f7 ‘ A “7“" _ 7 ‘9 ‘ J. E1 19411111": . . H‘b‘ ‘ H‘L’l‘tfi‘ 1 I. I“! . ‘.‘ g r" 1 .3 ('1 -,C 4-. M 3“. .—. - u. . A ~ 09-- . .u 233:; . EC ~ 0 m “.4533; -...... F?“ ‘x .m': ..._._.:,:.__ 1..- .._._. 1- MY my - .9... . . ‘ “vi ‘ «v- .w at» an o. ”'33“:— m * [A a.” u ~4- _, 3.3:. N... ~. I ~—. 1 1' 11? 11 \ 'l I. I :14 ' - £117.55 '1Hi' ' . 11 2:111? of, I . fl ,‘ I A 1.‘ 1 I . i '10". - .. v u x . '3 "':'.'.".~ 0 u . E if ~ 5 l ‘i z'!’ ‘t . .' :‘ . :33: "g '- g: .3. " r; ’33? ..-:1‘—t' A 1<»-<'1.-.'f . V . . 1 ,w.-4-.~\ .. , {#510 o . . - u ”v.-. . . 'l 1" IV .. A " ' - . u :5" 73‘ by”. -. 4—5. m‘ .31.; ‘3‘" ' . 1.. - -« u c .-‘4~ o .,. .n'. . ‘4’; ' n v ‘ "—4 . -. .DJ‘ ‘ - _ . .. .21.’ - - ..- .. .' u < - ”1-0:. . . . ' "h p.” . u v‘ c on... ‘. Mo; ' o . k . 5: 1'4 u - .- t 1’ 315 . . 1r '14 zl‘i'ifir": v0 " .111 . ' w”- . ’1? 51115121511 15.13311, , ,5 1:111:111111111,1,,g g. 1‘ '1 ' ' 5' In fir“ i: :. . 1. : . w . 4.. '1‘ ”:1: '1:"\.<'nL' ‘ ‘ . , 1 :1” ‘. I ' < . n. ‘1 L. ‘ , 2 'l‘iiflu ‘qh ,s .. 11:11::9—1‘1EJ- -.’ '1 3121'. . 1.11:1 ' 1111111711“ M 42111515311111-‘11 .1“ 1' =1 2' is . r 1111' ~ ‘ L 1 ' ' ”‘5 .,l rim-e" Y .T'_!' I»: . «z -t ‘f, 5 | ~ 1‘” , - . v 1 1 t v . . I ' o :1 u’fiiv‘j A. 1:; -l . 11.7 '3 “I :11:- :‘ ' ~" a it . .. . vm‘ . . ‘7‘ .‘LJ‘ ._. .,. ' . ' $51 21:. 1 . . . :6" 9.3315“; .11 I . l _ ~ {115%}? RE : 1“ n fl" :, ’ '5'" i" 1 ' . ‘ “3113:5115" "f1 ., r. - 1 L 5 ......... ‘: . '-~".-_ 3 1...? 1 ‘ tfréf fiig‘iéigfg“ ‘ ' ' .:?"11h=112}3z’2.' 2': -" fr- 13 t ' 1111,1113 -‘ u; - 1 n n. "tn-5 :1. ._.‘. ‘ :10 Z. ‘? ' C“ I ' ~. 1:" 5‘ . iiéi‘{s}‘li::1iifit"1" " ‘ 5:11 41‘}. “1 kg, ‘3 J! 5' -. -': . ..,‘-.1‘—- - at: ' -"' :1' “3 "" 'E‘»'-! . :' v ’ 'l.‘ '1' .- Y?!“ .‘h‘ ”1'5? . ‘v‘ru Law: ' ’i- ' {.13 @1333" 11w. 1:11., :11'if511aw g'HiMM’ :19 ‘- '11:” 1' “VJ-11.; l’gz’F' ”If? £11331 4:51":.-"’\bEr'--r"‘#ig21“"7"'3:fi19'1 “fl. A {‘31 - 1 3; {jai-lfgafiuggymtg};:5 mpmggifi 1’;$"i{1%£ 14121;) p 1:73" iifi‘f't" ' ‘9‘" 1‘5 " 1- L. 1"“5'33; 1 ’1 ' f. ‘-:‘.' z 1’: - 11111 1122111121 1111-1131141 1 ,.._,,:'-- 11-411411. "an-.41 61» .':- «as: . " 1 - "».~s.n&-1-1,':J’~ «11.03% 'z‘ o" in w am” 1 ‘11va fl“ 1?": :1. 1 3";“.~...‘,'r¢. '1‘ .11" .-1_- ‘ ”.1" 7*“ ‘ i] 1‘ ' $1,»: ‘ A I)” .3“ J. 3.1! .1193 . "fig-.PH'Jng'W n"! "J -8 51151121111115111115111 . JHMfififm , 'r 'f . k ., . : *2?wa 7-213 ‘11-'44!”- zi' 31...},1 415931}? ' ~ 1* “- -'." 7.1 "'51 1;?‘"{:‘;’ " 41:11 ‘1“ ' . 1" 1,13; 11:1 - ' " - .7... 1: ,- ”151. '11;‘~“f1. 1’ (:71? ". 1 ‘ 1J1. 1.1 ’1 15%" M15?" 2‘ 3&3. '- .h I . - M1 :': .1 .1 41113-1111 ‘11»: I hlh'u 73013159! 11116111111111111111111 31293 00551 4538 I. LIBRARY Michigan state University This is to certify that the thesis entitled POLITICAL NARROWCASTING: A STUDY OF THE RELATIONSHIP BETWEEN TELEVISION PROGRAM AND CABLE CHANNEL VIEWING AND POLITICAL ATTITUDES AND BEHAVIORS presented by Mitchell Steven Shapiro has been accepted towards fulfillment of the requirements for M.A. d . Telecommunicat ion egree 1n W W Major professor Date mekjayé'flid? 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution IVIESI.J RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from -_—I-_ your record. FINES will be charged if book is returned after the date stamped below. u" ’v' ”#0 “£83. .15 3.1994 MAY 1.6 1995‘; 1 1 POLITICAL NARROWCASTING: A STUDY OF THE RELATIONSHIPS BETWEEN TELEVISION PROGRAM AND CABLE CHANNEL VIEWING AND POLITICAL ATTITUDES AND BEHAVIORS BY Mitchell Steven Shapiro A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Telecommunication 1988 ABSTRACT POLITICAL NARROWCASTING: A STUDY OF THE RELATIONSHIPS BETWEEN TELEVISION PROGRAM AND CABLE CHANNEL VIEWING AND POLITICAL ATTITUDES AND BEHAVIORS BY Mitchell Steven Shapiro Cable television's growth has greatly increased the number of video channels available to a majority of American homes, increasingly fragmenting the television audience. This produces both problems and opportunities for political communicators who previously could reach nearly all American homes via the three broadcast networks. One possible response to this trend is "political narrowcasting," the targeting of specific political messages to relatively narrow audience segments, e.g., viewers of particular programs, program types or cable channels. This study, analyzing data from a survey of 321 voting- age Lansing, MI, residents, explores the relationship between broadcast and cable viewing patterns and political attitudes and behaviors. It discovers that some audience Segments differ from others in terms of political characteristics that are significant to political campaigns. Based on such findings, implications are drawn regarding campaign media buying and message design strategies and suggestions are made for incorporating similar analyses in campaign polls. TABLE OF CONTENTS LIST OF TABLES .0...0.0.0..........‘OOIOO......OOOOOOOOOO.0v I. INTRODUCTION ........ ................................. 1 II. LITERATURE REVIEW .......... .......................... 5 Program Types ..... . ..... . ........ ...... ..... ..... 5 Geodemographics ....... ......... ............ . ..... 8 Television Audience Segmentation ..... ..... . ..... 10 Age and Gender as Predictors of Political Characteristics .. ........... ........... ......... 15 III. RESEARCH QUESTIONS ... ..... ....... ..... . .......... ... 20 Iv. METHODOLOGY ......OOOOOOOOOO...OOOOOOOOOO0.0000...... 23 Sampling and Data Collection ........... . ........ 23 Measurement and Analysis ........................ 24 V. RESULTS ............. ................... . ............. 37 Bivariate Analysis . .................. . ............ 37 Multivariate Analysis ........ . .................... 44 VI. DISCUSSION .............. . ............................ 56 Program and Cable Channel Types ................... 56 Viewing Patterns and Political Characteristics: Implications for Political Narrowcasting . .................... 56 iii VII. SUMMARY AND RECOMMENDATIONS ............ ............. 67 Recommendations for Future Research .............. 71 APPENDIX: QUESTIONNAIRE ......... ...... ......... ....... ..... 74 LIST OF REFERENCES. ...... ... ........ ......... ............... 89 iv Table Table Table Table Table m UI h U N H .0 Table Table 7: Table 8: Table 9: Table 10: LIST OF TABLES Prime Time Viewing Factor Loadings............ Political Activity Factor Loadings ............ Political Trust and Efficacy Factor Loadings .. Foreign Policy Issue Factor Loadings .......... Law and Order Attitude Factor Loadings ........ Television Viewing and Political Characteristics: Correlations and partial correlations controlling for age and gender .................................... Correlations of Age and Gender to Political Characteristics .......... Correlations of Age and Gender to Teler-Sion vieWing .....OOOOOOOOOOOO‘OOOOO. ..... Partisanship, Partisanship-Ideology and Television Viewing (F values for ANOVA) ....... Targeting Voter Segments 25 28 34 35 48 54 54 I. INTRODUCTION In recent years, technological and regulatory changes have brought about a dramatic increase in the number of video channels available to the producers and consumers of political communications. While desirable economically because it offers video consumers increased options, and politically because of its potential to expand the marketplace of ideas, this proliferation of channels has made the job of those who transmit political messages to the American public more complex and perhaps more difficult. It has done so by fragmenting television's traditionally mass audience among multiple channels and thereby reducing the share of that audience reachable by the three broadcast networks which, in the past, have been able to collectively provide political campaigns with direct access to nearly all American homes (Brotman, 1981). 'fWith cable penetration surpassing 50% in 1987 (Multichannel News, December 9, 1987), and'VCRs in more than 40 percent of American homes, (Lachenbruch, 1987) there has been growing interest among researchers and campaign managers in the concept of "narrowcasting." Generally, "narrowcasting" refers to the targeting of particular video, audio, text or data transmissions to relatively narrow audience segments, in contrast to "broadcasting", which is associated with more limited channel availability and larger audiences per channel.’ Targeted political communications have become increasingly commonplace and important in political campaigns, with candidates making heavy use of targeted direct mail, often based on geodemographic analyses such as the Claritas Cluster System, C.A.C.I's ACORN, or Market Opinion Research's Custom Targeting (see Robbin, 1980: Mockus, 1980: Atlas, 1981: Schneider, 1983; Kramer and Schneider, 1985), as well as occasional use of videoconferences, videotapes and cable television (Meadow, 1985).‘/Narrowcasting is already a well established phenomenon in radio, with its numerous specialized formats, and this characteristic has been taken advantage of by some political candidates. For example, congressional candidates in Los Angeles have aired messages of special interest to black and Hispanic voters, and to voters within specific age groups (Brotman, 1981)f While cable television has so far been largely ignored by most political campaigns, some systems, such as one in Austin, Texas, have been heavily used for political advertising in recent local elections (Paley and Moffett, (1984). Probably the most compelling example thus far of a politician's use of cable television to his advantage is the successful 1982 campaign of Representative Barney Frank (D-Mass), who tailored commercials to a key Portuguese population in his district and aired them on the local Portuguese cable channel (Power, 1987). At the national level, cable, now in more than half of the nation's homes, is expected to play an increased role in presidential campaigns. Already media buyers for Republican presidential candidates Bob Dole and George Bush have told reporters that they see cable as.a good buy and will definitely be incorporating it into their 1988 campaign media strategies (Cablevision, December 7, 1987, p. 170). This study is intended to provide information that can lead to greater use of targeted political communications by exploring the narrowcasting concept as it relates to the video medium, seeking to identify politically relevant distinctions between viewers of specific types of television programs and cable channels. It first addresses the question of whether program or channel "types" exist. It then investigates the extent to which the viewers of these various program or channel "types" (as well as of specific programs and channels) can be identified in terms of unique political characteristics. This is done first without controls, and then controlling for age and gender, the two demographic variables used to segment ratings data. The demographically controlled analysis is intended to reveal those campaign-relevant audience segments which ratings data cannot adequately identify. In a secondary analysis, the relationships of age and gender to both political characteristics and viewing patterns are explored in terms of their implications for a ratings—based political narrowcasting strategy. It is hoped that the study will generate findings useful to political communicators who will be designing and disseminating political messages in an increasingly multi- channel environment. II. LITERATURE REVIEW Four general areas of research are reviewed in this chapter. First, studies which have attempted to develop television "program types" are reviewed in order to provide a substantive and methodological context for our own exploration of such a typology. Next we examine ”geodemographic” analysis, the communication targeting methodology which has been most heavily used in recent political campaigns. We then focus on television audience segmentation studies with relevance to political communications, followed by a brief discussion of "narrowcasting" as it is practiced by today's political campaign consultants. Finally we examine data on the relationship of age and gender--the two segmentation variables included in program ratings data which political communicators usually rely on to target their televised messages--to political characteristics. Program Types There are two basic research approaches to the identification of program types, a priori and empirical. Since an empirical method is used in this study, our discussion will focus on empirically derived typologies. The empirical approach to identifying program types has been developed mainly by researchers affiliated with business schools who sought empirical verification for the existence of program types through analyses of viewing data and respondents' ratings of programs (Gandy, 1984). Most of [these studies have used a form of factor analysis to identify underlying program dimensions, and most which have been published date back to the 19605 and early 70s. Kirsch and Banks (1962), applying factor analysis to viewing data from diaries, discovered six program factors-- ABC Westerns, CBS situation comedies, Music and Variety, ABC Action, and NBC Westerns. Ehrenberg (1968), in reviewing this analysis and one done by Swanson (1967) with similar results, pointed out that the correlations among programs are due not only to program content (type of show), but also to other influences such as time of day, day of the week, and network. For example, he concluded that the two major factors which Swanson identifies in terms of show types actually represent the ABC and NBC networks, and pointed to similar evidence of network loyalty in Kirsch and Banks' analysis, which yielded separate ABC-Western, NBC-Western, ABC-Action and CBS-Action factors. William Wells (1969) combined commercial ratings data with preference data from respondents in an analysis which concluded that the influence of program content was greater than that of time and channel as a predictor of viewing trends. In a later study, Dennis Gensch and B. Ranganathan (1974), applying a factor analysis to data from a national marketing survey, discovered eight underlying dimensions of eighty-one programs controlling for scheduling and demographic factors. These eight factors were as follows: 1. Movie factor 2. Action shows 3. ”Light entertainment" shows (e.g., Monkees, Carol Burnett, Jerry Lewis) 4. 4 Shows emphasizing nostalgia and traditional values (e.g., Andy Griffith, Green Acres, Lawrence Welk, Walt Disney) 5. Peyton Place factor, including its two weekly segments 6. The Dating Game and the Newlywed Game shows 7. Westerns 8. Variety shows In an analysis of data on 7000 British adults collected in a Leo Burnett Lifestyle Research study, Goodhardt, et. al, (1987) found six program types. Analyzing the responses to questions about specific programs which inquired whether respondents: "Really like to watch it": "Watch it only because someone in my family watches it”: ”Watch it when there's nothing better" or: "Don't watch it", the study derived the following typology: 1. Sports 2. Current Affairs . Light Entertainment . Adventure . Children's . An unnamed category which Leo Burnett refers to as "cult" programs. A review of these published studies provides some indication of the typologies which may emerge in such factor analytical studies of program types. However, most are fairly dated--for example, variety shows and westerns are rare in prime time today--while the only recently published study focuses on British audiences. In any event, as Gandy notes: Program types generated by factor analysis cannot be considered to be universal, or, for that matter, stable over time with the same population...we should expect factor structures to change over time as both tastes and program schedules will change.