +L - - _. _ __ 1 je’L1«———~"T’:f‘ -0 interest group - 4..- - " d .- fig’. -2 Figure 8b (Hypothesis C). Knowledge Comparisons Over Time, Impeachment Developments. 124 Both education and overall interest were related to factual and structural knowledge; we also identified a tendency for interest to correlate much more strongly with structural knowledge. Specific: The components of overall interest did not relate to knowledge as expected. Milieu interest emerged as the strongest predictor, leaving the other components behind. Specific components of interest did not do better than edu- cation as predictors of knowledge. A positive exception was milieu interest and potential milieu interest, which did emerge as stronger predictors for structural knowledge. Discussion There are several important points to dwell upon in this discussion. First, the emergence of milieu interest as the best predictor of knowledge clearly needs attention in further research. Presently, we have seen that in the con- text of one political and one sports event, interest stemming from perceived utility to milieu was at work throughout. The other interest components were not activated sufficient- ly as predictors of knowledge. The implication of this may be at least twofold; one would regard the nature of information made available through the media. That is, what type of events does it take to activate self-interest; also, are such topics aired over television? To begin answering this question, one has to .III I!" ll- nllIll'l.‘IIllll I I I ll III .III III I I I : II 125 encompass diverse publicized topics besides political and sports events; for example, economic, ethical or religious issues. These topical areas are mentioned as possibilities because they are likely to touch off areas where the direct economic well being, or principles held by an individual are at stake. Such an inquiry would help clarify the role, if any, played by the self-interest component. It may indi- cate that self-interest does emerge as a predictor of knowl- edge for only certain kinds of broadcast tOpics. Examples of that would be inflation developments, the controversy over absorption of large numbers of Vietnamese, or the con- troversy over Catholicism and abortion. In case self- interest does not yet seem activated, it would become necessary to question the very role of self-interest as a viable component of the predictor variable, overall interest. Furthermore, since milieu interest did play a domi— nant role, it is important to understand well what it stands for. The least complicated interpretation would emphasize the mere facilitation of talk, chat or discussion. The likelihood of having something to build conversation about, based on shared concerns at various levels of intensity. Earlier we referred to utility to milieu seen in terms of communicative facilitation related to an individuars social affiliation groups. However, this may hardly be doing justice to the en- tire picture. Building conversation over issues of various 126 levels of intensity implies involvement with different seg- ments within the social milieu, i.e. friends perhaps being placed closer to the individual than fellow workers. Viewed this way, the social milieu become less attractive as a catch-all notion; it looks more promising when used to distinguish among possible variability in utility percep- tions, depending on whg is the focus of concern. Thus we may embark on a more complicated, but compelling line of interpretation. For example, Greenberg* has suggested altruism as an interesting interpretation regarding perce- ived utility to milieu. Presently, the shape of our data precludes analysis along these lines, but the implications are clear. First, measurement might be such as to allow differentiation within the milieu (e.g., utility perceptions to milieu regarding close friends, as compared to perceived utility to milieu regarding distant members of the social circle). Also, it could be expected that some interesting interaction may take place between the self and milieu com- ponents; for example an inverse relationship may emerge be- tween self-interest and milieu-as altruism component, at least in certain cases. This last point is debatable, of course, if altruism is viewed as self-denial, yet awareness of the existence of self interest on a particular occasion. * Bradley S. Greenberg, informal communication, Spring 1975. 127 A final idea in interpreting milieu interest based on differentiation within the milieu suggests an overhaul of the self/milieu dichotomy. It emphasizes the point that just about everything gets discussed sooner or later, but issues are being discussed selectively. So it may turn out that what we presently term milieu interest regards discus- sion in the broader social circle; topics which would be primarily discussed with intimates may represent what we . presently term self-interest. Thus one way of providing a flexible linkage between the two would be to view communicative utility in terms of one's primary and secondary groups. Furthermore, one may have to consider interest which does not necessarily entail discussion; so utility categories other than the communica- tive variety emerge; for example, issues where behavioral or gratification outcomes take precedence. Another point of discussion here regards the relation- ship between education and interest and their predictive power for factual and structural knowledge. As already noted, they were associated with each other, and each also was related to factual and structural knowledge. Moreover, while their predictive power for factual knowledge was roughly comparable, interest was much stronger than educa- tion in relation to structural knowledge. Finally, the com— bined predictive power of education and interest taken together was only slightly better than the predictive power 128 of interest taken alone. Taking this last outcome in par- ticular consideration we wonder whether viewing the assoc- iation between education and interest as a part-whole relationship may not be helpful, where education contributes to the whole configuration of existing interests. We did the preliminary checks possible at this time, by computing the partial correlation coefficients for both events at two points in time. The purpose of doing this was to see how much change in the magnitude of correlation be- tween interest and knowledge would occur when the influence of education is removed. If the notion of part/whole rela- tionship is on the right track, then partialling out the in- fluence of education would not diminish the interest- knowledge association drastically. Table 59A shows the re- sults of this check. As can be seen there, partialling out education slightly diminishes the magnitude of correlation between interest and knowledge. Table 59A. Partial Correlation Check for Interest, Knowl- edge and Education NFL T1 NFL T2 IMP Tl IMP T2 a) .317 .355 .345 .391 b) .307 .346 .301 .362 a) = correlation between interest and knowledge. = same as above, with education partialled out. 129 It is also noticeable that the correlations at time two have increased, and the explanations of this are not likely to be very crisp at this time. It is conceivable that the first time around respondents answered to the interest questions in a more or less stereotypic way, particularly since the commitment of further thought was to come later on with the knowledge questions. This possibility can be backed up with the existences of "switchers" at time two, i.e. peOple who changed their mind a week later and decided that they really were interested in the discussed event. The other possible explanation is less pleasant; that is un- avoidable sensitization may have occurred with some people, where the mere presence of the survey aroused interest. This however can be contradicted by the presence of those "switchers" who at time two lost interest in the event(s).2 Indeed it seems that further exploration of the issue is necessary in order to understand the consequences of all this with respect to knowledge about different kinds of events. The place where we would begin is establishing what are the types of cases for which education becomes a size- able component of interest in the event, more so that we have presently found for one political event. The goal of such exploration would be to decide in the long run whether lFor NFL n=4; Impeachment n=40; 2for NFL n=78; Impeachment n=44. 130 using the combined predictive power of education and inter- est is necessary, if overall interest would do about the same job most of the time. We would like to add to this discussion one further aspect which has not been touched upon. After testing our three general hypotheses as stated specifically for the study, we saw fit to test the underlying causal flow from interest to knowledge. We used the cross-lagged correla- tional technique, applying the Rozelle - Campbell baseline criterion for causal relationships between two variables at two points in time. Also, since potential interest was not measured at time two, we computed the cross-lagged analysis both with and without the potential interest component for each event; the pattern remained stable both ways (Figure 8c). Since cross-lagged analysis is revealing of mutual causation and sensitive to time lag equivalencies, we thought such analysis particularly interesting for two reasons. These are, the short time lag used in the present study, and also the difference in media display duration between the two events, football strike and impeachment developments. There was some possibility, therefore, that the test would show differences in the way the main causal process of con- cern to us will manifest itself in each instance; it will be seen that this is precisely what happened. As the results indicate the main diagonal f exceeds the baseline for both events, so that we can infer a causal . 