(Gandy, 1984, p. 237) Geodemographics Geodemographics was developed as an analytical tool utilizing census data, often supplemented by polls and election returns, to derive voter segments based on census districts known as Block Groups/Enumeration Districts (BG/EDs), each containing about 900 persons. Claritas' Cluster System, a pioneer in geodemographic analysis, uses factor analysis to reduce a large number of census variables into a smaller number of variables for easier analysis. Then, using cluster analysis, the 250,000 BG/EDs in the country (or the 36,000 5-digit ZIP Code areas) are divided into approximately forty relatively homogeneous groups or clusters. Each BG/ED is assigned to a particular cluster based on its value for the factors previously generated. The end result is that each cluster is made up of census areas from across the country in which people with similar demographic characteristics reside (Atlas, 1981). The use of this multivariate approach allows political communicators to develop geographically-based targeting strategies which transcend the "one-dimensional pigeon holes such as 'women,‘ 'blacks,‘ 'blue- collar,’ or 'senior citizens'" traditionally relied upon in political campaigns (Robbin, 1980). Since its emergence as the "hottest fad” in the 1982 election (Schneider, 1983), geodemographic analysis has evolved to incorporate data on political characteristics from surveys and election returns in addition to census data. Among the methods used by this new generation of research, such as Market Opinion Research's Custom Targeting, is the inclusion of partisanship information (not included in census data) in a cluster definition by establishing the correspondence of political precincts to .census units in order to estimate the partisan composition of the latter based on precinct election returns. Other approaches include survey driven clusters which allow clusters to be defined in terms of campaign-relevant variables such as candidate preference, turnout, fundraising and major campaign issue positions (Schneider, 1983). Television Audience Segmentation As valuable as geodemographic analyses have been in political campaigns, their value to a campaign's television media buyer is limited. While it may provide some information useful in placing spot buys with specific stations or cable systems to the extent census areas correspond to broadcast coverage or franchise areas, geodemographics reveals nothing about which programs or channels are best suited to carry particular messages to particular audiences. Television audience studies that measure politically relevant variables are needed to free media buyers from reliance on the age and gender breakdowns found in commercial ratings data. However, because audience studies are usually sponsored by commercial advertisers whose interest in them tends to be more long-term than that of political campaigners, very few such studies have paid much attention to viewer characteristics of importance to the latter. For example, one study of audience segments examined the relationship between "audience interests" and program viewing behavior, using cluster analysis to derive fourteen "interest segments" from a nationwide sample (Frank and Greenberg, 1980). Three of the segments were predominately male (mechanics and outdoor life, money and nature's products, family and community centered); four were 10 primarily female (elderly concerns, arts and cultural activities, home and community centered, and family integrated activities): three were composed mainly of young people (competitive sports and science/engineering, athletic and social activities, and indoor games and social activities): while four were demographically mixed (news and information, detached, cosmopolitan self-enrichment, and highly diversified). The study compared the different clusters' viewing of nineteen a-priori program types using an F-ratio statistic for each of the nineteen program types. Though all of the nineteen tests were significant at the .005 level, it is difficult to imagine how these "interest" categories could have very much significance to a political campaign's message designers and media buyers. Another, more relevant, multivariate analysis of television viewing patterns is the "values and lifestyles" research undertaken by SRI International. SRI's VALS program uses a discriminant analysis of "values and lifestyles" variables for a nationwide sample to develop a typology consisting of four major groups-~"need-driven", ”outer-directed", "inner-directed” and "integrated" individuals--which are further subdivided into nine lifestyle groups (Mitchell, 1983). Though focused primarily on commercially relevant variables, e.g., consumption patterns, SRI's research does include a number of politically significant variables, including partisan and 11 ideological affiliation, political trust, and attitude toward military spending. VALS researchers have discovered that different "lifestyle" groups differ in terms of these political characteristics as well as in their media behaviors. For example, the heaviest overall television viewers--the "need- driven" segments of the population (also the lowest-income segments) and the ”outer-directed” "belongers" and "emulators" are made up largely of Democrats. Among these, the two "need-driven" groups--the "survivors" and the "sustainers”--as well as the "emulators" tend to be quite distrustful of political and economic leaders, while the ' ”belongers" have the highest level of trust of any of the groups. While a majority of "survivors" consider themselves to be conservative, only 298 of ”sustainers" do so. They are more middle-of-the-road" politically (48%), as are the ”belongers” (46%) and the ”emulators" (44%). Among these four groups, only the "survivors" feel that too much is spent on the military. The two ”need-driven" groups are both heavier than average viewers of game shows and soap operas, though the more conservative "survivors"--the oldest group (median age of 66) and the group with the highest proportion of females (778)--are heavier news and variety show viewers, while the much younger, more middle-of-the-road "sustainers" (median age of 33) watch more comedies, crime dramas and movies. 12 The latter's preference for comedies and movies is shared by the similarly young (median age of 27) and middle-of—the- road "emulators." The more politically trusting ”belongers" (median age of 52 and 68% female) are heavier than average viewers of news shows only and are infrequent viewers of sports programming. Among the lighter television viewers are found the outer-directed ”achievers" and the three inner-directed groups--the I-Am-Me's, the ”experientials" and the ”societally conscious." The ”achievers" (60% male with a median age of 43) are predominantly Republican, conservative, inclined to support military spending and to trust military and corporate leaders (though less so elected officials), and are heavier than average viewers of sports programming only. The ”inner-directed” groups, in contrast, tend to be liberal, independent, less supportive of military spending and less trustful of society's leaders. Among these, the "I-Am-Me's”--the youngest (median age of 21) and most predominantly male (64*) among all groups--are the least strongly opposed to military spending, and the least distrustful of the country's leaders (particularly military leaders). They are also heavier than average viewers of comedies and movies. The most liberal, independent, distrustful of leaders and opposed to military spending are the "societally conscious" individuals. Though light television viewers in general, they are heavier than average 13 viewers of educational programming. They are also the oldest of the inner-directed groups with a median age of 39, and are split about evenly between men and women. Though SRI's study is notable in that it measures program-type viewing as well as certain political characteristics, its VALS program does not appear to directly measure the relationship between these two sets of variables. In addition, it ignores a number of variables important in political campaigns such as likelihood of voting, ticket splitting, opinion leadership, political interest and participation, and candidate preference. [Though there appear to be no audience segmentation studies published by practicing political consultants, there is evidence that some of the leaders in this field employ narrowcasting concepts in their message design and placement strategies. For example, Larry Sabato, in his book, The Rise of Political Consultants (1981, pp. 182-184) cites political media wizard Tony Schwartz as describing "the object of time-buying as 'narrow-casting' rather than broadcasting" and reports that another highly respected consultant, Robert Squier, produced seventy spots for a gubernatorial campaign, "most with a specific audience and time slot in mind." Noting that dissimilar kinds of people watch and listen to different sorts of programs at various times of the day, Sabato cites the following examples: better educated, information-oriented, undecided voters have 2 y 14 been found to cluster around late night news shows: middle- aged housewives can be influenced by family-oriented or charismatic image spots strategically located within afternoon soap operas: and rural Americans, southerners, and blue-collar workers can be reached with down-home, musical or even humorous advertisements during Bee-Haw. While perhaps of great value to political campaigns, these "narrowcasting" insights, with the exception of the first, nevertheless still rely on demographic "locators" of political characteristics rather than on direct measurement of the relationship between political characteristics and program viewing. Age and Gender as Predictors of Political Characteristics In an effort to evaluate the efficacy of the indirect demographics-based approach to narrowcasting, we will now briefly examine the relationship of the two demographic variables included in commercial ratings data-~age and gender--to campaign relevant political characteristics. Political participation. While the difference in levels of registration and voting between men and women once greatly favored men, this differential has steadily declined to the point where it now no longer exists. As of 1982, 63.7 percent of men and 64.4 percent of women reported to the Census Bureau that they were registered to vote, while the numbers who reported actually voting were 48.7 percent for men compared to 48.4 percent for women. While there is 15 somewhat more variation between genders within specific age cohorts, these differences are not in any consistent direction. There is no difference in the 18-24 year old group, women show slightly higher percentages for ages 25- 54, and for the 55 and over age group men are somewhat more likely to vote (Poole and Ziegler, 1985, pp. 121-126). Regarding the relationship between age and voting, analysis of survey and census data indicate that voter turnout is lowest among the youngest voters, highest among the middle-aged, and somewhat lower among the elderly (Abramson, 1986, p. 111; Pierce, 1982, pp. 300-301). In terms of political participation other than voting, the general trend has been that housewives do not participate to the same degree as employed women, but employed women are just about as active as men, most strongly so in presidential campaigns. In 1972 and 1980 women were actually somewhat more active than men, though they appear to be less active in non-presidential election years. The one area where men consistently participate significantly more than women is in their efforts to influence another person's vote (Poole and Ziegler, 1985, pp. 141-147). Partisanship and presidential preference.i In presidential elections since 1968 men and women have both tended to favor conservative over liberal and Republican over Democratic candidates, but women consistently less so 16 than men, with the exception of Democrat Jimmy Carter's victory in 1976, when there was practically no difference between the sexes. In 1980, for example, 60 percent of white males voted for Reagan, while 53 percent of white females did. The numbers for 1984 were 66 percent for men and 61 percent for women (Abramson, et. a1, 1982, pp. 98-99: 1986, pp. 136-137). The relationship of age to candidate preference and partisanship is somewhat unclear in light of the conflicting findings of various studies of recent presidential elections. However, it seems fairly clear that third party candidates did best among the youngest voters in 1968 and 1980, as did President Reagan in 1984. This is generally interpreted as evidence of the relatively weak party loyalty of younger voters (Ibid., p. 100, p. 139), though some .interpret Reagan's popularity among young voters as a movement toward the Republican party. The former interpretation is, however, supported by the relatively high levels of declared "independence" among younger voters (see, for example, Gallup Poll, March 10, 1983), suggesting that a "dealignment" rather than a "realignment" is occurring. Issue attitudes: foreign policy and crime. Gallup Polls taken during the period when the data for this study was collected, as well as other periods, indicate that men tend to take a somewhat harder line than women on foreign policy issues such as the relative value of arms buildup versus negotiations with the Soviet Union. Regarding age 17 differences on these issues, liberal, "dovish" views tend to be strongest among the 18-24 and 25-29 year old age groups, with the percentage of respondents expressing conservative, ”hawkish" views increasing somewhat with age (April 21, 1983). One partial exception to this apparent correlation between age and conservatism on foreign policy issues emerged in response to the question "Do you think the USSR is doing all it can to keep peace in the world?" For this question, the frequency of negative responses peaked for the 30-49 year old group and then decreased with age (Ibid.). In a May 1, 1983 Gallup Poll, when asked if they approve or disapprove of President Reagan's handling of the Central American situation, 30 percent of males and 21 percent of females expressed approval, 49 percent of males and 43 percent of females disapproved, and 21 percent of males and 36 percent of females had no opinion. In terms of age, younger respondents, especially those 25-29, expressed somewhat lower approval ratings than older ones, though the maximum variation in ratings between age cohorts was only 9 percent for approval and 6 percent for disapproval, with no consistent relationship to age appearing throughout all the age groups reported. A number of studies have shown that women tend to be less supportive than men regarding defense spending, military intervention and the death penalty (Holloway, 1986, pp. 45-46). Young people also appear less supportive of the death penalty than older people, with the elderly 18 tending to take the "hardest" line on law and order issues, presumably because of their sense of vulnerability to crime (Ibid. p. 48: Erikson, 1980, p. 171). Political efficacy. Recent studies of political efficacy reveal little or no difference between political efficacy scores of men and employed women, though housewives score significantly lower than both of these two groups (Poole and Ziegler, pp. 137-139). In terms of age, there is no consistent pattern except that the elderly consistently show lower levels of efficacy than other voters. The demographic variable most strongly related to efficacy is education, and it has been suggested that its influence is largely responsible for the observed differences noted above (Abramson, 1983, 182-189). Political Trust. The level of reported political trust has declined steadily in the two decades prior to Ronald Reagan's election for all groups in the population. During the Reagan years this tendency has been somewhat reversed. Women have tended to be less trustful of government than men, with this difference being greater for housewives than for working women (Poole, p. 129-134). There seems to be a fairly clear pattern in the relationship of life-cycle (age) to trust, with young adults having relatively high levels of trust which declines rather rapidly as they enter the mid-20 to mid-30 age group, followed by further but more gradual erosion in later years (Abramson, 1983, pp. 234-38). 19 III. RESEARCH QUESTIONS This study addresses several related research questions confronting political media buyers seeking to practice narrowcasting. Primary among these is whether viewing of particular types of television programs and cable channels is related to campaign-relevant political attitudes and behaviors. Because the relationship of age and gender to program viewing is readily available to media buyers in the form of ratings data, the study also seeks to discover which of these political characteristics is correlated with age and/or gender. By combining these correlations with ratings data (or with the correlations of age and gender to those viewing measures included in this data set), media buyers may be better able to "locate" high concentrations of viewers with particular political characteristics. The study further seeks to identify those political characteristics for which such use of age and gender as proxies for political characteristics is a relatively inefficient targeting strategy. This is done by measuring the partial correlations between content viewing and political characteristics controlling for age and gender. In those instances where such partial correlations are significant we can conclude that political media buyers would improve the quality of their targeting by looking to the results of this or similar studies rather than relying 20 solely on television ratings supplemented by knowledge of the relationship of age and gender to political attitudes and behaviors. The specific research questions addressed in the study are listed below. 1. Are there prime-time television program types which can be identified in terms of audience viewing patterns? 2. Are there cable channel types which can be identified in terms of audience viewing patterns? 3. Is viewing of particular programs, program types or cable channels associated with the following political attitudes and behaviors: b. C. d. f. Likelihood of voting Political activism Opinion leadership Tendency to split one's ticket between parties Partisanship Liberal vs. conservative self-identification Interest in politics Political efficacy Political trust Intention to vote for Reagan in 1984 Attitude toward law and order issues 21 l. Attitude toward U.S. military intervention m. Attitude toward the USSR and socialism? What campaign relevant political characteristics and content viewing measures are significantly related to age and gender? Which of the significant zero order correlations discovered between viewing measures and political attitude and behavior variables remain significant when subjected to age and gender controls? \ 22 IV. METHODOLOGY Sampling and Data Collection The data used in this study were collected in a telephone survey of Lansing area residents over a period of six days in late October and early November of 1983. A sample of 321 individuals was drawn from the Lansing telephone book using systematic sampling procedures. Respondents were screened so that only those 18 years or older were interviewed, since the target population of the study was those of voting age. The telephone survey was administered by trained and supervised college students. Each phone number included in the sample was called at least three times. The non- completion rate, including households that could not be contacted after three attempts (15%), refusals (14%) and households where an 18 yr. old was not available (7%), totalled 36% percent. A copy of the questionnaire used in the study is included in the appendix. The data was recoded on the questionnaire when necessary, and then transferred to codesheets. It was then keypunched into an MS-DOS based computer equipped with SPSS- PC software. Using that software, a system file was created for analysis. 23 Measurement and Analysis Program-viewing measures were developed using questions as to the frequency of viewing 29 prime-time television shows and five prime-time fringe shows. Prime-time shows were defined as those running between 8 and 11 p.m., while prime-time fringe shows were defined as non-news shows running between the hours of 5 and 8 p.m. 'The questions used three response categories, "Regularly", "Occasionally" and "Practically Never". As Table 1 shows, measures of ”program-type” viewing were developed by subjecting the prime-time show items to principle components factor analysis with varimax rotation. This yielded seven factors which together explained 58.4 percent of the total variance. The first factor, referred to as "Less Sophisticated Comedy/Entertainment" included comedies such as Happy Days, Dukes of Hazzard, Alice, The Jeffersons, as well as the two light-entertainment magazine shows, That's Incredible and Real People. This factor explained 22.7 percent of the variance. The second factor included the somewhat more sophisticated sitcoms such as Family Ties and Newhart and explained an additional 8 percent of the total variance. The third factor included the evening soap-operas such as Dallas and Dynasty and explained an additional 7.2 percent of variance. The next two factors included most of the (action/adventure/police shows. One consisted of three ABC shows--Fall Guy, Matt Houston and T.J. Hooker--all with 24 Table 1: Prime Time Viewing Factor Loadings Less lore Evening 'lacho' Cop 'Charninq' Cop lel use less/ Sophisticated Sophisticated Soaps Action/Adventure lotion/Adventure Series Info. Conedy/ Coaedy lagazine lntertainnent llice .67 Happy Days .64 That's Incredible .59 .54 Dukes of Hazzard .53 Real People .51 .61 The Jeffersons .50 Three’s Colpany .49 .44 One Day at a line .44 .44 Fanily Ties .77 facts of Life .67 leuhart .61 line to Five .47 Falcon Crest .82 Dallas .80 Dynasty .68 Knots Landing .62 Fall Guy .69 latt Houston .67 TJ Hooker .67 lapnua 0.1. .70 Silon and Sison .67 Bart to Hart .48 Hill Street Blues .77 St. Elselhere .71 leninqton Steele .50 Cheers .50 20/20 .73 60 linutes .63 Total variance accounted for: 22.78 8 7.2 6.8 5.6 4.4 3.8=58.48 £35 relatively "macho" heroes, and explained 6.8 percent of the variance. The second included three shows with relatively "charming" heroes--Magnum P.I., Simon and Simon, and Hart to Hart. This factor explained an additional 5.6 percent of variance. The sixth factor was cemprised of four NBC shows which did not fit into a single program genre as easily as the others included in the study--Hill Street Blues, St. Elsewhere, Remington Steele and Cheers. It explained an additional 4.4 percent of variance. The final factor included all of the magazine-format shows and explained 3.8 percent of variance. Prime-time fringe shows were not included in this factor analysis because they were televised at a time when many people are either still at work, driving home, or eating dinner. Since the effects of viewers' daily work schedules would seem much more likely to distort preference- based viewing patterns for these shows in comparison with prime-time shows, it was decided to omit them from the factor analysis of the latter. In addition, because there were so few fringe shows included in the study, and because they were repeated daily and thus tended to compete directly 'with each other for viewers, each was included in the subsequent analysis as a separate variable. Respondents were also asked whether they were cable subscribers and whether they owned a VCR and a home computer. If they were cable subscribers they were asked to 26 estimate the weekly amount of time they viewed each of the basic cable channels. Submitting these data to a principle components factor analysis yielded no factors with 30 iterations. An examination of the correlation matrix for these variables also revealed a lack of intercorrelation among them. Therefore, each cable channel viewing measure was used separately in the subsequent analysis. Respondents' estimates of the total time they spent each day watching television and the number of times each week they watched television news shows were also solicited. Several aspects of voting behavior were measured. Respondents were asked if they had voted in the 1980 election, whether they intended to vote for President Reagan in 1984, and the extent to which they were inclined to split their ticket between parties. The latter was measured by offering the respondents three choices: ”always vote for candidates of the same party," ”usually vote for most of candidates of the same party," or "vote without much consideration of which party a candidate is affiliated with." Political activity in addition to voting was also measured, with the responses subjected to a principle components factor analysis with varimax rotation (see Table 2). Opinion leadership was measured by two questions regarding whether the respondent had tried to convince anyone of their political ideas lately and whether they had 27 Table 2: Political Activity Factor Loadings Attended political rally Worked on campaign Contributed to campaign Belong to political organization Contacted political leader or public official Tried to convince someone of political ideas Was asked for political advice Written letter to editor Total variance accounted for: 28 Political Activity .75 .75 .71 .65 .55 14* Opinion Leadership .76 .67 .66 38% = 52% been asked for their advice on a political question recently. Other political activities such as "worked on a political campaign," "attended a political rally," and "contributed to a political campaign" were measured using three point response scales with the following values: "more than once", "only once” and ”never." In addition, respondents were asked if they belonged to a political club or organization. The factor analysis of political activity measures yielded two factors explaining 52 percent of the total variance: one representing opinion leadership and explaining 38 percent of variance, and the other associated with actual physical or financial participation in political campaign activity which explained an additional 14.1 percent of variance. Partisanship was measured using a seven-point scale from "strong Republican" to "strong Democrat." The scale was used for correllational analysis and was also collapsed into three categories--Republican, Independent and Democrat- -for use in an analysis of variance. In addition, respondents were asked to identify themselves on a five point scale from "very conservative" to "very liberal." These two variables were also combined to create a typology consisting of nine combinations of three partisanship categories--Republican, Independent and Democrat--and three ideological categories--conservative, .middle-of—the-road and liberal. This typology was used in 29 subsequent analyses to explore the possibility that particular subgroups--e.g., conservative Democrats, liberal Independents, liberal Republicans--had different viewing patterns. This was done largely in response to some of the findings of SRI's VALS study, for example, that the conservative Democratic ”survivors" had substantially different viewing patterns than the more liberal Democratic ”sustainers.” Creation of this composite variable was also felt to reflect the fact that each party's membership as well as the growing ranks of political independents span the full range from liberal to conservative. This being the case, it seems desirable, particularly on certain issues, for politicians to be able to address these different subgroups separately. Respondents were also asked whether they were "very interested", "somewhat interested" or "not very interested" in politics and political campaigns. - Respondents' trust in political and economic leaders was measured using several questions. One asked whether the respondents felt that "the government in Washington was run by a few big interests looking out for themselves or for the benefit of all the people." Two other questions asked how often respondents felt that the government in Washington and the leaders of private industry did what was best for the country, using a four point scale ranging from "always" to "practically never." Political efficacy was measured using 30 two Likert-scale questions ranging from "strongly agree" to "strongly disagree." One concerned perceived efficacy with regard to action, while the second concerned one's perceived ability to understand politics and government. Both the trust and efficacy related questions were subjected to a principle components factor analysis with varimax rotation which yielded one trust factor explaining 34.4 percent of total variance and one efficacy factor explaining an additional 26.7 percent of variance, a total of 61.1 percent of variance explained by the two factors (see Table 3). A number of items measuring attitudes regarding foreign policy issues were also subjected to factor analysis. These included three Likert items with four response categories ranging from "strongly agree” to ”strongly disagree.” One- of these asked respondents whether they felt that the U.S. had ever fought in an unjust war, another attempted to measure respondents' acceptance of socialism, while a third addressed their willingness to support negotiations in El Salvador even if it meant that Communists would get some power in that country's government. An additional two items solicited respondents' attitudes toward the Reagan administration's policies in Grenada and Lebanon using four response categories ranging from "strongly approve" to "strongly disapprove." Two final items, using a forced- choice format, addressed respondents' relative evaluation of reliance on arms buildup vs. arms negotiations and the 31 Table 3: Political Trust and Efficacy Factor Loadings Political Political Trust Efficacy Government run by few interests looking out for themselves or for .72 the benefit of all the people? How often do you think that the leaders of private industry do .77 what's best for the country? How often can you trust the government in Washington to do what's right for .74 the country? It's no use worrying about public affairs; I can't do anything about .82 them anyhow. Sometimes politics and government seem so complicated that a person like me .81 can't really understand what's going on. Total variance accounted for: 34.4% 26.7% = 61.1% 32 extent to which the U.S. should attempt to get along with the Soviet Union. A factor analysis of these items yielded -two factors explaining 49.7 percent of variance. The first factor included items which addressed attitudes toward the use of U.S. military troops abroad--the question about U.S. involvement in an unjust war and the two questions about support for Reagan administration policies in Lebanon and Grenada--and explained 32.8 percent of variance. The second factor, which explained an additional 16.9 percent of variance, included the other four items which addressed attitudes toward relations with the Soviet Union and toward socialism and communism (see Table 4). Three items included in the study addressed attitudes toward law and order and appropriate forms of punishment for criminals. Two of these items used Likert scales measuring levels of agreement or disagreement with a statement about the relative merits of rehabilitation versus punishment and about whether the police should hesitate to use force to maintain order. A third question asked respondents whether they supported or opposed capital punishment as a deterrent against crime. When subjected to a principle components factor analysis with varimax rotation all three items loaded on to a single factor which explained 49.3 percent of variance (see Table 5). 