3258 . 3414 (Couputed with potential interest oonponents) 131 (Computed without poten- tial interest oouponents) T1 int . 5184 ‘T2 'int . 3726 . 2909 . 3551 e V W know1 .7557 ; 1010.72 baseline = .2357 e, f > .2357 T2 1 . 5085 )lntz T1 int e V v knowl .4332 E know baseline = .1611 e, f > .1611 _NZ'E Tl .5593 \T2 . 7 . intl int2 f .3551 e7f kI‘Iow 7559 my 1—'———-> 0W2 baseline = .2264 e, f > .2264 % T1 T2 intl .5099 > int2 f .3913 £72 V v know1 .4332 > ow2 baseline = 1732 e, f > .1732 T1 a T2 n = 243: baseline = d V b Figure 8c. Cross-lagged Analysis of the Causal Flow. 132 flow from interest at time one to knowledge at time two. More importantly, diagonal 2 also exceeds the baseline in both cases. Since both diagonals exceed the baseline, it will appear that causality flows in both directions, and the question becomes one of time lag equivalences; e.g., which process is faster, the causal flow from interest to knowl- edge, or the flow from knowledge towards interest. In the case of a short time lag, which we have here, the slower process (f) would not have manifested itself. This seems to be happening in the case of NFL. With impeach- ment, the main causal flow from interest to knowledge has already manifested itself, since the main diagonal f exceeds BREE the baseline and diagonal g. This outcome is particu- larly interesting given the differences in mass media dis- play duration between the two events. Impeachment was of longer standing, thereby allowing sufficient time for the slower process to emerge in view. Further work should focus on events of equivalent durations in mass media display in order to ascertain whether the above interpretation of re- sults is basically sound. Another point of interest would be ascertaining the difference in speed between the two causal processes as such, for various events and time lag conditions. Discussion of the above mentioned work in £939, and all that remains to be done, cannot be divorced from dis- cussing the present study's limitations. Let us note first I I.l III. I is: II! 'II .I‘..I I I . I‘ll 133 that to the extent to which this study was built without many specific conceptual or methodological precedents, some limitations were bound to emerge after the fact. Thus the failure of the specific hypotheses, where the postulated hierarchy of utilities did not emerge as expected, is a case in point. On the other hand, these latter findings, although disappointing, have the definite merit of raising important issues about the nature of information and sources of infor- mation which go beyond the parameters of this particular study. Furthermore, we have learned a few methodoloqical lessons as well. The design of future work wrestling with the implications of the interest model and the differential levels of information acquisition would surely have to take into account the shortcomings of this work. Let us first briefly outline some of these short- comings, should the same or similar design be considered. One is the need for longer time span between measures in the panel design, or keeping the same time span and then implement more than two measures. This would give a better feeling for variations in interest and the reasons for it, as the events unfold through mass media coverage. The prob- lem in doing this would be the generally limited duration of publicized events. One will have to be prepared to handle the problems that arise in studying events which may not per- sist in the media as long as the researcher would like. This happened with this study, where the Bell Telephone strike in 134 in the summer of 1974 was called off on a Sunday night, around midnight, so that the event dropped from the media and had to be dropped from consideration for this study. Another important point is the inclusion and study of a larger number of diverse events simultaneously. This im- plies that we would be taking more measures within the same time span. The considerations here are at least of two kinds. First, the appropriate selection of events is of cru- cial importance, as well as carefully keeping track of event developments, particularly if measures are to be taken at more than two points in time. Such a study may then be carried out much better through face-to-face interviews in the respondents' household, to allow sufficient time to go through the considerable battery of questions, and assure a reasonable amount of cooperation. However, cost and mortal- ity problems would enter the picture. With such changes, one could better clarify the issue of knowledge gap attenu- ation over time. Presently we can only say that results on Hypothesis C give a tentative answer; that is, it may be that with the information saturation over time, those less inter- ested have the opportunity to catch up on old knowledge; or, that accruing information would also generate some degree of interest where there was none before. Clearly, there is room for better testing and more definitive answers. Another change is the need to measure the potential interest components more than once during a panel. This 135 would indeed be much more sensible within a study using either longer time spans, or multiple measures in time. Then, shifts in the perception of potential utility could be justifiably expected to occur and be examined. Finally, knowledge items measuring the dependent vari- able should definitely move away from the dichotomous format, to refine the possible comparisons of differential knowledge levels. Immediate suggestions for format are as follows; first, questions which call for simple yes/no, or true/false reply should be eliminated so that there is no question about some of the respondents guessing, rather than truly knowing, the right answers. In this study, the large proportions of false replies given by respondents in answering questions where they could have guessed, alleviated our concern this time around. Yet there is no need to run such a risk again. Another suggestion is for use of knowledge questions construc- ted in a manner which allows the measure of degrees of knowl- edge among the "informed"; that is, carefully designing items which would allow us to go into as much depth and exhaustive- ness, as the actual knowledge of the respondent allows. This is probably the appropriate time to return briefly to the shape of hypothesized relationships presented earlier. After incorporating some of the changes suggested by test results, Figure 8d gives a diagram of relationships as they look now. The main change regards the hierarchy postulated 136 mmHnmCOHumHmm )_-..___-____..----- -------7------------(F -—---(s T R U c T U R A L)--—---—« A AR —’---- --r—-——--- ---——. AMHmmv smHHHE Amawmvo —---‘--—_—- -——-.L-——-----_-.I-d deHHE pmNHmmnuommm mo EMHmMHQ meH>mm EIJQDU‘IIE-IHOZ .Um musmHm 0A muHHHuD Um>Honmm 137 among the interest components. The dominant role of milieu interest over the self interest is, therefore, to be noted. We have also indicated a tentative linkage, rather than the original full juxtaposition of education versus interest. This tentative linkage is meant to indicate our reflection over the possible part-whole relationship, which we must leave as an Open issue for the time being. Finally, we have tried to visually indicate the tendency for overall interest to associate itself better with the structural component of overall knowledge, while factual knowledge is associated comparably with both education and overall interest. Apart from noting the concrete changes in the present model based on test results, we would like to briefly re- capitulate some of the remarks made earlier in relation to the self-milieu dichotomy and the latter's dominant role. These few ideas, while not developed at the present time, take us toward a possible alternative conceptualization of the problem area. Thus, while keeping the underlying notion of interest as a useful variable, it seems sensible to broaden the notion of utilities in which interest may be rooted. We would like to eventually identify a set of com- munication utilities as they are linked to one's primary and/ or secondary groups of human association. Furthermore, we would like to expand on the notion of utilities other than the communicative kind using notions found in the functions and gratifications approach in the past. All of this implies 138 covering a vast amount of ground, of course; one probable result, however, could be a useful configuration of inter— related typologies. Such a configuration may include the links among the set of primary, secondary and other groups of human association and the various communicative and non- communicative utilities as perceived in terms of carefully prepared typologies of events and information available through a variety of sources. We have come to recognize that the methodological make-up of an investigator's studies is likely to improve not only with the accumulation of experience (sine-qua-non), but also with his/her increasing opulence; this would make for fewer pressures due to limited resources and need to "cut corners." Given all of the above considerations and caveats, it remains for the next efforts focusing on the same problem area to build on the sound ideas and remedy the previous weaknesses. APPENDIX A Sample Characteristics APPENDIX A Sample Characteristics Tables 59, 60 and 61 give the demographic and mass media use profiles obtained for our sample. Briefly, 45.5% of the respondents were male, 54.5% were female, with a reasonable spread in terms of education, occupational status and age; the mass media use patterns indicated that the majority of respondents used mostly one (60%) or two (26%) media to get the news; in terms of specific media, tele- vision emerges as the favorite, followed