33 Table 4: Foreign Policy Issue Factor Loadings U.S. Military View of USSR Intervention and Communism/ Socialism Do you strongly approve, approve, disapprove or strongly disapprove .87 of President Reagan's handling of the problems in Lebanon? Do you strongly approve, approve, disapprove or strongly disapprove .81 of President Reagan's handling of the situation in Grenada? The U.S. has never fought .57 in an unjust war (Likert scale) To end the fighting in El Salvador, the U.S. should arrange for negotiations .82 with the rebels even if it means that the Communists would get some power in the government. (Likert scale) (The following two questions asked respondents to choose which of two opposing viewpoints they most agreed with.) It is important for the U.S. to try very hard to get along with Russia ' versus .66 It is a big mistake to try to get along with Russia. The U.S. should place less emphasis on military weapons and more emphasis on negotiations. versus .49 It is important that the 0.8. continue to develop new and more effective weapons and not depend too much on negotiations. It would be alright if socialism replaced capitalism if it meant a better life for .47 most people. (Likert scale) Total variance accounted for: 32.8% 16.996,= 49.7% 34 Table 5: Law and Order Attitude Factor Loadings Would you say that you are in favor of capital punishment as a deterrent to crime, or would you say that you are opposed to it? The police should not hesitate to use force to maintain order. (Likert scale) Our treatment of criminals is too harsh: we should try to rehabilitate, not to punish them. (Likert scale) Total variance accounted for: 35 Law and Order/ Punishment Factor .75 .59 49.3% The factor scores generated by the factor analyses of these variables were appended to the data file for use in further analyses. Two demographic characteristics were also included in analysis--age and gender. As noted earlier, these variables were selected because they are the only demographic variables currently available in the commercial ratings data which media buyers usually rely on for targeting purposes. The data were analyzed in three stages. First bivariate analyses were performed to determine the zero- order relationships between the independent variables measuring television content viewing and the dependent variables measuring political attitudes and behaviors. Second, correlations of both dependent and independent variables to age and gender control variables were examined. This provided information as to how the two demographic variables could be used by media buyers to "locate" viewers with particular political characteristics for targeting purposes. In the third stage, partial correlations between independent and dependent variables with controls for age and gender were calculated in order to explore the extent of direct relationships between viewing patterns and political attitudes and behaviors. This analysis was intended to reveal instances where age and gender would not serve as efficient proxies in locating audience segments with particular political attitudes or behaviors. 36 V . RESULTS Of 364 bivariate relationships between viewing variables and political attitudes and behaviors, 51 emerged as significant at a level of p-.05. When these zero-order relationships were subjected to demographic controls, 28 were still significant at p-.05 (see Table 6). Bivariate Analysis Political participation, interest, efficacy and trust. Several television viewing variables were negatively related to voting. Heavier viewers of both kinds of comedy shows were less likely to be voters than lighter viewers (r=-.18 for the "less sophisticated comedy/entertainment” factor: and r--.21 for the "more sophisticated comedy” factor). The same was true with regard to several cable channels: ESPN (r--.l7), MTV (rs-.42), and the USA Network (rs-.15). Only two viewing variables were positively related to voting. More frequent television news viewers and more frequent viewers of People's Court were more likely than less frequent viewers to vote (r-.13 and r=.ll, respectively). Cable subscribership was also found to be mildly related to voting behavior (r-.10), though not to any of the other political characteristics included in the study. Heavy television viewers were less likely to be opinion leaders than lighter viewers (r=-.23) and expressed less 37 interest in politics and public affairs (r=-.17) and less sense of political efficacy (rs-.23). A number of specific programs, program-types and cable channels were also related to opinion leadership and political activity, interest and efficacy, though none emerged as significantly related to political trust. Those who watched more television news tended to be more politically active (r-.15) and more politically interested (r-.20) than lighter news viewers. Heavy viewers of People's Court were less politically active (r=-.ll), less often opinion leaders (r--.18), and expressed less sense of their own political effectiveness (rs-.19) than light or non-viewers. The same was true for Wheel of Fortune viewers with regard to political activity (rs-.13) and political ‘efficacy (rs-.17), and for regular Monday Night Football viewers in terms of opinion leadership (r=-.13). Among the five prime-time fringe shows, only the viewing of MASH was positively related to a measure of political activity, showing a mild positive association with opinion leadership (r=.10). In terms of prime time viewing, "more sophisticated comedy” viewing was negatively associated with political activity (rs-.15), political efficacy (rs-.20), and interest in politics (r=-.14). Action/adventure series viewing was also negatively associated with political activity (r=-.15 for "charming cop" shows) and political efficacy (r=-.14 for 38 "macho cop" shows). Viewing of the shows comprising the ”NBC" factor, on the other hand, was positively related to feelings of political efficacy (r=.l3) and to political interest (r8.15). Regular C-SPAN viewers were also more politically involved than others. They were more active (r=.15) and were more likely to be opinion leaders (r=.l7). With regard to the demographic variables, women emerged as significantly less interested in politics than men (r=- .16), less trustful of government and economic leaders (r=- .14), less politically active (r--.12), less likely to be opinion leaders (r=-.13), but not significantly less likely to vote. Age showed a fairly strong positive relationship to both voting (r=.26) and other political activity (r=.24) (see Table 7). Partisanship. In the correlational analysis, heavier television viewers emerged as more likely to consider themselves Democrats than lighter viewers (r=.17), as did heavy viewers of "less sophisticated comedy/entertainment" shows (r=.21) and Monday Night Football (r=.11). When the partisanship variable was collapsed into three categories--Republican, Independent and Democrat--and its relationship to viewing examined using analysis of variance, the F-statistic was significant at the .05 level for viewing of three types of content--less sophisticated comedy/ entertainment shows, NBC-factor shows, and Monday Night 39 Football. Using the least-significant-difference paired comparison test, the following was discovered: Democrats were the most likely to watch the less sophisticated I comedies, followed by Independents, with both groups viewing these shows significantly more often than Republicans: Independents were the most likely to watch the NBC factor shows, but a significant difference was found between them and Republicans only, who were the least likely to watch these shows: Monday Night Football was watched most by Democrats, significantly more than Republicans, but not significantly more than Independents. Regular viewers of "charming cop" action-adventure shows revealed somewhat more party loyalty in their voting habits (r=.12), while fans of the new breed of NBC series exemplified by Hill Street Blues were more likely than others to split their ticket among different parties' candidates (r=-.13). Of those who reported an intention to vote in the 1984 presidential election, heavier television viewers were less likely than lighter viewers to report that they intended to vote for president Reagan (r=-.12). Among specific content variables, only USA Network viewing was associated with a relative (positive) preference for Ronald Reagan (r=.20). Neither age nor gender was significantly related to partisanship, though younger respondents appeared significantly less likely to vote for Ronald Reagan than 40 older ones (r=-.l4). They were also more likely to split their ticket (r=-.10), thus displaying less party loyalty than older generations. Liberal/conservative self-identification. Relatively heavy viewers of "NBC-factor" shows (r=.17), MASH (r=.11), C-SPAN (r-.15), and the Arts and Entertainment cable channel (r=.21) considered themselves more liberal than did less frequent viewers. The reverse was true with regard to frequency of viewing magazine format shows, where heavier viewing was associated with a more conservative identity (r=.11). . As expected, age showed a significant positive relationship to conservatism (r=-.11), though there seemed to be no relationship between gender and this variable (r=.0004) Partisanship/ideology. When this composite variable consisting of nine combinations of partisanship and liberal/conservative self-identification was tested for relationships with viewing variables in an analysis of variance, the F-statistic was significant for three relationships (See Table 9). Liberal Republicans were found to be the most likely to watch MASH, followed by liberal Independents and middle-of-the-road Independents. The least likely to watch this show were conservative and middle-of- the-road Republicans, and significant differences in viewing frequency were found for both these two Republican groups in comparisons with the three most frequent viewer groups. 41 Conservative Democrats emerged as the most frequent viewers of Wheel of Fortune, followed by middle-of-the-road “Independents. Members of these two groups were significantly more likely to watch this show than members of a wide range of other groups, including middle-of-the road Democrats (the least frequent viewers), liberal Independents, and conservative and middle-of-the-road Republicans. Conservative Democrats were also the most frequent viewers of the lessesophisticated comedy shows, with all paired comparisons with other groups significant at the .05 level. Law and order/crime and punishment. On the issue of law and order, heavy viewers of "macho cop" shows appeared to take a harder line against criminals than lighter viewers (r=.19), as did regular viewers of Wheel of Fortune (r=.12). While there was not a significant difference between men and women for this factor, older respondents did take a significantly harder line than did younger ones (r=.l6). Foreign policy. Macho cop show viewers tended to favor the use of military force abroad (r=.15) and to take a harder line against socialism and the Soviet Union (r=.17). Regular viewers of "more sophisticated" comedies, while also expressing relatively strong distrust of the Soviets and socialism in general (r=.17), were less likely than less frequent viewers to support the use of U.S. troops in military actions abroad (rs-.14). Those who regularly watched the shows comprising the "NBC-factor" seemed to have 42 relatively high levels of faith in negotiations and relatively little distrust of the Soviet Union and socialism -in general (r=-.32). Among cable viewers, relatively heavy viewers of the Arts and Entertainment channel were more likely than lighter viewers to oppose U.S. military intervention abroad (r--.21). Men took harder lines on both foreign policy issues than did women (rs-.23 regarding military intervention, and r=-.19 on attitudes toward the Soviet Union and socialism). Age, in contrast, cut in different directions with regard to the two foreign policy factors. While younger respondents were significantly less supportive of military intervention (r=.14), they were also significantly less trustful of the Soviet Union and accepting of socialism (rs-.17). Age, gender, and program and channel viewing. Older respondents tended to be heavier viewers of television news (r=.43), Wheel of Fortune (r=.23), People's Court (r=.l3), Hour Magazine (r=.25), and of television in general (r=.20) than younger respondents. They also tended to be heavier viewers of the magazine format programs (r=.25) and of CNN (r-.18). On the other hand, they appeared to watch MASH (rs-.20), "NBC factor" shows (rs-.20), and MTV (r=-.34) less often than did younger respondents (see Table 8). Women tended to watch somewhat more television than men (r=.10), particularly the "more sophisticated comedies" (r=.23), the evening soap operas (r=.24), and the Cable Health Network (CHN) (r=.13). On the other hand, they were 43 less likely than men to be heavy viewers of sports programming such as ESPN (rs-.36), Monday Night Football (r=-.40) and the USA Network (r=-.2l). None of the other correlations between age or gender and viewing variables emerged as significant at p=.05. Multivariate Analysis To find out whether the observed bivariate relationships between viewing patterns and political attitudes and behaviors had predictive power beyond that which was available through reliance on commercial ratings data, age and gender were included as controls in the second stage of the analysis (see Table 6). Political activity, interest, efficacy and trust. In terms of voting, the tendency of heavy comedy viewers to be less likely to vote than lighter viewers did not lessen when demographic controls were introduced. In fact, it increased, (r=-.22) for the "less sophisticated" comedy/ entertainment shows, while remaining the same as the bivariate correlation (rs-.18) for the "more sophisticated" comedy shows. A stronger negative correlation to voting when subjected to controls was also observed for ESPN (r=- .25) and the USA Network (r=-.19). The negative relationship of MTV viewing to voting also retained its strength (rs-.36). However, the positive relationships between voting and television news and People's Court viewing largely disappeared with demographic controls (r=.03 and .08, respectively). Age emerged as an extraneous 44 variable in both cases, with older viewers more likely both to vote and to watch both television news shows and People's Court. Among the observed bivariate relationships of television viewing variables to measures of political interest, activity, efficacy and trust, a good portion were still significant when subjected to age and gender controls. Partial correlations were significant at p-.05 for the following relationships: overall television viewing was negatively related to political interest (r=-.17) and efficacy (rs-.20): television news viewing (r=.19) and "NBC factor” viewing (ré.16) were positively related to political interest; C-SPAN viewing was positively related to political activity (r=.18) and opinion leadership (r=.18): viewing of Peoples' Court (rs-.16) and Monday Night Football (rs-.20) were negatively related to opinion leadership: Wheel of Fortune (r=-.18), People's Court (r=-.l4) and ”charming cop show" (r--.-.l6) viewing were negatively related to political activism. Viewing of Wheel of Fortune (r--.14) and People's Court (rs-.18), along with "more sophisticated comedies" (r--.18) and "macho-cop shows" (r=- .15) also retained a significant negative relationship to political efficacy after controlling for the effects of age and gender. Partisanship. Neither the relative party loyalty in voting revealed among regular viewers of "charming cop" shows nor the tendency to split one's ticket characterizing 45 regular viewers of the "NBC factor" shows retained their significance when controlling for age and gender. The relationship of total television viewing (r=.18) and "less sophisticated comedy/entertainment shows" (r=.21) but not of Monday Night Football viewing, to Democratic partisanship still held with controls (r=.16). The relationship of total viewing to a disinclination to vote for Reagan appeared even stronger with controls (r=-.15 compared to zero-order r=-.12). The same was true for heavy viewing of the USA network (r=.21 compared to r=.20 for the zero-order relationship) though in the opposite direction-— heavier viewers being more likely to vote for Reagan. Liberal/conservative self-identification. The association of liberal self-identification with ”NBC-factor" viewing (r-.15) and with Arts and Entertainment channel viewing (r=.21) retained their significance with controls. However, the demographically controlled relationship of C- SPAN and MASH viewing to liberalism was not significant at p=.05. The relationship of conservatism to information/magazine show viewing also was not significant in the analysis of partial correlations, with the viewers of this type of show tending to be both older and more conservative than non-viewers. Law and order/crime and punishment. 0f the two significant bivariate associations of viewing variables to a hard line position on the "law and order" factor, only one, the relationship to "macho cop" show viewing held up with 46 demographic controls (r=.20). Age appeared to be an extraneous variable in the relationship between this factor and Wheel of Fortune viewing, with older viewers appearing both more hard-line on law and order issues and more likely to watch this show. Foreign policy. Attitudes toward foreign policy issues were influenced by gender as an extraneous variable. The relationship of "more sophisticated comedy" viewing to both foreign policy attitude factors, as well as that between "macho cop" show viewing and the "military intervention" factor were no longer significant when subjected to controls. Apparently women were more likely both to watch these comedy shows and to take less hard-line positions, while men were more likely to watch "macho cop" shows and to have a more militaristic view of foreign affairs. However, the positive relationship of “macho cop" show viewing to a hard-line position on the second foreign policy factor-- "attitude toward the USSR and socialism" did not lose any of its strength with demographic controls (r=.18). "NBC- factor" viewing also retained a fairly strong relationship with this attitudinal factor, though in the opposite direction (r=-.29). 47 Table 6: Political TV Tine TV less VieIing Cahle Subscriber VCR Pole Conputer People's Court lash Iheel of Fortune Television Viewing and Political Characteristics: Correlations and partial correlations controlling for age and gender Opinion lctivity lasts; .051 (299) .149' (271) .059 -.078 (279) -.062 (209) .035 ‘(209) -.113‘ (200) -.147° .065 (280) -.l27c (279) -.183' Political Leadership Pactor .101 (299) ’ .084 (271) ‘ -.069 (271) -.082 (289) .035 (289) -.175' (200) -.1645 .1045 (200) .077 -.033 (279) a 8 significant at p=.001 b a significant at p=.01 c a significant at p=.05 Partial correlations are included for variable pairs with zero-order correlations significant at p<=.05 Political Efficacy Pactor -.225' (239) -.203' -.064 (215) -.022 (225) .000 (230) -.061 (230) -.109’ (223) -.179' -.067 (223) -.168’ (225) -.144‘ ‘48 Bconoaic Political Trust Factor Interest .022 -.167' (239) (309) -.174' -.043 .195' (215) (281) .185) .077 .073 (225) (298) .010 -.050 (230) (299) -.055 -.055 (230) (299) -.083 -.087 (223) (290) .010 -.004 (223) (290) -.057 -.002 (221) (289) has and Order Factor .009 (255) .030 (229) -.001 (239) .027 (246) -.039 (246) .034 (238) .090 (230) -.119‘ (237) -.007 87 Four lagazine Monday light Football Coledy I Factor (Less-Soph.) Couedy 2 Factor (lore Soph.) Evening Soap Factor Macho Cop Show Factor Charning Cop ShOl Factor Hill Street/ 88C Factor lagazine Show Factor Table 6 (cont'd.) Political lctivity Factor -.091 (281) -.083 (280) -.080 (277) -.098 (243) -.145' (243) -.106 -.018 (243) -.007 (243) -.1456 (243) -.152‘ .080 (243) -.028 (243) Opinion Leadership Factor -.007 (280) -.065 (280) -.128‘ (277) -.203| -.028 (243) -.091 (243) -.024 (243) -.083 (243) -.092 (243) .082 (243) -.009 (243) Political Efficacy Factor .109 (223) -.041 (223) -.078 (223) -.061 (198) -.195- (193) -.m= -.058 (198) -.144c (198) -.l2l -.104 (198) .133c (198) .112 -.041 (198) 449 Political Bcononic IBEiJEQEI -.009 (223) .011 (223) .008 (223) ~.072 (198) -.076 (198) .106 (198) .010 (198) .048 (198) -.030 (198) .021 (198) Political Interest .016 (291) .109 (290) .039 (290) -.075 (252) -.139' (252) .060 -.073 (250) -.043 (250) .026 (252) .145” (252) .158c .033 (252) La! and Order Factor -.001 (238) -.085 (238) -.038 (234) -.065 (203) .032 (203) -.089 (203) -.194b (203) -.197' .002 (203) .060 (203) .046 (203) Cll ESPN MTV DSI letlork Cable Health letsork C-Span 8876 Satellite Prograu letlork Table 6 (cont'd.). Political Activity Factor .038 (158) .015 (153) .049 (149) .023 (146) .095 (147) .1495 (145) .1766 .012 (144) .004 (147) Opinion_ Political Leadership Efficacy Factor Factor -.012 .012 (158) (136) .067 -.044 (153) (132) .080 .028 (149) '(129) .041 -.102 (146) (128) -.080 -.018 (147) (128) .1656 -.011 (145) (127) , .176“ .112 -.023 (144) (125) -.117 -.086 (147) (128) 5() Political Fconolic Trust Factor .127 (136) .033 (132) .117 (129) -.040 (128) -.006 (128) .116 (127) -.035 (125) .102 (128) Political Interest .011 (166) .007 (160) -.051 (156) -.004 (154) -.114 (152) .009 (153) .021 (152) .042 (154) Law and Order Factor -.082 (141) .069 (139) .130 (134) .124 (134) -.033 (135) .021 (133) -.025 (131) -.062 (135) TV Tine TV less Vieving Cable Subscriber VCR Pole Couputer Peoples Court lash Vheel of Fortune KT Hour Magazine Table 6 (cont'd.). Voter -.059 (284) .132“ (259) .025 .100“ (266) .099 -.0484 (274) -.096 (274) .114“ (266) .080 -.032 (266) -.127“ (266) -.125“ .051 (267) -.047 (267) honday flight -.030 Football (267) Partisan- shin. .174) (287) .184“ .099 (260) .024 (269) .003 (277) .033 (277) .044 (268) .070 (268) .100 (267) .026 (270) .047 (268) .110 (266) Reagan Voter -.120' (297) -.149' -.014 (271) -.088 (278) -.032 (288) .060 (288) .056 (279) -.030 (279) .042 (278) .060 (280) .004 (279) -.004 (278) 51. Liberal vs. Ticket Conservative Splitter -.024 -.077 (304) (299) -.046 .012 (276) (273) -.021 -.035 (285) (281) -.031 -.056 (294) (291) -.022 -.001 (294) (291) .034 .011 (285) (282) .107c .056 (285) (282) .080 -.058 .035 (284) (281) -.029 .068 (286) (283) -.120 -.034 (285) (282) .085 .041 (282) (279) Military Action {2219; .036 (164) .112 (152) -.024 (154) -.094 (159) ~.018 (159) -.080 (154) -.035 (154) -.032 (153) .004 (154) -.054 (154) .098 (234) View of USSR 6 Socialisa .104 (164) .007 (152) -.027 (159) -.033 (159) -.036 (159) .059 (154) -.039 (154) .059 (153) -.090 (154) .122 (154) -.070 (234) Table 6 (cont'd.). Voter Coaedy I -.181' Factor (234) -.218| Conedy II -.214“ Factor (234) -.183“ Evening -.050 Soap Factor (234) Eacho Cop -.035 Shov Factor (234) Charning Cop .001 Shov Factor (234) Hill Street/ .067 EEC Factor (234) Eagazine Shov .047 Factor (234) Partisan- 14.11 .209I (234) .207“ .072 (234) .029 (234) .052 (234) .101 (234) .056 (234) -.064 (234) Reagan Voter .104 (229) .031 (229) .016 (229) -.079 (229) -.017 (229) .074 (229) -.057 (229) Liberal Vs. Ticket Conservatisl Splitter .049 .054 (248) (245) .099 .034 (248) (245) .048 -.014 (248) ' (245) .032 .046 (248) (245) .007 -.117“ (248) (245) -.115 .165“ .131“ (248) (245) .147“ .111 -.113“ .062 (248) (245) -.088 552 Military Action Factor -.046 (138) .142“ (138) .082 :.006 (138) -.148“ (138) -0144 -.053 (138) -.037 (138) -.127 (138) Viev of USSR 6 Socialisn -.094 (138) .172“ (138) .151 .053 (138) .165“ (138) .178“ .091 (138) -.319I (138) -.290‘ .057 (138) Cl] 8888 MTV 0S1 letvork Cable Eealtb Eetvork C-Span lrts Satellite Progran letvork Voter .106 (153) -.172“ (147) -.246' -.418| (145) -.355' -.147“ (142) -.159“ -.040 (142) -.044 (141) -.033 (141) .107 (144) Table 6 (cont'd.). Partisanship -.044 (152) .080 (147) .184) (150) .150 -.203' (148) -.218“ -.089 (144) -.073 (142) .019 (141) .003 (144) Reagan Voter .001 (159) .085 (153) -.183’ (150) -.144 .203“ (148) .208' -.045 (148) .027 (147) .105 (146) -.012 (149) 513 Liberal Vs. Ticket Conservative Splitter -.030 .030 (160) (159) .002 -.062 (155) (152) -.027 .120 (152) (149) .082 -.010 (150) (147) .051 -.030 (151) (148) .146“ -.100 (149) (146) .138 .206“ -.087 (147) (145) .206“ .073 -.126 (151) (148) Eilitary Action 138121. -.060 (86) .072 (83) .116 (81) .088 (79) .009 (79) -.018 (79) .210“ (76) .169“ -.052 (79) View of USSR 6 Socialisa -.022 (86) .119 (83) -.138 (81) .001 (79) -.124 (79) -.049 (79) -.165 (76) -.106 (79) Table 7: Correlations of Age and Sex to Political Characteristics Political Voter Partisan- Liberal/ Reagan Ticket Political Opinion Lav and Efficacy Interest ship Conservative Voter Splitter Activity Leader Order AGE .0799 .2642 -.0013 -.1129 -.1378 -.1024 .2370 -.0889 -.1559 -.0241 ( 303) ( 279) 284) ( 299) ( 293) ( 295) ( 296) ( 296) 251) ( 233) P= .083 P8 .000 P3 .491 P: .026 P: .009 P: .040 P= .000 P: .063 P= .007 P= .357 SEX -.1641 -.0991 -.0573 .0004 .0336 -.0667 -.ll73 -.1317 -.0101 -.0923 ( 294) ( 268) 273) ( 289) ( 282) ( 284) ( 287) ( 287) 245) ( 226) P: .002 P: .053 P8 .173 Ft .497 P: .287 P: .131 P: .024 P= .013 P: .437 F3 .083 Trust Eilitary View of Action USSR AGE -.0886 -.1360 .1743 ( 233) ( 161) ( 161) = .089 P: .043 P: .013 SEX -.1410 .2389 .1894 ( 226) ( 155) 155) P8 .017 P8 .001 P: .009 Table 8: Correlations of Age and Sex to Television Viewing TV Tile TV News (heel of People’s How HASH Entertaiment Comedy Comedy Evening C-Soan Fortme Court Haoazine Tonight 1 2 Soaos AGE .1976 .6250 . 2250 .1269 . 2516 -. 2008 -. 0696 . 0823 -. 0716 -. 0272 -. 0866 ( 305) ( 278) ( 285) ( 286) ( 286) ( 286) ( 287) ( 250) ( 250) ( 250) (150) P: .000 P: .000 P: .000 P: .016 P2 .060 P: .000 P: .202 P: .097 P: .130 P: .336 P=.152 SEX . 0989 .0359 .0821 . 0279 .0630 -. 0761 -. 0091 - . 0672 .2306 .2357 -. 0060 ( 296) ( 266) ( 27s) ( 276) ( 276) ( 276) I 277) ( 261) ( 211) ( 211) (113) P: .065 P: .280 P: .087 P: .322 P: .168 P: .110 P: .660 P: .233 = .000 P: .000 P=.152 Nacho Chenino NBC Havazine Honda)! COD Shows COD Shows Shows Shows Night CNN ESPN 717V USANET CHN ARTS Football AGE -. 0005 .0262 -. 1991 .2698 -. 0628 .1806 . 0303 -. 3366 -. 0778 .0573 -. 0369 ( 250) ( 250) ( 250) ( 250) ( 283) ( I61) ( 157) ( 156) ( 151) ( 152) (169) P: .697 P: .352 P: .001 P: .0113 P: .236 P: .011 P: .353 = .000 P: .171 P: .261 P=.327 SEX -.0601 .0069 —.0585 -.0119 -.6026 -.1037 -.3616 .0877 -.2111 .1368 .1369 ( 211) ( 261) ( 261) ( 261) ( 273) ( 160) ( 156) ( 152) i 119) 1 150) (167) P: .268 P: .670 P: .183 P= .627 P= .000 P: .096 P: .000 P: .161 P: .005 P: .050 P=.012 56 Table 9: Partisanship, Partisanship-Ideology and Television Viewing (F values for Analysis of Variance) Partisanship Partisanship-Ideology TV Time 3.00 1.4 TV News 2.8 1.6 Cable Subscriber 1.3 . 1.6 VCR 1.0 1.0 Home Computer 1.1 , 1.5 People's Court 1.8 1.1 HASH 2.4 3.1b Wheel of Fortune 1.6 . 2.6b ET 0.1 1.1 Four Magazine 0.1 1.8 Monday Night Football 3.1° . 