by newspapers and radio, with magazines ranking last. Also, 29 respondents were reluctant to cooperate further, after having replied to the initial interview, and were thus excluded from the usable sample (e.g., 7% don't- call-next-week). We deemed it necessary to check for possible systematic bias, by comparing those who did not agree to participate in the second wave, with the respondents who agreed to do so. Table 62 gives the results of this com- parison between the study sample and the non-cooperative group. All tests indicated lack of significant differences on any demographic dimension, or in terms of reported interest, discussion and knowledge on the topics used for the study. The conclusion, therefore, was that both the un- cooperative group and the study sample have been drawn from the same population. Furthermore, Tables 63 and 64 give 139 140 sample comparisons with available census data. Table 59B. Demographic Profile of the Sample1 Relative Absolute Frequency Variable Frequency (Percent) Sex Male 115 45.5 Female 138 54.5 253 100.0 Age 18-20's 86 34.0 30's 62 24.6 40's 36 14.2 50's 32 12.6 60's 15 5.9 70-80's 22 8.7 253 100.0 Occupation Retired 33 13.0 Housewife 50 19.8 Labor, Service 43 17.0 Craftsman, Foreman 11 4.3 Sales, Clerical 40 15.8 Professional 52 20.6 Official, Manager 12 4.8 Student 12 4.7 253 100.0 Education 6th grade 4 1.6 Junior high-some high school 38 15.0 Finished high school 74 29.2 Some college 71 28.1 Finished college 38 15.0 Graduate work 28 11.1 253 100.0 1 n=253 141 Table 60. Use of Separate Media TV Newspapers Radio Magazines n % n % n % n % Yes 161 63.6 125 49.4 76 30.0 37 14.6 No 92 36.4 128 50.6 177 70.0 216 85.4 253 100.0 253 100.0 253 100.0 253 100.0 Table 61. Overall Mass Media Use Profile Number of Media Absolute Mentioned Frequency Percent Use none 3 1.2 Use one 152 60.1 Use two 66 26.1 Use three 15 5.9 Use four 17 6.7 253 100.0 142 Table 62. Bartlett's Test Comparisons of the Sample and the Uncooperative Group Sample Group 2 Variable Variance Variance df x Sex .326 0.233 1 1.2201 n.s. Age 2.586 2.254 1 .2471 n.s. Education 1.642 1.278 1 .7027 n.s Occupation 6.266 6.713 1 0.1051 n.s. Overall media use .794 1.238 1 2.7750 n.s. Interest index on NFL strike 1.227 1.078 1 0.1604 n.s. Discussion of strike 1.060 0.970 1 0.0885 n.s. Overall knowledge NFL strike 3.206 3.965 1 0.5865 n.s. Interest index impeachment developments 2.114 1.780 1 0.3264 n.s. Discussion of impeachment deve10pments 1.902 1.958 1 0.0249 n.s. Overall knowledge impeachment developments 2.968 3.123 1 0.0027 n.s. 1 n=253 and n=29 respectively 143 Table 63. Comparison of Sample Demographic Characteristics and 1970 Census Data Ihian quacUajsth: Lamflng-+EBmexz='Rfials % Samfled98 Sex: .Male 39,276 46,236 85,512 48.1 45.5 Female 44,763 47,465 92,228 51.9 54.5 177,740 Age: 18-20's 29,151 49,966 79,117 44.5 34.0 30's 13,620 12,541 26,161 14.8 24.5 40's 13,660 11,751 25,411 14.3 14.2 50's 11,871 8,894 20,765 11.7 12.6 60's 8,485 5,792 14,277 8.0 5.9 70-80's 7,252 4,757 12,009 6.8 8.7 Emmxuion: 7___ less than 6 years 4,214 1,565 5,779 4.8 1.6 F, samehbfilsdxxfl 5; (including 7+8) 23,087 12,432 35,519 29.6 15.0 C; finished high school 24,015 16,916 40,931 34.0 29.2 fi sane college 7,969 8,206 16,175 13.5 28.1 finished college 3,484 6,499 9,983 8.3 15.0 " graduate work 2,982 8,802 11,784 9.8 11.1 a 18-24 year olds g w/o high school 4,687 2,707 7,394 - - The area covered by our study did not fully coincide with the census definition of "urban balance" or Lansing area; also the age and education categories available in the census data made direct comparisons impossible. The above figures then, represent an approximate comparison. This com- parison indicates a higher overall level of education mani- fested in our study sample. 144 Table 64. Census Data for Occupational Classifications for Lansing SMSA (breakdown for Lansing Area + Urban Balance unavailable) Occupational Category Total POpulation % Sampled % *Labor, service (including Operatives) 48,798 26% 25.0% Craftsmen, Foremen 19,393 10% 6.5% Sales, Clerical 40,452 21% 23.5% Professional 25,831 14% 30.6% Officials, Managers 10,507 6% 7.0% Students 43,778 23% 7.0% *1abor and service only: 23,844; Operatives and transport: 24,954 Our study also included housewives and retired persons in the sample. However, since these categories are not reported in census data, they are omitted from the com- parison and the sampled percentages recomputed based on com- parable categories. The occupational categories which 335$ available in the census data were measured and compared with the adjusted sampled percentages. The figures above represent that comparison. (There were two other occupational categories reported in the census data, but these** were judged to be negligible for purposes of comparison. These were also omitted since they weren't reported separately in the study. See below.) ** farm managers and workers: 3,204; private house- hold workers: 1,246 145 The two large discrepancies, professional and student, can be explained as follows: Students--the census is done in April while MSU is in session; our study was done in August, and furthermore, dorm phones were discarded from the sample. Professiona1-—as noted elsewhere we did not sample the entire SMSA. It is plausible that more professionals live in Lansing (city) and the urban balance than in the remainder of the SMSA. We have some indication of this when we examine the education levels. Over 86% of people with five years or more of college live in the sampled area, and over 78% of the college grads do so, as compared to about 72% of the total pOpulation. APPENDIX B General Measurement Procedures APPENDIX B General Measurement Procedures Here we shall present: (a) an overview of the general measurement pro- cedures employed for both events, and (b) the preliminary variable deve10pment work. Overview The purpose of this study dictated the creation of eight variables in total. These were the four basic pre- dictor variables (self interest, milieu interest, potential self interest and potential milieu interest) and in addition, overall interest; also, the two basic criterion variables (factual knowledge, structural knowledge) and in addition, overall knowledge. We also measured education as a corollary demographic predictor variable in contrast to ours. As mentioned earlier, the behavioral measures needed here were devised largely ex novo, given the lack of many useful precedents in prior research. Thus, the deve10pment of variable classes underwent three phases: (a) operationalization of component measures at the questionnaire deve10pment stage. (b) check on component measures configuration based on the obtained data. 146 147 (c) obtaining final measure of our variables, by indexing the appropriate component measures. By necessity we made some a priori decisions at the questionnaire deve10pment stage, with careful consideration to conceptual fit to the interest mode. Next we checked the actual data configuration for the measurement components underlying each variable class, for each event, at two points in time. This was done by means of factor analysis, where varimax rotation appears to be most appropriate for our needs; we used principal axis solution, and stepped at three factors, after subsequent solutions did not contribute to data interpretation. (See Tables 65—67c (NFL), 68-71c (IMP.)) This check was done in order to avoid distortions introduced by undue a priori variable structuration, while disregarding indications given by response data, reflecting the conceptual set of queried individuals. In short, we stood by the variables as postulated in our model but re- mained Open to revisions regarding some component measures for these variables. We thought such flexibility appropriate, since from the beginning our argument had been for receiver orientation, so that sensitivity to manifest respondent be- havior becomes the raison d'etre of our model. Once we established a sensible set of component measures for each predictor class, we applied the following identical procedures for all. Using factor analysis, we obtained factor scores for the 3 measures in each predictor 148 class and combined them into a single index. This method was chosen because it enables us to come up with an index which conceptually taps all predictor class measures, while combining them in such a way that greater weight is given to the components which emerge as central within a given class. This method was also useful in gleaning the relation- ship among component measures in a way which would help in further work, but was missing at the planning stage of the present study. Throughout, we used principal axis solution, quartermax rotations. With respect to principal axis solu- tions, we settled for three factor solution, since further factoring did not markedly change, i.e. the solution stabilized; orthogonal quartermax rotation was chosen since it simplifies variable structure and is thus useful in measurement procedures. The choice of final factor was guided by three criteria: (a) proportion of the variance accounted for in all component variables; (b) proportion of the variance accounted for in a given variable by a given solution (i.e. communality); (c) factor purity of the variable on a given factor. Our goal was to base the index on a factor which provides the best conceptual fit with the predictor variable postu- lated in the model. 