1.2 Comedy 1 4.6b 1.6 Comedy 2 1.2 1.9 Evening Soaps 0.4 0.9 Macho Cop 0.1 2 1.3 Charming Cop 0.6 0.9 NBC Factor 2.7 1.9 Magazine Show 0.6 0.9 CNN 2.5 1.4 ESPN 0.5 0.5 MTV 1.3 0.7 USA Network 1.8 1.7 CHN 1.3 0.4 C-SPAN 0.6 1.6 ARTS 0.3 1.4 55 VI. DISCUSSION This chapter discusses the results, implications and limitations of this study. It first examines the findings regarding program and cable-channel types. Following this it considers the implications for potential political narrowcasters of the study's findings regarding the relationship of viewing patterns to political characteristics. Readers seeking to draw conclusions frOm the results of this study should keep in mind its limitations. First of all, the sample is limited to residents of the Lansing, Michigan, area. Secondly, the data are four years old at the time this report is completed, having been collected in the fall of 1983, approximately one year before the 1984 presidential election and several months before the presidential primaries began to receive heavy press coverage. Data collected--even from the same respondents-- might vary considerably if the survey was conducted in the heat of a presidential campaign. Additional research would thus be required to test the generalizability of its findings to other locales and time periods. Program and Cable-Chgnne; Types In general, the factor loadings of the various programs are conducive to meaningful interpretation of "program- types," though they can also be viewed as revealing some 56 network loyalty and scheduling effects. For example, two of the factors included shows of only one network--ABC's "macho-cop” shows (all scheduled on different nights) and the group of NBC shows led by Hill Street Blues (all on different nights save for Cheers which precedes Hill Street Blues). Possible evidence of both scheduling effects and network loyalty could also be observed in the make-up of the two comedy factors. Within the "less sophisticated" group, three CBS comedies--Alice, One Day at a Time, and The Jeffersons--ran one after another on Sunday night, while two ABC shows--Happy Days and Three's Company--ran back-to-back on Tuesday night. For the "more sophisticated" comedy factor, two NBC shows, Family Ties and Facts of Life, ran back-to-back on Wednesday night. However, neither network loyalty nor scheduling effects can fully explain the loadings on these factors, since each factor included at least one show from each of the three networks as well as shows scheduled on at least four different nights. Among the "charming-cop" shows, Magnum P.I. and Simon and Simon ran back-to-back Thursday night on CBS, while ABC's Hart to Hart could be seen on Tuesday night. Among the evening soaps, CBS dominated with three of the four, two of which, Dallas and Falcon Crest, ran back-to-back on Friday night, while ABC's Dynasty was scheduled on an evening in which none of CBS's shows of that genre appeared. 57 In summary, the program groupings generated in the factor analysis do appear to reflect legitimate ”program- types," given the common characteristics of the shows loading on each factor (in the case of the NBC-factor, the program "type" involved appears'to have more to do with plot and production devices than with those elements traditionally used to distinguish genres) and the overall network and schedule mix which most of the factors embraced. As for the observed network loyalty and scheduling effects, they may reflect network executives' desire to retain and expand their audience by imitating a successful format and running several shows of the same type one after another on the same evening as much as they do audiences' network loyalty and viewing inertia. In contrast to these findings for broadcast program- types, the results of both correlational and factor analysis yielded no evidence pointing to the existence of cable channel types. Viewing Patterns and Political Characteristics: Implications for Political Narrowcasting Political involvement. In general, the data reveal a tendency toward relatively low levels of involvement in politics among heavy television viewers. Only four measures of content viewing were positively related to any form of political involvement, with two of these--news and C-SPAN-- devoting a large share of their airtime to direct coverage 58 of politics and public affairs. Only two television programs or program types with no current affairs or explicitly political programming were positively associated with any measures of political involvement. MASH was positively related to opinion leadership, and the "NBC- factor” was associated with relatively high levels of political interest and efficacy. The positive relationship of television news viewing to voting, as well as to other forms of political participation, was largely the result of older viewers' higher level of both. However, the negative relationships of a number of specific content variables to voting retained their strength subject to the same controls. The relative disinclination to vote of MTV, ESPN, USA Network, Wheel of Fortune, and both types of comedy show viewers, independent of age and gender, suggests a direct relationship between low voter motivation and viewing of these types of content. In light of this finding, these shows would seem to be particularly inappropriate vehicles for reaching likely voters. Though Wheel of Fortune was watched more by older than by younger viewers, it was associated with low levels of voting, political activity and efficacy. Thus, while a ratings-based narrowcasting strategy might expect to find a high concentration of active voters among Wheel of Fortune's older viewers, a strategy which considered the direct 59 relationship of political participation to specific television content viewing would discover that this is not the case. Heavy viewers of comedies, action-adventures, sports, music television (MTV), Wheel of Fortune and People's Court all revealed relatively low levels of political interest, efficacy and/or activity, and most of these relationships remained significant with age and gender controls. The one exception was the "more sophisticated comedy" shows, with women's lower level of political interest and participation and more frequent viewing largely explaining the observed relationship between these political characteristics and viewing of these types of shows. The low level of political involvement among viewers of these shows should not necessarily discourage political campaigns from buying time on such programs. In fact, research indicates that low involvement viewers may be more readily influenced by political messages than more interested and'involved voters: McClure and Patterson (1974), testing the influence of political messages on viewer attitudes, found that the greatest attitude change took place among heavy viewers with low political interest; studies by McLeod et al. (1974) and Iyengar et al. (1983) found less interested persons to be more prone to agenda- setting effects than more interested individuals; Schoenbach and Weaver (1985) found that the most politically ignorant 6O and uninterested voters gained the most consistency in their political cognitive structures as a result of an election campaign and: Rothschild (1978) found that in the low involvement situation it was possible to cause a change in behavioral intent without affecting attitude formation to the same degree. However, the extent to which low involvement viewers can be influenced by televised political messages will of course be affected by the type of message used. For example, long and complex messages providing less interested viewers abundant opportunity to switch to another channel, leave the room, or become otherwise distracted, could be expected to be less effective than shorter messages based on images and visual appeal (Perloff, 1985) which require minimal cognitive involvement and ”encourage [viewers] to perform their civic duty without [having] to evaluate information." (Rothschild, 1978). Because People's Court viewers appear more likely than non-viewers to vote, but also scored significantly lower than non-viewers on the political activity, opinion leadership and efficacy factors, this prime-time fringe show appears particularly well suited for a political advertising campaign designed to have an impact on the low involvement voter--for example, a well designed high-frequency, high visual impact, 30-second spot campaign designed to convince low-involvement viewers of the importance of voting for a 61 particular candidate. In addition, given the low cost of cable advertising, a similar campaign geared to the low- involvement viewers of ESPN, MTV and USA Network might be .able to generate new voters at relatively low cost. The effectiveness of such campaigns would likely be further enhanced if their messages emphasized the closeness of a political race and the differences between the candidates (Rust, et. al., 1984) and were geared toward the particular demographics of these networks--predominantly male for ESPN and USA Network and primarily young for MTV. Partisan and ideological targeting. The positive relationship of television viewing to Democratic partisanship and the apparently related disinclination to vote for Ronald Reagan actually became stronger when age and gender were controlled for. This finding seems to run counter to the image of President Reagan as television's "Great Communicator,” and is thus difficult to explain. However, it does suggest that Democratic candidates are more likely to reach their supporters via television than are Republicans. Only one content variable--USA Network viewing--could be identified as running clearly counter to this tendency. Its heavier viewers were considerably more likely to identify themselves as Republicans than non-viewers, as well as more likely to vote for Reagan if they voted. This cable network would therefore seem a good candidate for placement 62 of Republican campaign spots. However, heavy USA Network viewers are also less likely than others to vote. In spite of this, a Republican campaign might find it worthwhile to take advantage of this cable network's low advertising rates and relative abundance of availabilities (in comparison to broadcast networks) by running political spots designed to activate this relatively dormant support for their party. A similarly undermotivated Democrat constituency emerged as regular viewers of the less sophisticated comedy/entertainment shows, making these programs relatively good vehicles for messages designed to expand that party's base of regular voters. Conservative Democrats in particular could be most efficiently reached via these shows as well as by Wheel ovaortune, which was also associated with low levels of voting and political activity and efficacy. As noted in the above discussion of low-involvement viewers, these seemingly dormant partisan constituencies could probably be best activated by campaign messages which: sharply contrasted the candidates of the opposing parties; pointed to the closeness of the race (Rust, 1984); required minimal cognitive involvement (Rothschild, 1978): and, where appropriate, took into account the demographic makeup of the audiences (male for USA Network and older for Wheel of Fortune). As noted earlier, "NBC-factor" viewers emerged as particularly interested in politics and public affairs, a 63 finding which seems consistent with the fact that these shows portray more complex and controversial socio-political issues more frequently and more realistically than most other television shows. Political campaigns running spots on these shows could therefore reasonably expect to find more attentive audiences willing and able to digest messages dealing with relatively complex political issues. On the other hand, these viewers would tend to be less susceptible than less interested viewers to the influence of political advertisements which were not consistent with their previously held views. . NBC factor viewers also tended to consider themselves Independents and were relatively frequent ticket splitters. This relatively weak party identification combined with their high interest level suggests that these viewers would be particularly responsive to ads emphasizing a candidate's experience rather than his or her party affiliation (Shama, 1974). Since the viewing of this program type was also associated with self-reported liberalism, even when controlling for the effects of demographic variables, placement of political spots within these shows would seem to be particularly effective for liberal candidates. Self-reported liberalism was also characteristic of Arts and Entertainment network viewers, though they did not reveal a high level of interest in politics, even without controls. They also did not exhibit the relatively tolerant attitude toward communism and willingness to deal with the 64 Soviet Union which characterized the liberalism of "NBC factor" viewers, though they were significantly less militaristic than non-viewers. In light of these findings and the different nature of the two types of content involved, it is suggested that the liberalism of Arts and Entertainment viewers is somewhat different than that of "NBC factor" viewers. Perhaps the former is more strongly associated with the "liberal" arts and humanities and with refined, pacifist sensibilities which stand in contrast to the aggressive character of militarism. The liberalism of NBC factor viewers, on the other hand, appears to be less an abhorrence of conflict and violence (certainly Hill Street Blues has its share of both) as it is an active interest in social liberalism and left-of-center politics. "Macho cop" show viewing was positively related to hard-line positions on law and order issues and toward the USSR and communism, both with and without demographic controls. Perhaps this can be at least partly explained in terms of the good-guys versus bad-guys themes so unambiguously expressed in both this type of show and these kinds of socio-political attitudes. "Macho cop" show viewers also expressed more support for military intervention than non-viewers, though the partial correlation for this relationship was not statistically, significant at p=.05. These findings indicate that this type of show is probably the best vehicle for anti- communist, anti-crime and pro-military political messages. 65 Given the importance of opinion leaders in the dissemination of new ideas and information, MASH would seem to be a good vehicle for political messages attempting to introduce change. In light of the relatively liberal leanings of the show's viewers, advertising on MASH seems particularly well suited for a liberal candidate or campaign offering an innovative program and seeking to get its message out to an audience likely to be both receptive and inclined to pass it on to others. Interestingly enough, when MASH-viewing was analyzed in terms of partisan/ ideological self-identification, liberal Republicans, followed by liberal Independents, emerged as the heaviest viewers of the show. Thus, a liberal but strongly partisan message of a Democratic candidate might lose much of its effectiveness by alienating liberal Republicans and Independents. On the other hand, a liberal Republican or Independent candidate with a "new" message could reach a uniquely concentrated audience of potential supporters, many of them relatively young (r=-.20). The show might also be a good advertising vehicle for a candidate of any party who wanted to appeal to those Republicans who have become ' alienated from the party's conservative wing which has to a large extent dominated the party during the Reagan presidency. 66 VII. SUMMARY AND RECOMMENDATIONS This study has discovered a number of relationships between general and content-specific television viewing patterns and political attitudes and behaviors. While some of these disappear when controls for age and gender are introduced, quite a few are independent of these two demographic factors. This suggests that potential political narrowcasters might find it worthwhile to "look beyond the ratings" by pursuing this line of research further in order to develop a more refined understanding of the relationship of program and cable channel viewing to political characteristics. Table 10 summarizes the results of this study in terms of the most appropriate television programs or cable channels for reaching voter segments likely to be targeted in political campaigns. Programs or channels listed in the second column were found to be significantly related to the political characteristic described in the first column. Those programs or channels which are marked by an asterisk are those for which the partial correlations controlling for age and gender were also significant. In the third column of the table statistically significant age or gender characteristics of each voter segment are presented as well. The information in this column, combined with ratings data, could also be used to reach particular voter segments, 67 Table 10: Targeting Voter Segments Description of Target Voter Segment Likely voters Unlikely voters Opinion leaders Politically active Political inactive Interested in politics Not interested in politics High efficacy Host Appropriate Content (largest "r" listed first) News shows People's Court Cable subscribers HTV* Less sophisticated comedies* More sophisticated comedies* ESPN* Wheel of Fortune* C-SPAN* HASH News shows C-SPAN* Charming copy“shows* More sophisticated comedies* Wheel of Fortune* People's Court* News shows* "NBC factor" shows* Television generally* More sophisticated comedies "NBC factor" shows (e.g., Hill Street Blues) AgeZGender Older Younger Males Older/ Males Young/ Females Males Females All variables listed in columns 2 and 3 exhibited significant bivariate relationships with the corresponding "target voter" characteristic at p=.05. An asterisk (*) indicates that the partial correlation between a viewing measure and a political characteristic, controlling for age and gender, was also significant. 68 Low efficacy Liberals Conservatives Democrats Republicans Independents Liberal Republicans Conservative Democrats Likely Reagan voters Unlikely Reagan voter Ticket Splitters Party voters High trust in leaders Television in general* More sophisticated comedies* People's Court* Wheel of Fortune* Macho cop shows* Arts and Entertainment Channel“ "NBC factor" shows‘ C-SPAN MASH Magazine Shows Less sophisticated comedies* MTV Television generally‘ Monday Night Football USA Network* "NBC factor shows" MASH Wheel of Fortune More sophisticated comedies USA Network* TV generally* MTV "NBC factor" shows Charming cop shows 69 Younger Older Older Younger Younger Low trust in leaders Law and order oriented Support military intervention abroad Oppose military intervention abroad Hard-line anti-Soviet, anti-Communist Detente oriented: more willing to deal with Soviets, Communists Macho-cop shows* Wheel of Fortune Macho-cop shows Arts and Entertainment‘ More sophisticated comedies Macho-cop shows* More sophisticated comedies "NBC factor" shows‘ 70 Women Older Male/ Older Female/ Younger Male/ Younger Female/ Older instead of or, most effectively, in combination with the type of findings summarized in column two. Recommendations for Future Research The results of this study suggest to the author that it would be worthwhile to undertake similar surveys using current programs. Given the difficulty in obtaining accurate and reliable recall data, it may be preferable for such follow-up studies to measure viewing variables with questions asking respondents about what shows and channels they watched during the past week rather than in general as was done in this study. It is the authors view that the former would tend to be a more accurate reflection of actual viewing, as it might be affected by other family members' viewing choices, etc., while the approach used in this study appears more likely to reflect respondents' individual viewing preferences, independent of external influences which impact on their actual viewing. While the author believes that this approach is likely to yield data more readily interpretable, media buyers may be more interested in getting more accurate viewing data than in developing a coherent understanding of the relationships between respondents' political attitudes and behaviors and their viewing preferences. It might also be desirable for future studies to measure fewer variables than did the survey upon which this study was based, so that they can reach a larger sample 71 within comparable time and budgetary constraints. A larger sample would allow for more statistically robust analysis of cable channel viewing, since only about half of a given sample are likely to be cable subscribers, and among these, a much smaller percentage are likely to be even occasional viewers of certain cable channels. Expanding the sample would also be valuable in that it would allow for statistically meaningful analysis of data for specific demographic categories--e.g., males 18-24. This would increase a study's comparability to commercial ratings and polling data which tend to break out their data in terms of these demographic categories. By controlling for demographics in this way rather than by partial correlations, the study could also avoid potential problems in analysis associated with the latter when it must deal with variables that first increase and then later decrease with age, such as voting. With cable viewing and audience fragmentation an ever- increasing reality, the conceptual approach taken in this study seems a timely one for an actual campaign setting--for example, in a baseline study done at the start of a campaign or one undertaken just prior to a period of heavy television time buying. A campaign could combine a set of viewing questions with another set testing respondents' attitudes toward major campaign issues and images of the candidate (as well as some of the general measures included in this study, 72 such as likelihood of voting). By comparing these two sets of variables, campaign media strategists might discover especially appropriate (or inappropriate) programs or channels for presenting particular positions or qualities of their candidates. Because cable channel advertising is often sold by time period (e.g., a 30 sec. spot in a particular four-hour time block), instead of by specific program, questions about what time of day respondents watched particular cable channels would likely be useful to media buyers, though they would, of course, lengthen interviews. To compensate, prime time viewing questions might be limited to the ten or twenty most widely viewed shows among voting-age viewers. For those with more theoretical interest in this subject, the author suggests undertaking in-depth analysis of program types and themes such as that practiced by Chesebro (1976). Such research, particularly when combined with the kind of survey research practiced and discussed in this study, could be very useful in helping both political communicators and scholars' understand the nature of program-types as well as their relationships to political attitudes and behaviors. 73 APPENDIX: QUESTIONNAIRE TELEPWNE N0 . ( ) -... '; 1983 new USE/POLITICAL VARIABLES ID NO- ( ) ammo . 337-2oo2 - FINAL smus or rmnvrev; (1; counter: (2) INCOMPLETE (3) REFUSUAL 4 orscouum (,5) 00m: (6) omen: out a rmuvreven me one me ggsuu . 1. ' 2° 82 - busy 3° , NA ' no answer . 7' ‘CB - call back (give time) 5. _ _ _— INTRODUCTIM: --_ - =-- ‘==—= ....... HELLO. my name is ____and I'm calling from Michigan State University. We're doing a study of television viewing and public opinion in the Lansing area. If there is a MALEéFEMAL§_(INTERVIEWER --.ALTERNATE EVERY OTHER ONE AND CIRCLE) 18 years old or older at t is number, HIS’HER time to answer a few questions would be very helpful. Is there a MALEfiFEMALE 1 can speak to? IF FIRST CHOICE IS NOT AVAILABLE. ASK FOR THE OTHER SEX: Is there a NALELFEMALE I could speak to instead? REPEQT INTRO IF RESPONDENT HASN'T HEARD IT. THEN PLUNGE RIGHT IN: Thank you. n O... _1. On a typical day. how much time-do you spend listening to the radio? (I? ZERO. 60 T0 0.5) 2; Whit is your favorite radio station? (IF NONE, GO T0 0.5) 3. What is your second favorite station? 4. Do you ever listen to ‘All Things Considered" the evening news report on Public Radio WEAR? 1- yes 2- no 9- dk/refused (IF YES, ASK:) As. How many times have you listened to it in the past week? 5. On a typical weekday, how much time do you spend reading newspaper articles about politics and public affairs? 74) Page 2 b. C. 10 - 11. On a typical weekday, how much time do you spend watching television? 12. How about on a typical weekend? __ 0f the following daily newspapers. please tell me whether you read. each one of them frequently, occassionaly or practically never. _F_r_gg_. Occas 'lly Pro. .never USA TODAY ' - 3 2 l The Wall Street Journal 3 2 l The New York Times 3 2 “I Do you read any news or public affairs magazines? layes z-no_ 9-dk/refused (15' YES. nan) ‘ Which ones do you most often read? l-Time z-Newsweek 3-US News and World Report 4-other (specify): How much time during the past week did you spend reading new magazines? . - Are there any other magazines that you read quite regularly? l-yes 2-no 9=dklrefused (IF YES, ASK:) .Which ones? (LIST UP TO FIVE) (I? NONE, GO TO Q. 26) 4 (STARTING.0N THE SECOND MONDAY, 11/21, ASX:) lla. Did you watch the recently televised movie “The Day After"? 1' yes Z-no 9-ak. In. 909007? 1'- now going to read you a list of TV programs that are on each weekday. Please tell me how many times if any you've watched each sh type of show in the past week. (READ TITLES) ow or at a. b. c. d. The local news at 6 p.m. The nightly news at 11 p.m. The 7 to 9 a.m. early morning news shows The CBS Evening News with Dan Rather 75 .The NBC Nightly News with Tom Brokaw The ABC World News with Peter Jennings The McNeil-Lehrer News Hours Hour Magazine with Gary Collins The Peoples Court M*AfS*W* The Wheel of Fortune Entertainment Tonight _ The Nightly Business Report O 13. How often do you watch television during the weekday hours of 14. ' otppractically never? 9 a.m. to 5 p.m. ...very often. fairly often. only occasionally, lsvery often 2=fairly often 3-occasionall 4=practically never 9cdk/refused ~ (3F ONgY.OCCASIONALLY OR PRACTICALLY NEVER OR DK/REFUSED. GO TO 13a. On a typical weekday. how much time do you spend watching the daily serials. or soap operas? ~13b. How about the daytime quiz and game shows? How much time do you spend each day watching these kinds of shows? 13c. How many days a week, if any. do you watch the Phil Donahue show? How often do you watch television after ll:30 at night?.... very gften, fairly often, only occasionally. or practically never l-very often 2=fairly often 380ccasionally 4=practically never 98dk/refused IF ONLY OCCASSIONALLY, PRACTICALLY NEVER OR DK/REFUSED GO TO 0. 15) IF VERY OR FAIRLY OFTEN. ASK:) 14a. In the past week, how many times have you watch the following shows on late night television? (READ TITLES) a. The Tonight Show with Johnny Carson b; Nightline with Ted Koppel c. Thicke of the Night with Allen Thicke 76 15. What about on weekends? Do you usually watch a lot of TV during the day gn weekends. some TV, or not vegy much TV EEFTfig the day on week- en 5 l- a lot 2. some 3- not very much 9- dk/refused (IF NOT VERY MUCH. OR DK/REFUSED GO TO 0. 16) (IF A LOT OR SOME. ASKz) 'lSa. 0f the following types of TV shows, please tell me if you watch - each type of show uite lot, occassi ll . or hardl ver on weekends. Zfi (ES 3 a. Public affair and interview shows b. Sport Shows c. Religious program d. 01d movies !_— a k 16. How much time would you estimate you spent watching programs on Public Television in the past week? . 17. Have you watched any of the episodes in the Public Television series entitled "Vietnam - A Television History”? layes Z-no 9=dklrefused (IF YES ASK:) HOW MANY EPISODES HAVE YOU NATCHED? (RECORD NUMBER GIVEN or “ALL” IF THEY HAVE NATCHED EVERY EPISODE SO FAR) 18. I'm now going to read you a list of weekly television shows that are on during the evening prime-time hours this season. For each show that I name, please tell me whether you've been watching it pretty regulagy, only occassionally, or practically never. REGULARLY OCCASSIONALLl PR. NEVER 95 Nardcastle and McCormick ------------- 3 2 l 9 60 Minutes --------------------------- 3 2 l 9 The A Team-------; .................. 3 2 l 9 ‘Threes Company ———————— =t=-=__=-— 3 2 'l 9 Simon and Simon ---------------------- 3 2 l 9 Magnum P.I --------------------------- 3 2 l 9 9 to 5 -------------- - ---------------- 3 2 l 9 Webster ------------------------------ 3 2 l 9 Lottery --------- - -------------------- 3 2 l 9 Newhart ------------------------------ 3 2 l 9 77 - 5 WtGULAHLY ECIASIONALLI. PR. NEVER (5 Hill Street Blues --------------------- 3 2 l '9 I Monday Night Football-e --------------- 3 2 l 9 Jeffersons ---------------------------'-3 2 1 9 Matt Houston -------------------------- 3 2 l , 9 We Got It Made----------------a ------- 3 2 l 9 One Day at A Time-----------------L--+3 2 l 9 Scarecrow and Mrs. King --------------:-3 2 1 9 Hart to Hart ---------------,— --------- 3 2 .1 9 Cheers ---------------------------- b--+3 2 l ' 9 Benson -------------------------------- 3 2 l 9 Love Boat ----------------------------- 3 .2 l 9 Fall Guy ------ ----------------------- 3 2 l 9 AfterMash ----------------------------- 3 2 l 3 9 Emerald Point M.A.S ------------------- 3 2 l 9 .41 Thats Incredible ------- - ------------- a 2 1 9 0h Madeline --------------------------- 3 ‘2 -l 9 Remington Steele ---------------------- 3 2 l 9 Bay City Blues ----------------------- 3 2 l 9 Just Our Luck ------------------------ 3 2 l 9 Happy Days ---------------------------- 3 ‘ 2 l 9 Whiz Kids ----------------------------- 3 2 l 9 Real People --------------------------- 3 2 l 9 Facts of Life ----------------------- 3 2 l :9 Family Ties -------------------------- 3 2 l 9 St. Elsewhere--—----e ----------------- 3 2 l 9 Dynasty ------------------------------ 3 2 l 9 Hotel -------------------------------- 3 2 l 9 Mamas Family ------------------------- 3 2 l 9 Dukes of Hazzard --------------------- 3 2 l 9 20/20 -------------------------------- 3 2 l 9 Dallas ------------------------------- 3 2 1 9 Falcon Crest ------------------------- 3 2 1 9 Trauma Center-----: ------------------- 3 2 l i 9 T.J. Wooker- -------------------------- 3 2 l 9 Alice ................................ 3 2 l 9 78 21. 22.’ 23. 24. 25. 25. Do you own a home computer? layes Z-no 3-no computer but say they have a video game unit 9-dk/refused Do you own a videotape recorder? layes 2-no 9-dk/refused 1: cable televiison available in your area? layes 2-no 9-dk/refused (IF NO 0R OK. 90 T0 0 25) Do you have cable television in your household? l- yes z-no 9-dk/refused y(IF N0 OR OR. GO TO 0.25) Do you subscribe to any of the pay TV channels on cable? l-yes 2—no 9-dk/refused (IF YES’ ASK' ) Which ones? Now 1.m going to read you a list of cable television channels. Please give me your best estimate of how much time. if any, you ' spent.wmtching it in the past week. a. Cable News Network b. ESPN Sports Program , c. MTV. Music Television d. USA Network e. Cable Healt Network f. C-SPAN g. ARTS h. Satellite Program Network 263-2232227223: -°“(-“’§% :22 $883235??? Lu: a. how often do you watch TV for information? 