149 For the dependent variables, we used the standard- ized scores to combine the n measures of knowledge in each criterion class into a single index. The resulting index reflected the relative status of any score in the overall distribution of responses to knowledge items. Preliminary Data Development Work Following is a description of the varimax check on measure configurations for both events at two points in time. NFL Strike NFL SELF INTEREST was seen in terms of the component measures: . NFL effect . NFL talk relatives . NFL talk at work . NFL talk others bWNI—J Questionnaire items: (1) Have you discussed it with friends? . . . (2) With relatives? (3) With people at work? (4) Anybody else? NFL POTENTIAL SELF INTEREST was seen in terms of the component measures: . NFL potential effect NFL potential price NFL potential job NFL potential enjoyment NFL potential keep up WubWNI-J 150 Questionnaire items: (1) (2) (3) (4) (5) Do you think the NFL strike could affect you in any way in the near future? Would that be an effect on the cost of living or prices for you? Could it affect your job, or the job of someone close to you? Will the strike have an effect on your enjoyment watching the games? Is the NFL strike the kind of thing you will want to keep up with? NFL POTENTIAL MILIEU INTEREST was seen in terms of the component measures: DMNH NFL potential friends NFL potential relatives NFL potential work NFL potential others. Questionnaire items: (1) (2) (3) (4) Do you think you will talk about it with friends? With relatives? With peOple at work? Anybody else? The factor analytic check on component measures con- figuration for NFL strike gave the following indications: (a) the measure of game enjoyment shifted from the expected self interest group, actually alining itself with the milieu interest group of component measures. 151 (b) the measure of keeping up with the event failed to align itself consistently with either group of interest measures.1 The problem was consistent for both items one and two, so that the final configurations we established for further indexing were as follows: NFL SELF INTEREST 1. NFL 2. NFL 3. NFL effect job cost NFL MILIEU INTEREST 1. NFL 2. NFL 3. NFL 4. NFL 5. NFL enjoyment talk friends talk relatives talk at work talk others NFL potential self interest and potential milieu interest followed the same pattern. Following on next page are the tables and figures illustrative of the above mentioned procedures. 1As this component measure was also inconsistent in its placement for impeachment developments, it was removed from data anlaysis. 152 Table 65. Varimax Configuration Check, NFL, Time One, Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 1. NFL Effect 0.16314 0.80977 0.17565 2. NFL Cost 0.07744 0.73238 0.02193 3. NFL Job 0.09344 0.51029 0.08571 4. NFL Enjoyment 0.35297 0.09122 0.28008 5. NFL Keep Up 0.18443 0.15012 0.70786 6. NFL Talk Friends 0.88299 0.08082 0.28037 7. NFL Talk Relatives 0.61222 0.13299 0.37210 8. NFL Talk Work 0.71054 0.09485 0.15474 9. NFL Talk Ohters 0.56809 0.14516 -0.04408 Horizontal Factor 1 1* vertical Factor 2 1: 1k 9: * * 2 * * * * 3 * * * * * * * 5 * 9 7 * 4 8 6 * ******************************** * * * * * 'k * * Figure 9. Vector Representation 153 mmN meoN.o mmvo.o 9.850 fine dz EN 836 836 #8: fig. dz 36 $26 836 6933.8 58. dz EN 826 2:6 «and fine dz EN 836 £36 8 83 dz mmm 22.6 336 ".ng dz RN 63.4.6 :36 non dz RN 886 886 68 dz Rm 236 3.26 68mm dz 88 838 883m 88. 63685 .6835 83:38 8 85. .dz .890 838508 g 08 33. 8864 8626 «:86 62.36 32.66 8626 $366 6286 2686 380 :3. dz 8864 633.6 632.6 «64.86 8686 6336 3626 2.636 85: fine. dz 8684 8366 28.86 3366 H886 3626 «:26 £383 fine dz 8684 6:86 3E6 mmmz6 22:6 6836 3d fine dz 8864 «836 8636 8656 3:6 8 89m dz 6884 8686 $686 8686 uanm dz 8864 N-mm6 3566 now dz 8864 2:86 88 dz 6886 83m dz Brno fin: .9338 88nd gammzdz ”.8...onqu 8ndz umoodz 88.de fine dz fine dz fine dz fins dz 386338 833808 888.5 533538 .68 m5. .dz 3.85 8038380 885 63 389 28.3.6 8850 fine dz 5.3.6 fie: 58. dz 8636 69388 #3. dz «2.36 882 fine dz 336.6 a: menu dz 9336 ugHfi dz 662 66 £366 m 6356 non dz N66 66” 83:6 N 8396 88 dz 6.3 «.8 8366 H 66:6 88mm dz 86 80 .85 no 8 888d Spud BHEEEB 38g ”.8435 efiaflxfimm .8 mad. dz .865 gfismflufio .855, .56 289 154 *********** Table 66. Varimax Configuration Check, NFL, Time One, Potential Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 1. NFL Potential Effect 0.24882 0.59278 0.37225 2. NFL Potential Prices 0.12597 0.89621 0.05916 3. NFL Potential Job 0.00152 0.65848 0.07733 4. NFL Potential Enjoyment 0.42849 0.28052 0.75976 5. NFL Potential Keep Up 0.57592 0.16152 0.15217 6. NFL Potential Friends 0.94100 0.09608 0.10987 7. NFL Potential Relatives 0.85860 0.04186 0.10258 8. NFL Potential Work 0.71482 0.11236 0.26829 9. NFL Potential Others 0.4306 0.04380 0.12772 Horizontal Factor 1 ,* vertical Factor 2 * 'k * 'k * * 3 * 'k * 1 'k 'k * 'k * 4 * * 5 * 8 * 6 * 9 7 * it * * * * 'k 'k * Figure 10- Vector Representation ******************** 155 mmm mumcuo Hm-.o mmmo.o emzucmuoa azz mmN Hemm.e Neee.o x603 deducmuoe 5E2 mm~ mmem.e mmefi.o mm>wumzmm Hmflucmuom azz mmm 6666.6 mmo~.e mcewzue emfiucwuoa szz mmm wmmv.o Haem.o do ammz HezucOOOE qzz mm~ nmmm.o mmmH.o acmezoncm HmHucmuom qmz mmm Hamm.e mmmo.o nee Hazucouoe 562 mmw mmmm.e HNNH.6 mmozna 5668:6606 S62 mmm om~e.o HH6~.o pumoLm Amusemeom ndz mmmmo cozumfl>mo eumwcuum cam: wznmfium> ummnmucH smeflezxuamm Hmeucmuom .mco made .gzz .xomno coflumusmzmcoo xmeflum> .066 manna ooooo.H mommm.o Hmome.o mmomm.o HHH-.6 meem~.o memoe.o Hence.o omomH.o mumeuo Hmfiucmuom Adz ooooe.~ H6666.o Heome.o HH~H6.6 oaomm.e mmmme.e oeo-.o Hemam.o xuoz deflecmuom azz oeooe.a Hoemm.e oezmm.o mzeme.e memmo.o Hmeee.o momom.e mm>aumamz HmHucwuom adz ooooo.H vm~mm.o mmdzm.o ”6666.6 meema.e Heomn.o mwcmzud Hezuewuom azz ooooo.z Heaze.o momeo.o eeme~.o eaem~.e 6: ammz Hmflucmuom gdz ooooo.H oooe~.e Hemem.o ~m~em.o acmeoncm HmHucwuom qdz ooeeo.e omemm.o mmmme.o non Hmaucmuom azz ooooo.H Nemem.e mmofiue Hmducmuoa azz ooooo.~ powuum Hazucmuoe ezz mumcuo x403 HmHu mm>flunfimz xfiae mecmfiud game 6: amoz nemeoncm ace Hafiu mmoaua nooemm HMHUCBOQ Hmz Icmug E 33:38 152 3.35.88 .sz HmHucqum E HdHucmuOm .Hmz Icwuom .Hmz HMUGBOA .Hmz HMHUCBOm Hmz mucwflofluwmoo coflumHmuuoo ammumucH amdefizxwzwm Hwflucmuoa .mco mEHe .gdz .xumgu coHumusmflucoo xmeauu> .mee manna vooomd ”350 83:38 .52 nmmomd €03 83598 .52 35:6 mgHumHma 83:38 E mnmomé mpcmCm HmHucmuOd «$2 mmomm.o a: amoz HMHucmuom gmz mmmmmd uEQSODQ HmHuGOuom Hmz 0.03 H4. mmmmmd m mmmmeé now 83:38 .52 m.mm m.m~ 56264 N nmmmmd mooHum HmHucmuOd .Ez Hfio .73 bmvmmd A 233.0 uomuum HoHucmuOm .EZ yum Ed .85 mo pom 3Hm>cwch uOuUmm 33092.58 Snag—“.5 umwumucH cmHHHsQMHmm Hwflcwuom .96 95.9 :52 £090 50328350 meHHmS Jam 0.33. 156 Table 67. Varimax Configuration Check, NFL, Time Two, Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 1. NFL Effect 0.10465 0.72879 0.35189 2. NFL Cost 0.11680 0.86715 -0.03952 3. NFL Job 0.09102 0.54916 0.07378 4. NFL Enjoyment 0.36526 0.13717 0.47472 5. NFL Keep Up 0.55047 0.20841 0.42947 6. NFL Talk Friends 0.93906 0.07619 0.13760 7. NFL Talk Relatives 0.79156 0.16620 0.14439 8. NFL Talk Work 0.82414 0.10746 0.25386 9. NFL Talk Others 0.35764 0.06554 0.26068 Horizontal Factor 1 Vertical Factor 2 * 'k * 2 'k * * l 'k * * * 3 * * 'k 'k * * 5 * 7 * 4 8 * 9 6 * ******************************** 'k 'k 'k * 'k * 'k * Figure 11. Vector Representation 157 mg 2.36 $85 828 fine dz 2a 326 $26 do: fine dz 2a 235 8:6 @93de fine dz new 885 :26 88nd fine dz Gm 336 3:6 8 mums dz 2a 5.35 35.0 ufisflond dz 2a 236 $86 now dz new and 2.85 u80 dz 2a 886 88.0 8.8mm dz 396 8335.8 gm 5...: «33:5 3835 83238 .99 Ed. .dz £86 Sflflsmficoo g .03 «38. 8°84 3.de dead 332° 2525 «885 2535 8806 £035 ammo fine dz 8254 «:25 :33 £235 2.5.5 «335 «$55 :33. do: find. dz 88°; 82.2. @336 Sand @826 58.0.5 «Emu... 8.53 £8. dz 8084 83...... 23:6 H235 38:5 34.35 £82 £8. dz 88°; 2536 Sofia and... $83. 8 5 dz 88o; 833. $83. 833. ufiazofld dz 8254 @886 3.96 nos 9% 8°84 1.68.0 .88 dz 8084 88mm dz muflno do: 8533mm 882 ma mud dz ugfld dz non dz 3.8 dz 83mm dz gflmz fig 395 a; 353338 832.58 #835 323.538 65 mad. .dz .fimfi 83856380 unsung é... manna 3°86 mafia fine dz ommmsd x83 58. dz momnod 8.33de fine dz mmoomé 882 58. dz $635 8 302 dz $2.35 agnfi dz 0.03 Tm 556 m 23.3 non dz 9: 0.8 $5qu N 3325 #8 dz TS 33 283” H 339° 88% dz 3m So “as do 8m «:3qu “Bond Bag «Bonus $835 sfldzhawm .03. Ed. .dz £88 Sflgmdfio gag £5 manna ‘.illl|llllllllll£'iI-l'l‘ll‘fl" Impeachment Development 158 At the questionnaire development stage, our expecta— tions regarding variable structure was reflected in the following operationalizations. IMPEACHMENT SELF INTEREST was seen in terms of the component measures 3 Impeachment cost Impeachment job UluwaH .0 Questionnaire items: Impeachment effect . Impeachment satisfaction . Impeachment keep up (1) Do you think the impeachment events have an effect on your life in any way? (2) Do impeachment events have an effect on the cost of living or prices for you? (3) Do you think these events have an effect on your job, or the job of someone close to you? (4) Do impeachment events have an effect on your general satisfaction with things around you? (5) Are the impeachment events the kind of thing you just want to keep up_with? IMPEACHMENT MILIEU INTEREST was seen in terms of the component measures : Impeachment talk Impeachment talk . Impeachment talk . Impeachment talk hWNI—J C. Questionnaire items: friends relatives at work others (1) Have you discussed the impeachment with friends? (2) With relatives? III. I‘.lllll\ I III-II (3) (4) 159 With people at work? Anybody else? IMPEACHMENT POTENTIAL SELF INTEREST was seen in terms of the component measures: 1. 2 3 4. 5 Impeachment potential effect Impeachment potential prices Impeachment potential job Impeachment potential satisfaction Impeachment potential keep up Questionnaire items: (1) (2) (3) (4) (5) Do you think impeachment developments could have an effect on your life in the near future? Could there be an effect on the cost of living or prices for you? Could there be an effect on your job, or the job of someone close to you? Will impeachment events have an effect on your general satisfaction with things around you? Are impeachment events the kind of thing you will want to keep up with? IMPEACHMENT POTENTIAL MILIEU INTEREST was seen in terms of the component measures: hWNI—d Impeachment potential talk friends Impeachment potential talk relatives Impeachment potential talk at work Impeachment potential talk others Questionnaire items: (1) (2) (3) (4) Do you think you will be talking about it with friends? With relatives? With people at work? Anybody else? 160 The factor analytic check on component measures con- figuration for impeachment developments gave the following indications: (a) the measure of keeping up with the event was inconsistent in its placement, as with NFL, and was therefore removed from further data analysis.1 The remaining pattern was consistent for both times one and two, so that the final configurations we established for further indexing were as follows: IMPEACHMENT SELF INTEREST l. Impeachment effect 2. Impeachment cost 3. Impeachment job 4. Impeachment satisfaction IMPEACHMENT MILIEU INTEREST . Impeachment talk friends . Impeachment talk relatives . Impeachment talk at work . Impeachment talk others think)?“ Impeachment potential self interest and potential milieu interest followed the same pattern. Following on next page are the tables and figures illustrative of the above mentioned procedures. 1Impeachment keep up shifted from the milieu con- figuration at time one, to the self interest group at time two. 161 Table 68. Varimax Configuration Check, Impeachment, Time One, Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 l. IMP Effect 0.22645 0.56636 -0.03272 2. IMP Cost 0.11615 0.60869 0.07431 3. IMP Job 0.04994 0.63600 0.42272 4. IMP Satisfaction 0.18349 0.38488 0.01469 5. IMP Keep Up 0.52656 0.20450 -0.05123 6. IMP Talk Friends 0.84949 0.12887 -0.04l30 7. IMP Talk Relatives 0.69493 0.21305 0.00667 8. IMP Talk Work 0.61166 0.21311 0.25030 9. IMP Talk Others 0.40724 0.09629 0.23160 Horizontal Factor 1 *********** Figure 12. *I-X-fl-X-X-fl-fl-********************* Vertical Factor 2 9 Vector Representation ******************** 162 mmm Hmmvé mhwmd ago v2.3. g mmm ~mmv.o womwd fig an. mzH mmm ovmvd 935.0 sewn—nag wanna. g mmN oNovd vmahd 3382 an. a mmN mmvmd hammd g @03— 3 SN Nvové mmwoé gwmnwm a mu mmmcd mmmmd non EH mmN omwvd vwmwd um8 s CH wmmvd mmmhd gmmm ma mmmuo 8.3358 35m 594 03> Hagan.” smuggudmm .95 3a. 58.50005" .3090 gang fig .Umm Gang. 8884 8310 Sofia 82.3. 83.4.6 «835 8335 £266 321° E930 fine .5 8°84 823.0 6335 Sand 2:36 SENS 8286 5.36 #83 x3. in 8884 838° «536 $35 «$35 $82. 38:. «933mm fine EH 2584 $82. «$85 32:6 «826 «$3.0 gum fine 3 8984 £835 39:6 821° and“... do and an 8084 End... Edda Nomad 568mg an 8254 28.3 38:. non. .5 8°84 $1.36 .88 .5 8°84 88mm an 9.850 88.. ugfiflmz and... 899?: .8368 song «80.3 88%.! 58. an id. an fine .5 fine .5 umflmm 85 flfiaoduooo 833880 $835 sfladhamm .08 man. $850895 .3 9.88.9395 gas do manna Shawna Emfio “as ea 2%.; x8: 0:8. an Rama ugflflmz x89 .qu 38:. 88B 58. an H225 as 89 .5 881° 83338 .5 0.2: fl... 262.0 m 889° non 9n «.3 98 3885 N down... #8 .5 Md 2: mama.” H dmnmd 83mm fin 8m 50 us do 88 83.8.3. 8qu Edna—Bo Edda... umwuwucH amflgwawm .96 958 .ucgfiumwng £090 ggwwfiuo .3553, .cww manna. l1 [I'll-l" 163 Table 69. Varimax Configuration Check, Impeachment, Time One, Potential Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 l. IMP Potential Effect 0.12709 0.70452 0.21558 2. IMP Potential Prices 0.11533 0.76460 0.11036 3. IMP Potential Job 0.17947 0.65772 -0.05893 4. IMP Potential Satisfac- tion 0.17762 0.40701 0.22494 5. IMP Potential Keep Up 0.29434 0.17562 0.39247 6. IMP Potential Friends 0.89681 0.09677 0.25738 7. IMP Potential Relatives 0.80237 0.12899 0.22920 8. IMP Potential Work 0.71552 0.19858 0.02850 9. IMP Potential Others 0.45094 0.19463 0.07316 Horizontal Factor 1 Vertical Factor 2 * * * * 'k '1: * * * * 'I: * 'k 'k * * * 5 8 * 7 * 6 * it******************************* ‘I: * * it * 'k 'k * Figure 13. Vector Representation 164 $6 836 £86 830 88:38 9: RN 886 $86 #83 88:38 9: mm” 836 886 3838 88:38 8: mg 8.86 826 338 8838 9: RN 286 626.6 8 89. 8838 9: mmm 886 8.56 588888 88.38 9: SN 83.6 836 non 8838 9: SN 836 3.26 32.8 8838 9: «8 22.6 $2.6 38m 8838 e: 3o 88888 33m 3 2883 833:: 9888\33 883 .30 8:9 .833 .6080 883880 38> 6% 838. 886.: 8:36 @886 6886 2886 $~86 2.686 8:86 88:6 330 :838 9: 6686.: 65.86 21.56 3.036 32:6 6386 8266 6286 fie... 8838 9: 688.: 6886 8886 8686 3886 .2886 6836 3888a 8838 9: 6668.: 1.526 8286 8686 88:6 8:.«6 3:8 8838 a: 688.: @886 6226 $886 «83.6 8 A32 8838 9: 688.: 6886 2:86 8866 8808883 8838 9: 688.: 2.8m... 8636 new 8838 a: 6686.: 8886 32.8 88588 .2: 6686.: 838 88:38 9: 330 x83 88 3838 3:8 as 3 8:88.088 now 88 3088 3% 88:38.2: .383 8838 a: 8838 9: 8838 a: 8838 9: 38 9: 8838 9: 8838 9: 35:33.80 833300 833:: «333:3 8838 .88 38. 5383:: .3 8883880 38> 6% 389 o.ooa ~.om m.mw O‘VQ I U . C‘QM \DN omona.o cmmma.a momha.m HNM mmmvm.o wamm.o meHh.o hmmbm.o HmHhN.o Hmhvm.o mmmwv.o oHoam.o mmwmm.o 39:5 8838 9: 803 8838 9: 388:8 8838 9: 3:8 8.338 93 as 3 8838 9: 83888 8838 9: non 8838 9: 388 8838 9: 38m 8838 a: 03883 8335 633588. 88.38 65 3:. .8383: :36 8883880 38> 88 £63. 165 Table 70. Varimax Configuration Check, Impeachment, Time Two, Self/Milieu Interest (Milieu) (Self) Factor 1 Factor 2 Factor 3 l. IMP Effect 0.16300 0.68441 0.25378 2. IMP Cost 0.10061 0.42701 0.52315 3. IMP Job 0.13854 0.20277 0.68096 4. IMP Satisfaction 0.23001 0.29010 0.27792 5. IMP Keep Up 0.15213 0.45814 0.10972 6. IMP Talk Friends 0.73724 0.28274 0.03196 7. IMP Talk Relatives 0.77739 0.18296 0.07287 8. IMP Talk Work 0.53180 0.15733 0.14339 9. IMP Talk Others 0.41366 0.01331 0.24189 Horizontal Factor 1 *********** Figure 14. *X-I-fi-X'X-fl'****fl-**X-*****I~******** Vertical Factor 2 9 Vector Representation ******************** mcw ommvd mvbmd mambo an. g MVN owwvd :Nwé xuox g g mvm mmavd 3.2.6 33mg yawn. EH mvw ooomé hhvmé gum and? EH mvm mhmmd mmomd n5 3 g mvm hamvd Numhd gmmjflm g mvm mhmvd vvvvé non QZH mew wmmcd adhé umoo 2 mg. mmwmd mmamd ubmmwm mZH M88 8.33.58 wanna—3m nag manmg $835 awfingwawm .2 “a .535 .3090 gawg g5, .02. wanna. 166 2584 $056 omommé 850.5 >385 «836 2:25 2.de 331° flufio fine am 88°.” «836 mmvmvd 5‘26 mmommé $925 Smfld mmmomd fie: fine an 8804 H886 381° Emfio $33. 3°26 Smamd agflflma fine EH 8084 $32. 88w... 0336 mmmmmé mmoomd «ESE fine an 2584 :3de «3.36 Band .5an as name .5 2584 2:2. $036 H386 Sgumflum EH 8084 «835 33:. non an 8254 3%}. #8 EH 2584 gum an «.350 «.8: 8333a 88.3” a: 982 g 8308 new an $8 an 88% m5 53% as Ema Ema umflmmg 386338 Sflflflu8 ummumofi aofiaguamm .03“. gm. 58.50% 3.090 gunsuflcoo awning .mon wanna. Hammad Eufio “:8. g Nammé #8: fine 9m 2266 mgflflmm 53. ea 3486 882 52. m5 momvmd as amen 9a 832. 530338 an 92: md mmfima m «2.35 non 9n 5.: flan Boga N 381° 38 an ode 9% ~m~$§ H $336 885 g pom 5 .3, mo “.8 83>:me 8qu bag—=8 Bfiflg $335 83233 62. £5. 555895 3.86 Sflfiaflfio Eng d2. Same APPENDIX C Instrument 353-3237 - office APPENDIX C Calls: lst - Key: NA No Answer Busy Not in service Refusal 332-3797 - home 2nd - B = Bissy Genova 2:: : N: : Project Director Day Interviewer's name Time Respondent phone Hello, I am MEDIA NEWS STUDY from the Department of Communication at Michigan State University.... We are doing a study on the news in the mass media and their importance to people. talk to the man (lady) of the house, please? I would like to ask you about two tOpics; it will only take a few minutes of your time.... (another time - probe) CARD ONE COLUMN \Dmflm 10 ll ITEM Card one Deck 1 Subject no. (1) Where do you get most of your news?.... INDEX OF MEDIA USE (2) How interested are you in the National Foot- ball League strike? 167 May I HP‘P‘H (AMI—'0 CODE TV newspapers radio magazine not at all a little some a lot COLUMN 12 l3 14 15 16 17-18 19 20 21 22 23 24 (3) (4) (5) (6) (7) 168 ITEM Do you think the NFL strike has an effect on your life in any way? Do you think the strike has an effect on the cost of living, or prices for you? Do you think the strike has an effect on your job, or the job of someone close to you? Does the NFL strike have an effect on the enjoyment you get out of watching the game? Is the NFL strike the kind of thing you just want to keep up with? INDEX INTEREST (8) (9) (10) (11) Have you discussed it with friends?.... With relatives? With people at work? Anybody else? INDEX DISCUSSION (12) Do you think the NFL strike could affect you in any way in the near future? CODE 0 = no (GO TO ITEM 6) l = yes, no sure, depends, maybe, don't know 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (If no also to item 2, go to item 12) l = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (Go to item 16) l = yes, maybe, don't know lilil'f’u' COLUMN 25 26 27 28 29-30 31 32 33 34 35 36 (13) (14) (15) (16) 169 ITEM Would that be an ef- fect on the cost of living or prices for you? Could it affect your job, or the job of someone close to you? Will the strike have an effect on your en- joyment watching the games? Is the NFL strike the kind of thing you will want to keep up with? INDEX POTENTIAL INT. (l7) (l8) (19) (20) Do you think you will talk about it with friends? With relatives? With people at work? Anybody else? INDEX DISCUSSION (21) What are some of the demands of the NFL players? CODE 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (Go to 21) l = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 1 = Salary increase; Elimination of reserve and op- tion clauses; Veteran veto on trades and waivers; Limit the author- ity of Commis- sioner Pete Rozelle; 0 = Don't know, other COLUMN 37 38 39 40 41 42-43 (22) (23) (24) (25) (26) (27) 170 ITEM What star quarterbacks have crossed the picketlines? How long has the strike been on? What is Ed Garvey's role in the NFL strike? Have exhibition games been successful with rookies and free agents playing? Do you think veterans lose money by remain- ing on strike? Does anything else come to mind in con- nection with the NFL strike? 44-45 KNOWLEDGE INDEX 46 (28) Is knowing about the NFL strike of any use to you? w 1 = Hadl; Staubach; Griese 0 = don't know, others 1 = over a month about 40 days 0 = don't know, others 1 = Executive Director of NFL Players Association (Secretary): Negotiates with the NFL Manage- ment 0 = don't know, others 1 = no; attendance low; 0 = don't know, others 1 = yes; up to $1000/ day 0 = no; don't know 0 = no If YES write in full reply: 0 = no If YES, in what ways? (Write in full reply): CARD TWO COLUMN 1 2 171 ITEM Card two Deck one 3-5 Subject no. 6 10 11 12-13 14 15 16 (l) (2) (3) (4) (5) (6) (7) (8) (9) How interested are you in the impeachment de- ve10pments these days? Do you think the im- peachment events have an effect on your life in any way? Do impeachment events have an effect on the cost of living, or prices for you? Do you think these events have an effect on your job, or the job of someone close to you? Do impeachment events have an effect on your general satisfaction with things around you? Are the impeachment events the kind of thing you just want to keep up with? INDEX INTEREST Have you discussed the impeachment with friends? With relatives? With peOple at work? CODE 2 = impeachment l 0 = not at all 1 = a little 2 = some 3 = a lot 0 = no (Go to item 5) l = yes, not sure, maybe, depends, don't know 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (If no also to item 1, go to 11 l = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes COLUMN 18 19 20 21 22 23 24-25 26 27 28 29 30 172 ITEM INDEX DISCUSSION (ll) (12) (13) (14) (15) INDEX (16) (17) (18) (19) Do you think impeach- ment deve10pments could have an effect on your life in the near future? Could there be an effect on the cost of living or prices for you? Could there be an effect on your job, Or the job of some- one close to you? Will impeachment events have an effect on your general sat- isfaction with things around you? Are impeachment events the kind of thing you will want to keep up with? POT. INT. Do you think you will be talking about it with friends? With relatives? With peOple at work? Anybody else? INDEX DISCUSSION CODE 0 = no (Go to item 14) l = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (If no also to l = yes item 14, go to 20) l = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes COLUMN 31 32 33 34 35 36' 37 38-39 40 (20) (21) (22) (23) (24) (25) (26) (27) (28) Now just a few final questions... 173 ITEM Is Vice President Ford in favor of impeach- ment? How much is a presi- dential pension currently? Does Senator Griffin favor resignation of the president? Is a simple majority in the House of Repre- sentative sufficient to obtain impeachment? If Nixon is censured does he remain in office? If the president re- signs, would he lose his pension? Does anything else come to mind in con- nection with the im- peachment events? Knowledge index Is knowing about what's happening with impeachment of any use to you? CODE 1 = no; he favors censure at the most 0 = don't know, yes 1 = $60,000 0 = don't know 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = no, he keeps pension 0 = don't know; loses pension 0 = no 1 = yes (write in full answer) 0 = no 1 = yes (write full reply): 174 COLUMN ITEM 41 42 43 44 (29) What was the last grade in school you finished? (30) What work do you do currently? (31) Would you say your age is in the: (32) Record sex of respondent CODE (”NI—‘0 mam» b mun \DQO‘WDWM \Dmfl II II II II II II II II II NH less than 6 some high school finished high school some college finished college graduate work no response retired housewife labor service, operative craftsman, foreman sales, clerical professional, technical official, manager student no response 20's 30's 40's 50's 60's over 60 no response male female Thank you for your time and help today. In order to plete this study, we may need a final, brief talk CORT --:-—! Wltl h you. . . . Would it be all right with you if I called next week, at the same time? Don't call OK, may call Thank you very much, I appreciate it. Good night. 175 353-3237 Office Calls: lst - Key: NA 2nd - B 332-3797 Home 3rd - R 4th - Bissnyenova Project Director Interviewer's name Subject number Respondent phone no. Recommended day... Time Other comments MEDIA NEWS STUDY FOLLOW-UP Hello, I am from the Department of Communication at Michigan State University.... Last week I believe I spoke to you [the lady of the house, the man of the house] about the news 0 O 0 May we take about 5 minutes now, and complete this study?... Thank you. CARD ONE COLUMN ITEM CODE _1_ 1 Card one 1 = NFL _2_ 2 Deck 2 2 3-5 Subject # COLUMN 6 10 11 12-13 14 15 16 17 18 (1) (2) (3) (4) (5) (6) INDEX (7) (8) (9) (10) 176 ITEM These days now, how interested are you in the National Football League strike? Do you think the NFL strike has an effect on your life in any way? Do you think the strike has an effect on the cost of living, on prices for you? Do you think the strike has an effect on your job, or the job of someone close to you? Does the NFL strike have an effect on the enjoyment you get out of watching the game? Is the NFL strike the kind of thing you want to keep up with? INTEREST Have you discussed it with friends?... With relatives? With people at work? Anybody else? INDEX DISCUSSION CODE 0 = not at all 1 = a little 2 = some 3 = a lot 0 = no (GO TO ITEM 5) l = yes, not sure, depends, maybe, don't know 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no (If no also to item 1, go to item 11 on next page) 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes 0 = no 1 = yes COLUMN 19 20 21 22 23 24 25 26 27 28 (ll) (12) (13) (14) (15) (16) (17) 177 ITEM Are veterans going to play in exhibition games in the coming weeks? What is Ed Garvey's role in the NFL strike? How long has the strike been going on? How long is the cooling-off period supposed to last? What is the decision of the Minnesota Vikings regarding the cooling-off period? Can veterans walk out again if agreement is not reached in two weeks? Index Str. Know (11, 12, 16) Index Fac. Know (13, 14, 15) Index Overall Know (ll-16) Does anything else CODE 1 = yes 0 = no, don't know H II Executive Direc- tor of the NFL Player's Assoc- iation Negotiates with the NFL Manage- ment 1 = over 40 days more than a month 0 = other, don't know l = 2 weeks (begin— ning this Wednes- day) 0 = other, don't know 1 = stay on strike (not report to camp) 0 = return to camp, other, don't know 1 = yes 0 = no, don't know 0 = no If YES write in full reply: COLUMN 29 3O 31 32 33 34 35-36 37 38 39 40 41 (l8) (19) (20) (21) (22) (23) INDEX (24) (25) (26) (27) 178 ITEM How interested are you no! in the events following Nixon's resignation? Do you think these deve10pments have an effect on your life in any way? Do these developments have an effect on the cost of living, or prices for you? Do you think these events have an effect on your job, or the job of someone close to you? Do these deve10pments have an effect on your general satisfaction with things around you? Are these events the kind of thing you want to keep up with? INTEREST Have you discussed these deve10pments with friends?... With relatives? With people at work? Anybody else? INDEX DISCUSSION 222 bowl-‘0 l—‘O l—‘O l-‘O I—‘O l-‘O not at all a little some a lot no (Go to item 22) yes, not sure, may- be, depends, don't know no yes no yes no yes no (If no also to item 18 go to item 28 on next page) no yes no yes no yes no yes COLUMN 42 43 44 45 46 47 48 49 50 51 (28) (29) (30) (31) (32) (33) (34) 179 ITEM Will Nixon keep his presidential pension now? How much is this pension? What are the most frequently mentioned names for a possible new Vice President? What Michigan con- gresswoman may be called to join the Ford administration? What is President Ford's Domestic Summit Meeting supposed to deal with? Who is to make a decision for or against further pros- ecution and indictment of Mr. Nixon? Str. 33) Index Know (28, 32, Index Fac. Know (29, 30, 31) Index Overall Know (28-31) Does anything else come to mind in con- nection with the im- peachment events? CODE 1 = yes 0 = no, don't know 1 = 60,000 0 = other, don't know 1 = Bush and Rocke- feller 0 = other, don't know 1 = Martha Griffiths 0 = other, don't know 1 = inflation 0 = other, don't know 1 = Leon Jaworski 0 = other, don't know 0 = no 1 = yes (write in full answer) Thank you very much for your KINDNESS and COOPERATION. Good night. # [‘55 iiltI-IIII'III'I‘J III II‘ 5' ‘g BIBLIOGRAPHY BIBLIOGRAPHY Adams, John B., J.J. Mullen, and H.M. Wilson. Diffusion of a minor foreign affairs news event. Journalism Quarterly, 1969, 46, 545-551. Allen, Irving L., and J.D. Colfax. The diffusion of news of LBJ's March 31 decision. Journalism Quarterly, 1968, 45, 321—324. ANPA News Research Center. News and editorial content and readership of the daily newspaper. A National Survey, Washington, D.C., April 1973. Atkin, Charles. Information utility and information seeking. In Kline and Clarke (eds.), Communication Research. Vol. 2. Information seeking and information process- ing. Sage Publishing Company, 1974. Atkin, Charles. Anticipated communication and mass media information seeking. Public Opinion Quarterly, 1972, 36, 188-199. Atkin, Charles and Bradley Greenberg. Public television and political socialization. Michigan State University report, March, 1974. Atkin, Charles, Lawrence Bowen, Oguz Nayman, and Kenneth Sheinkopf. Quality versus quantity in televised political ads. Public Opinion Quarterly, 1973, 37 (2), 209-224. Atkinson, J.W. and Feather, N.T. (eds.). A theory of achievement motivation. New York: Wiley and Sons, Inc., 1966. Augedal, Egil. Patterns in mass media use and other activi- ties. Acta Sociologica, 1972, 15 (2), 145—156. Stockholm. Bagdikian, Ben H. The information machines. New York: Harper and Row, Inc., 1971. 180 181 Baldwin, Thomas and Bradley Greenberg. A comparison of public and community leader attitudes toward local television programming needs. Journalism Quarterly, 1969, 13 (2), 111-123. Ball, Samuel and Gerry Ann Bogatz. The first year of Sesame street: An Evaluation. Educational Testing Services, Princeton, New Jersey, 1970. Berelson, Bernard, P. Lazarsfeld, W. McPhee. Voting. University of Chicago Press, 1954. Berger, J., M. Zelditch, Jr., Bo Anderson. Sociological theories in progress. Vol 1. Boston: Houghton- Mifflin Co., 1966. Berkowitz, Leonard and Donald Cottingham. The interest value and relevance of fear arousing communication. Journal of Abnormal and Social Psychology, 1960, 37—43. Bernstein, Basil. Elaborated and restricted codes: Their origins and some consequences. In R.D. Hess, et al., (eds.), Conference on develgpment of cross-natIEnEI research on the educatiOn of children and adolescents. CO Project #0-015, University of Chicago, 1964. Bernstein, Basil. Social class and linguistic deve10pment: A theory of social learning. In A.H. Halsey, gt 31., (eds.), Education, economy, and society. New York: Free Press, 1961. BishOp, Michael. Media use and democratic political orien- tation in Lima, Peru. Journalism Quarterly, 1973, Spring, 50 (1), 60-67. Bishop, Michael and Pamela McMartin. Toward a socio-psy- chological definition of transitional persons. Journal of Broadcasting, 1973, 17, 3. Block, Carl. Communicating with the urban poor: An explor- atory inquiry. Journalism Quarterly, 1970, 47 (1), 3-110 Blumler, Jay, J.R. Brown and D.M. McQuail. The social origins of the gratification assoCiated with tele- vision viewing. November, mimeo, 1970. Blumler, Jay G. and Denis McQuail. Television in politics. University of Chicago Press, 1969. 182 Bogart, Leo. The spread of news in a local event: A case history. Public Opinion Quarterly, 1950, 14, 769-772. Bower, Robert T. Television and the public. New York: Holt, Rinehart, and Winston, Inc., 1973. Budd, Richard, 33 31. Regularities in the diffusion of two major news events. Journalism Quarterly, 1966, 43. Buss, Linda. Motivational variables and information seeking in the mass media. Journalism Quarterly, 1967, 44 (1), 130-133. Campbell, A. and P. Converse. (eds.). Indicators of social change. New York: Russell Sage, Inc., 1970. Cantor, Joan. Transfer of stimulus pretraining to motor paired-associate and discrimination learning tasks. Vol. 2. In Lipsitt, L., et 31., (eds.), Advances in child development and behEVior. New York: Academic Press, 1965, 19-58. Cantril, H. The pattern of human concerns. New Brunswick, N.J.: Rutgers University Press, 1965. Carlsson, Gosta. Change, growth and irreversibility. American Journal of Sociology, 1968, 73. Chaffee, Steven, et g1. Mass communication and political socializatIEn. Journalism Quarterly, 1970, 47, 647-659. Chaffee, Steven. The interpersonal contact of mass communica- tion. In Kline and Tichenor, (eds.), Current perspec- tives in mass communication research. Vol. 1. Sage Publishing Co., 1973. Chaffee, Steven and Jack McLeod. Individual versus social predictors of information seeking. Journalism Quarterly, 1973, 50 (2), 237-245. Chaffee, Steven and C.K. Atkin. Parental influences on adolescent media use. ABS, 1971, 14 (3), 323-340. Chaffee, Steven, gt 31. Experiments on cognitive discrep- ancies and communication. Journalism Monographs, 1969, 14. Chaffee, Steven. Parent - child similarities in television use. AEJ Conference, Washington, D.C., 1970. 183 Chaffee, Steven and Jack McLeod. Communication as social- ization: Two Studies. AEJ Conference, Boulder, Colorado, 1967. Clarke, Peter. Parental print use, social contact about reading and use of the print media by teen-age boys. Pacific Center, American Association for Public Opinion Research, Napa, California, 1969a. Clarke, Peter. Parental socialization values and children's newspaper reading. Journalism Quarterly, 1965, 42, 539-546. Coleman, James, gt it. Equality of educational opportunity. 0.5. Government Printing Office, Washington, D.C., 1966. Dervin, Brenda. The information needs of urban residents: A conceptual context. A working paper, AEJ Conference, Port Collins, Colorado, 1973. ° Dervin, Brenda. Communication behaviors as related to information control of black low income adults. Ph.D. dissertation, Michigan State University, 1971. Dervin, Brenda and Bradley Greenberg. The communication environment of the urban poor. Project CUP, Michigan State University report, #15, 1972. Douglas, D., B. Westley and S. Chaffee. An information campaign that changed community attitudes. Journalism Quarterly, 1970, 47, 479-487. Douvan, Elisabeth and Joseph Adelson. The adolescent experi- ence. New York: Wiley and Sons, 1966. Dunlop, R.A. The emerging technology of information utilities. In Sackman, H. and N. Nie, (eds.), Information utility and social choice. Montvale, New Jersey: AFIRS Press, 1970. Emmet, Brian. A new role for research in broadcasting. Public Opinion Quarterly, 1969, 32, 654-665. Etzioni, A. Cable television: Instant shopping or partici- patory technology. New York: Social Policy Co., 1973. Fathi, Asghar. Problems in develOping indices of news value. Journalism Quarterly, 1973, 50 (3), 497-508. 184 Fitzsimmons, Steven and H.A. Osburn. The impact of social issues and public affairs television documentaries. Public Opinion Quarterl , 1969, 32, 379-397. French, Elisabeth G. Some characteristics of achievement motivation. Journal of Experimental ngchology, 1955, 50, 232-236. Genova, B.K.L. The knowledge gap phenomenon. A paper, ICA, 1975. Gibson, Eleanor. Perceptual development. In H. Stevenson (ed.), 62nd Yearbook of the National Society for the Study of Education and Child Psychology. University of Chicago Press, 1963. Glessing, Robert and W. White. (eds.). Mass media, the invisible environment. Chicago, SRA, Inc., 1973. Goldhammer, Herbert. (ed.). The social effects of commun- ication technology. Rand Report R-486, RSF, 1970. Greenberg, Bradley and J.R. Dominick. Race and social class differences in teen-agers' use of television. Journalism Quarterly, 1969a, 13, 331-344. Greenberg, Bradley and Hideya Kumata. National sample pre- dictors of mass media use. Journalism Quarterly, 1968, 45 (4), 641-646. Greenberg, Bradley S. Mass communication and social behavior. Michigan State University. Mimeo, 1973. Greenberg, Bradley and Thomas Baldwin. A comparison of public and community leader attitudes toward local television programming needs. Journal of Broadcast- lpg, 1969, 13 (2). Greenberg, Bradley. Person-to-person communication in the diffusion of news events. Journalism Quarterly, 1964b, 41, 489-494. Greenberg, Bradley, James E. Brinton, and Richard S. Farr. Diffusion of news about an anticipated news event. Journal of Broadcasting, 1965, 9, 129-142. Greenberg, Bradley. Diffusion of news of the Kennedy assassination. Public Opinion Quarterlyl 1964a, 28, 225-232. 185 Greenberg, Bradley, gt El. Use of the mass media by the urban poor. New York: Praeger Publishing Co., 1970. Hanneman, Gerhard and B. Greenberg. Relevance and the dif- fusion of minor and major foreign affairs events. Journalism Quarterly, 1973, 50, 443-437. Hill, Richard and Charles Bonjean. News diffusion: A test of the regularity hypothesis. Journalism Quarterly, 1964, 41, 336-342. Hyman, Herbert H. and Frank B. Sheatsley. Some reasons why information campaigns fail. Public Opinion Quarterly, 1947, 11, 412-423. Jaffe, J. Language of the dyad a method of interaction analysis in psychiatric interviews. Psychiatty, 1958, 21, 249-258. Jencks, C. and Reisman, D. The academic revolution. New York: Doubleday and Co., 1968. Johnson, Norris. Television and politization: A test of competing models. Journalism Quarterly, 1973, 50, 447-455. Katz, D. The functional approach to the study of attitudes. Public Oplnion Quarterly, 1960, 24, 163-204. Katz, Elihu, Jay G. Blumler, and Michael Gurevitch. Utili- zation of the mass media by the individual. New York, May. Mimeo, 1973. Katz, Elihu, pt 21- On the use of mass media for important things. Tokyo, August. Mimeo, 1972. Katzman, Nathan. The impact of communication technology: Some theoretical promises and their implications. NIH Colloquium in Information Science, Stanford University, 1973. Kendler, Tracy. DevelOpment of mediating responses in children. Society for research in child development. Monograph #28, 1963, 2, 33-48. Kernan, Jerome and Richard Mojena. Information utilization and personality. Journal of Communication, 1973, 23, 3, 315-328. Key, V.O. Public opinion and American democracy. New York: Knopf Publishing Co., 1961. 186 Kline, F.G. and P.J. Tichenor. Current perspectives in mass communication research. Beverly Hills: Sage Publishing Co., 1972. Kline, Gerald, gt El. Family and peer socialization and . autonomy related to mass media use. Paper presented to International Association for Mass Communication, Constance, Germany, 1970b. Krieghbaum, H. Science and the mass media. New York: University Press, 1968. Krieghbaum, H. The news and the public. New York: University Press, 1958. Levy, Sheldon. How population subgroups differed in their knowledge of six assassinations. Journalism Quarterly, 1969, 46, 685-698. Lyle, Jack and H.R. Hoffman. Children's use of television and other media. In E.A. Rubinstein, et al., (eds.), Television and social behavior. Washingtafi, D.C., NIMH, Vol. 4, 1971a. McClelland, David. The achieving society. New York: Free Press, 1961. McClelland, David. Toward a theory of motive acquisition. American Psychologist, 1965, 20, 321-333. McClure, Robert and Thomas Patterson. Television news and political advertising. Communication Research, 1974, 1 (1), 3-32. McCombs, Maxwell and David Weaver. Voter's need for orien- tation and use of mass communication. Paper presented to ICA, Montreal, 1973. McCombs, Maxwell. Mass communication in political campaigns: Information, gratification, and persuasion. In Kline and Tichenor, (eds.), Current perspectives of mass communication research. Vol. 2. Sage Publishing Co., 1973. McCombs, Maxwell, Eugene Shaw, Donald Shaw, and L. Edward Mullins. Working papers on agenda setting. Series #1. University of North Carolina, Chapel Hill, 1973. McCombs, Maxwell and Donald Shaw. The agenda-setting power of the press. Public Opinion Quarterly, 1972, 36, 176-187. 187 McGuire, William. The nature of attitudes and attitude change. In G. Lindzey and E. Aronson, (eds.), Handbook of social psychology. Vol. 3, 136-314, 1968. McLeod, Jack and Daniel Wackman. Family communication: An updated report. AEJ Conference, Boulder, Colorado, 1967. McLeod, Jack and Steven Chaffee. The construction of social reality. In J. Tedeschi, (ed.), The social influence processes. Chicago: Aldine-Atherton, 1971. McLeod, Jack, pt 21' The mass media and political informa- tion in Quito, Equador. Public Opinion Quarterly, 1969, 32, 575-587. McLeod, Jack and J. Swinehart. Satellites, science and the public. Ann Arbor, Michigan Survey Research Center, 1960. McNelly, John. Cosmopolitan media usage in the diffusion of international affairs news. Journalism Quarterly, 1968, 45, 329-332. McNelly, John. The broadcast media and the redistribution of information about population and family planning. National Broadcast Training Centre, Kuala Lumpur, Malaysia, 1972. McNelly, John and Julio Molina. Communication, stratifica— tion and international affairs information in a develOping urban society. Journalism Quarterly, 1972, 49, 316-326. McNemar, Quinn. Psychological statistics. New York: John Wiley & Sons, 1962. McPhee, William, gt al. A model for simulating voting systems. In J. Dutton and W. Starbuck, (eds.), Computer simulation of human behavior. New York: Wiley and Sons, 1971. Maslow, A.H. The need to know and the fear of knowing. Journal of General Psychology, 1963, 68, 11-24. Medalia, Nahum and O.N. Larsen. Diffusion and belief on a collective delusion: The Seattle windshield pitting epidemic. American Sociological Review, 1958, 23, 180-186. 188 Mesthene, Emmanuel. Technological change. London, The New English Library, Ltd., 1970. Miller, G.A. The magical number seven, plus or minus two: Some limits on our capacity for processing informa- tion. Psychological Review, 1956, 63, 81-97. Miller, G.A. What is information measurement? American Psychologist, 1953, 8, 3-11. Miller, James G. Living systems: Basic concepts. Mental Health Research Institute, University of Michigan, Ann Arbor, 1970. Miller, James G. Psychological aspects of communication overloads. In R. Waggoner and D. Carek, (eds.), Communication in clinical practice. Boston: Little, Brown and Co., pp. 201-224, 1964. Miller, James G. Adjusting to overloads of information. In David Rioch and Edwin Weinstein, (eds.), Research Publication - Association for research in nervous and mental disease. Vol. 42, 87-100, 1964. Milbrath, Lester W. Political pgrticlpation: How and why do people get iiVolved in_politiCS? Chicago: Rand McNally, 1965. Moore, Wilbert. Social change. Englewood Cliffs, N.J.: Prentice Hall, Inc., 1963. Moynihan, Daniel P. Sources of resistance to the Coleman report. Harvard Educational Review, 1968, 38 (l). Nayman, Oguz, gt al. The four-day work week and media use: A glimpse ii the future. Journal of Broadcasting, 1973, 17, 3, summer. Nie, Norman H., Dale H. Bent, and C. Hadlai Hull. Statis- tical package for the social sciences. New York: McGraw Hill, 1970. Oettinger, Anthony and Nukki Zapol. Will information tech- nologies help learning? The Annals of AAPSS, 1974, March issue, 116-127. O'Keefe, Timothy and B.C. Kissel. Visual impact: An added dimension in the study of news diffusion. Journalism Quarterly, 1971, 48, 298-303. O'Keefe, Timothy. The first human heart transplant: A study of diffusion among doctors. Journalism Quarter- ly, 1969, 46, 237-242. 189 Olien, C.N., gt gl. Research on purposive communication: An evaluational model. Fort Collins, Colorado, August. Mimeo, 1973. Parker, Edwin B. Technological change and the mass media. Standord Unviersity. Mimeo, 1971. Parker, E.B. and W.J. Paisley. Patterns of adult informa- tion seeking. Stanford Institute of Communication Research, 1966. Reiss, Albert. Occupations and social status. New York: Free Press, 196I. Robinson, John. National election study. Ann Arbor, Survey Research Center, 1968. Robinson, John. Public information about world affairs. Ann Arbor, Survey Research Center, 1967. Robinson, John. The audience for national television news programs. Public Opinion Quarterly, 1971, 35, 403—405, Fall. Robinson, John. Television and leisure time: Yesterday, today and (maybe) tomorrow. Public Opinion Quarterly, 1969, 33, 210-222. Robinson, John. Toward defining the functions of tele- vision. In E. Rubinstein, gt gl., (eds.), Tele- vision and social behavior. NIMH Publication, Vol. 4, 1972. Robinson, John. World affairs and media exposure. Journalism Quarterly, 1967, 44 (1), 23-31, Spring. Rogers, Everett and F. Shoemaker. Diffusion of innovations. New York: Free Press, 1971. Rosen, B.C. Race, ethnicity and the achievement syndrome. American Sociological Review, 1959, 24, 47-60. Rosen, B.C. and R.G. D'Andrade. The psychological origins of achievement motivation. Sociometry, 1959, 22, 185-218. Rosen, Ephraim. Factor analysis of the need for cognition. Psychological Reports, 1964, 15, 619-625. Rosengren, Karl E. Mass media consumption as a functional alternative. University of Lund, August. Mimeo, 1971. Ilillilli‘l’lv lili!‘ ‘ Ill 190 Rosengren, Karl E. Uses and gratifications: An overview. University of Lund, October. Mimeo, 1972. Rosengren, Karl E. News diffusion: An overview. Journalism Quarterly, 1973, 59, 83-91. Rota, Joseph. News diffusion. Department of Communication, Michigan State University. Mimeo, 1973. Rummel, R.J. Applied factor analysis. Evanston: North- western UniVersity Press, 1970. Samuelson, Merrill, Richard F. Carter and Lee Ruggels. Education, available time and use of mass media. Journalism Quarterly, 1963, 40, 491-496. Schramm, Wilbur. Information theory and mass communication. Journalism Quarterly, 1955, 32, 131-146. Sears, David O. and J. Friedman. Selective exposure to information: A critical review. Public Opinion Quarterly, 1967, 31, 194-213. Sewell, W.H. and V.P. Shah. Parents' education and children's educational aspirations and achievements. American Sociological Review, 1968, 33, 2, 191-209. Sherif, M. and C. Sherif. An outline of social psychology. New York: Harper and Row, 1956. Spitzer, Stephan P. and N.K. Denzin. Levels of knowledge in an emergent crisis. Social Forces, 1965, 44, 234-237. Star, Shirley and Helen Hughes. Report of an educational campaign: The Cincinnati plan for the United Nations. American Journal of Sociology, 1950, 50, 389-400. Stempel III, Guido H. Increasing reliability in content analysis. Journalism Quarterly, 1955, Vol. 22, 4, Stern, A. Time magazine article, October 18, 1971. Strodtbeck, Fred. Family interaction, values and achieve- ment. In D. McClelland, 22.21-r (eds.), Talent and society. Princeton, New Jersey: Van Nostrand, pp. 135-194, 1958. Tannenbaum, Percy and Bradley Greenberg. Mass communication. Annual Review of Psychology, 1968, 19. 191 Tichenor, P.J., G.A. Donohue, C.N. Olien. Mass communica- tion research: Evolution of a structural model. AEJ Conference, Carbondale, I11., August. Mimeo 1972. Tichenor, P.J., G.A. Donohue and C.N. Olien. Mass media flow and differential growth of knowledge. Public Opinion Quarterly, 1970, 34, 159-170. Tichenor, P.J., G.A. Donohue and C.N. Olien. Science, mass media and the public, January. Mimeo, 1971. Tichenor, P.J., J.M. Rodenkirchen, C.N. Olien and G.A. Donohue. Community issues, conflict and public affairs knowledge. In P. Clarke, (ed.), New models for communication research. Vol. 2. Sage Publishing Co., Beverly Hills, California, 1973. Tipton, Leonard, Roger Haney, Jack Baseheart and William Elliott. Media agenda—setting in a state campaign. Fort Collins, Colorado, August. Mimeo, 1973. Tipton, Leonard. Effects of writing tasks on utility of information and order of seeking. Journalism Quarterly, 1970, 47, 309-317. Troldahl, Verling and Roy Carter, Jr. Random selection of respondents within households in phone surveys. Journal of Marketing Research, May 1964, 71-76. Troldahl, Verling and Don Costello. Media exposure patterns and interpersonal communication behavior of teen-agers. AEJ Conference, Iowa City, Iowa, 1966. Usdansky, G. and L.J. Chapman. SchiZOphrenia-like responses in normal subjects under time pressure. Journal of Abnormal and Sociological Psychology, 1960, 60, 143-146. Wade, Serena and W. Schramm. The mass media as source of public affairs, science and health knowledge. Public Opinion Quarterly, 1969, 33 (2). Walker, Helen M. and Joseph Lev. Statistical inference. Holt, Rinehart & Winston, 1953. Westley, Bruce. The functions of public communication in the process of social change. Michigan State Uni- versity, April. Mimeo, 1966. Westley, Bruce and Lionel Barrow, Jr. An investigation of news-seeking behavior. Journalism Quarterly, 1959, 36, 431-438. 192 Westley, Bruce and Steven Chaffee. An information campaign that changed community attitudes. Journalism Quarterly, 1970, 47, 479-487. White, Sheldon. Evidence for a hierarchical arrangement of learning processes. Vol. 2, 187-220. In Lipsitt, gt_gl., (eds.), Advanced in child development. New York: Academic Press, 1965. Winterbottom, Marian. The relation of need for achievement to learning experiences in independence and mastery. In J. Atkinson, (ed.), Motives in fantasy, action and society. Princeton, New Jersey: Van Nostrand, 1958. nICHIan STATE UNIV. LIBRnRIEs )IHI(I!)I"(I)(IIWIIM)I)I)IHNINHIHWHIHIIHI 31293101582843