1 2 .3. b. how'about fur excitement? ’ l 2 3 c. how about for amusement? l 2 3 d. how about in order to relax? l ‘2 3 e. and how often do you watch TV just for something to do? I 2 3 Which of the following would you say is your primary source of information about politics and public affairs? (READ THEM THE CHOICES) l-newspapers 2-radio 3-television A-magazines 5-personal discussions (STOP READING) 6-other 7-dk/refu. 79 _‘- 1 o- .- ...-..-... - .u . .. ~ v . sun's-a & 5 5 5 5 Often -9=dk 9'dk 98dk 9'dk 9¢dk -“ _.-—-— 27. 28. 29. 3]. 32. Among the other choices. which one would you say is your second most important source of political information? (READ REMAINING CHOICES If NECESSARY) l-newspapers- Z'radio 3-TV Asmagazines s-personal discussions 9-dk ' Some people pay a lot of attention to politics and political campaigns, while others don't. Now about you? Would you say that you are very interested, somewhat interestgg, or not very interested in politics andIEOlitical campaigns.? l-very 2-somewhat 3-not A-dk Wbuld you say that the government in Washington is pretty much run by a few big interests looking out for themselves or that it is run for the benefit of all the people? l-few big interests 2-fOr benefit of all 9=dk How much of the time do you think you can trust the government in Washington to do whats right for the country?...Alwa s. most of the time. some of the time. or practicallynever? l'always stt of time 3-sometimes Aspractically never 9-dk How about the leaders of private industry-~the executives that run America's large corporations? How much of the time do you think that they do what's right for the country? (READ CATEGORIES IF NEC.) . l-always Z-most of the time 3=some of the time A-practically never 9-dk I'm now going to read you a series of statements. Please tell me whether you stongly agree, agree, disagree or strongly disagree with each statements. 53. A o .52 or a. Its no use worrying my head about public affairs: 1 can't do anything about them anyhow l 2 3 4 5 b. Sometimes politics and government seem so complicated that a person like me can't really understand whats going on 1 2 3 A S c. voting is the only way that people like me can have any say about how the _ government runs things I 2 3 4 5 d. I feel that our political leaders hardly care what people like myself think or want 1 2 3 4 S 80 ' Page 8 There is practically no connection between what a plitician says and what he will do once he gets elected The U. S. has never fought in an unjust war Our treatment of criminals is too harsh: we should try to rehabilitate. not to punish them The police should not hesitate to use force to maintain order It would be all right if socialism replaced capitalism if it meant a better life for most people. In order to end the fighting in El Salvador. the U. 5. government should arrange for negotiations with the rebel forces even if it means that the Columnists would get some power in the government. It is the government's responsibility to coshelp people get medical care at low I I must admit that ’1 try to see what others think before I take a stand. A strong person doesn't show his emotions and feelings Nuclear energy is too dangerous to be relied upon as a major source of energy in this country. 18 l l A . 2 3 A .FOR THE NEXT GROUP OF QUESTIONS. I'M GOING TO READ TO YOU A NUMBER OF PAIRS OF STATEMENTS REPRESENTING OPPOSING POINTS OF VEIH ON A PARTICULAR SUBJECT. AFTER LISTENING TO EACH PAIR OF STATEMENTS. PLEASE TELL HE WHICH OF THE THO STATEMENTS YOU MST AGREE WITH AND WHETHER YOU STRONGLY AGREE WITH IT OR JUST TEND TO AGREE HITH IT. PERE IS TIE FIRST PAIR OF STATEFENTS FOR YOU TO CPDOSE FROM. (Read the o__r between the two statements) 81 35. 37. smnucu Tomas PREF TOWARDS 5190"ch A L HP b . 3 If citizens get together and apply pressure. they can frequently have some real influence on the government. 0" Even if citizens get together and try to apply pressure. they can hardly ever have any real influence on the government- NP b B A. There can't be much improvement in our'way of lives unless we are willing to question the old and accepted ways of doing things. 9:. People.who question the old and accepted ways of doing things usually end up causing trouble. NP b B fi—f The main cause of poverty in this country is that the American system doesn't give all people an equal chance. 2: Those who are poor usually have mainly themselves to blame. NP b B If our country does something that we feel is wrong. we should be prepared to criticize it. or Americans should be willing to defend their country. even if it has done . something that might be wrong. NP b B The main cause of violent rebellion in the less developed nations of the world is the poverty and social injustice that exist there. 2!. The main cause of violent rebellion in the less developed countries of the world are the effbrts of Russia and Cuba to export their Marxist revolutions to those countries. NP b B 82 ‘1. 42. TEND 'NO TEND STRONGLY TOWARDS PREF TOWARDS STRONGLY A If you act in faith with people. most of them w ll respond with fair- ness to you. 9!. If you give them the chance, most people will take advantage of you for their arm benefit A. Obedience and respect for authoritv are '1mpoztent virtues that every child should learn. 9; Self-reliance and a healthy skepticism towards authority are important virtues that every child should learn. The U.S. should give economic aid to the poorer countries of the world if they need our help. 9; The U.S. should concentrate on solving its own problems and let other countries make their own way as best they can. If a law is unjust. people should refuse to obey it. e: - Laws must be obeyed by everyone, even if they are sometimes unjust. ' All‘ the electronic and other techno- logical developments in recent years are inproving the quality of our lives 9.: All the electronic and other techno- logical developments in recent years are making our lives too complicated' 83 NP NP NP HP HP NP 1: ‘ ‘age II no Tend 43. A. This country'would be better off if we Just stayed home and did not worry abou: problems in other parts of the wor . 2: This country realistically cannot afford to ignore problems in other parts of the world or to remain uninvolved with them. The U.S. should place less emphasis on military weapons and more emphasis on negotiations . or It is important that the U.S. continue to develop new and more effective weapons. and not depend too much on negotiations. 45. The world has always been divided into different groups or nations. Since each nationality is different, this is as it should be. TEND STRONGLY TOWARDS PREF TOWARDS STRONGLY A a. NP b B A a NP b D A a up b '3 A a NP b B 2'; 'The world should not be divided into different nations -- all people should learn to live together as one community. l- str. approve Do you IF APPROVE OR STRONGLY APPROVE GO TO O. 47 IF DISAPPROVE OR STRONGLY DISAPPROYE, ASKz) Would you prefer that the U.S. strengthen its military presence in Lebanon or’would you prefer it if we moved in the direction of ‘6‘. withdrawing our troops from Lebanon? I'etrmngthen military 2-withdraw Orr 84 strongly;ggg:gye, aggrgyg, giggpgrove or strongly disapprove of President Reagan's handling of the problems in Lebanon? 2-approve . 3-disapprove 4-str. disapprove 9-dk 3-other 9-dk he 12 47. 49. 51. 52. 53. 54. Do you strongly approve, approve, disapprove. or strongly disapprove of President Reagan' s handling of the situation in Grenada? l-strongly approve 2-approve 3-disapprive 4-strongly disapprove 9-dk/refused Some people feel that its important for the United States to try very hard to get along with Russia. Others feel that its a big mistake to try to get along with Russia. Which of these two positions do you most agree with? l-should try hard Z-its a mistake to try hard 9-dk Would you say that you are in favor of capital punishment as a deterrant against crime or would you say that you are opposed to it? l-in favor 3-opposed Ztit depends 9-no opinion Generally speaking, do you usually think of yourself as a Republican, a Democrat. an Independent. or what? l- Republican 2- Democrat 3- Independent 4- Other 5-dk/refused 50a. (IF REP. OR DEN. ASK.:) Would you call yourself a strong or not very strong Rep. /Dem. ?) l-strong Republican 2-weak Republican G-weak Dem. 7-strong Dem. 98dk/refused 50b. (IF INDEP. OR OTHER ASK:l Do you think of yourself as closer to the Republican party or to the Democratic party? 3- Closer to Repub. 5- Closer to Demo. 4- niether Have you tried to convice anyone of your political ideas lately? layes Zeno 9-dk Has anyone asked your advice on a political question recently? l-yes Z-no 9-dk In tenms of your political views, would you say that you are.very conservative somewhat conservative. middle of the road, somewhat liberol or very liberal? l-very conservative z-somewhat conservative 4-somewhat liberal 5-very liberal 3-middle-of-road 9-dk/depends/refused Do you remember whether or not you voted in the 1980 presidential election and whom you voted for if you did vote? lsdidn't vote 2-voted for Carter 3-voted for Reagan 4-voted fon.Anderson 5-Clark G-Commoner 7-other 9-don't remember 85 :.--13 55. Assuming that President Reagan runs for reelection in l984 do you plan- to vote for him? 51.,” 2.1g depends] maybe/not sure/dk 3- no 9-refused (IF YES. GO TO 0. 57) ' ' 56. Can you name any of the men who have declared their candidacy for the l984 presidential nomination? l-yes Z-no (IF no. so TO 0.57) (IF YES. ASKz) Which ones? (Circle below those candidates which are named). Ratin s L a. Walter Mondale b; John Glenn c. Jesse Jackson d. Gary Hart e. Alan Cranston f. Senator Nollings g. Rubin Askew (NOW ASK:) Could you now tell me how much you support each of these (or this) candidate (s) on a scale of l to 9. with l meaning practically no support and 9 meaning complete support? (RECORD RATINGS IN SPACE PROVIDED ABOVE NEXT TO CANDIDATES NAMES) 57. When you cast your vote on election day. do you always vote for all the' candidates of the same party. do you usually vote for most of the candidates of the same party or do you cast your votes without much consideration of which party a candidate is affiliated with? l- always same party 2- usually party 3- don't consider party 4- don't vote 9-dk/refused 58. I'm now going to read you a list of activities related to government and politics. Please tell me whether you've participated in each activity more than once, only once. or never. more than once_ only once never g5, a. Attended a political rally 3 2 'l . 9 b. Contributed money to a political campaign 3 2 l 9 c., Worked for a candidate or in some other kind of political campaign 3 2 l 9 d. Written a letter to the editor of - a magazine or newspaper 3 2 l 9 86 Page'l4 .. f-‘A GD. 61. 62. 63. 3- yes. other family neuter does (Specify: ' more than once only once never dk e. Spoken or written to a political '— ' leader or official about a public . problem 3 2 l 9 f. Worked with other people or an organization in tryingto . something about a pub ic problem 3 2 ' l 9. Taken part in a sit-in. demonstration. or protest concerned with some public problem 3 . . 2 l 9 Do you belong to any political clubs or or organizations? ,l-tyes 2- no 3- dk/refused Do you belong to any other types of clubs or organizations? yes 2- not 9- dk/refused ("ye YES. AK: ) How many? Do you or any menber of your inmediate family belong to a labor union? (IF THEY DON'T SPECIFY WHO. ASK THEM) l- yes. I do 2- yes. my husband/wife does (Circle which one) 4- no 9- dklrefused. Now 1511:” organizational or other meetings have you attended in the past men And how many times in the past month have you gone out to a movie. a show or some other form of entertainment? Just a few more questions... How many persons live in your household? 0 l 2 3 4 5 6 7 8+ (”OK/REFUSED What is your age? ( ) Record exactly as stated What is your marital status? Single Cohabita ding (living with someone) Married Separated Divorced Widowed Dk/refused NOM.UN-l 87 ‘e 15 .1 67. 69. 70. 71. What was the last year of school you finished? 0 Less than high school Some high school High school graduate Trade or vocational school Some college or 2 yr. degree Bachelor's degree Graduate Training DK/refused ONGMbUN-fl Are you presently employed? l- yes 2- no (IF YES. ASK:) WHAT IS YOUR OCCUPATION? What is your racial or ethnic background; are you black,hispanic. white. or what? 1 Black 2 Hispanic 3 Oriental 4 White 5 Other 6 DK/refused What is your religion? Catholic Protestant, Jewish Other (Specify: ) DK/refused Ulhh’N-P Is your total annual household income more than $15,000? (----) res (1) No (Go to Closing) (9) DK/refused 52a. Is it more—than $25,000? — (g---) YES (2) No (Go to Closing) 52b. Is is more than $35,000? (rf-) YES _(2) no (Go to closing) 52c. In it more than $45,000? (5) YES (4) NO Thank you very much for your couperatinn. RECORD sex: (1). MALE (2) me 88 LIST 0? REFERENCES LIST OF REFERENCES Abramson, P., et al. Change and Continuity in the 1980 Elections. Washington, D.C.: Congressional Quarterly Press, 1982. Abramson, P. Political Attitudes in Americg: Formation and Change. San Francisco: W.H. Freeman and Co., 1983. Abramson, P., et al. Change and Continuity in the41984 Elections. Washington, D.C.: Congressional Quarterly Press, 1986. Atlas, M., ,"Gambling with Elections: The Problems of Geodemographics." Campaigns and Elections, Fall 1981, pp. 4-12. Brotman, "New Campaigning for the New Media." Campaigns and Elections, Fall 1981, pp. 32-34.. Capuzzi, Cecilia. "Wall Street's Affair With the Wire," Channels 1987 Field Guide: p. 68. Chesebro, J. W. "Communication Values and Popular Television Series--a four year assessment." In Newcomb, H. (Ed.), TelevisionL the Criticgl View. New York: Oxford University Press, 1976, pp. 8-46. Ehrenberg, A. S. C. "The Factor Analytical Search for Program Types." Journal of Advertising Research, Vol. 8, March 1968, pp. 55-63. Erikson, R., et a1. American Public Opinion: ng OriginsI Content and Impact. New York: John Wiley and Sons, 1980. Frank, R. and M. Greenberg. The Pleich Use oggTelevisipn: Who Watches and Why. Beverly Hills: Sage Publications, 1980. Gallup Poll. "Party Preference." March 10, 1983. Gallup Poll. "Foreign Policy Issues." April 21, 1983. Gallup Poll. "Approval of President Reagan." May 1, 1983. Candy, 0. H. Jr. "Television Audience Size and Composition." In Dervin, B. and Voight, M. J. (eds.). Progressgip 89 Mockus, J. "Geodemographics II: Targeting Your Turnout." Campaigns and Elections, Summer 1980. pp. 55-63. "My fellow subscribers, if elected-.." Cablevision, December 7, 1987, p. 170. Paley, W., and S. Moffett. ”The New Electronic Media- Instant Action and Reaction." Campaigns and Elections, Winter 1984. Pp. 4-12. . Perloff, R. "Personal Relevance and Campaign Information ‘Seeking: A Cognitive Response-Based Approach." In Kraus, S. and R. Perloff (eds.). Mass Media anq Political Thoaght: an information procesaing approach. Beverly Hills, London, New Delhi: Sage Publications, 1985. pp. 157-176. Pierce, J. et al. The Dynamics of American Pablic Opinion: Patterns and Processes. Glenview, Illinois: Scott, Foresman and Company, 1982. Poole, K., and L. Zeigler. Women, Public Opinion, and Politics: The Changing Political Attitages of American Women. New York: Longman, 1985. Power, J. '"Plug in to Cable TV.” Campaigns and Elections. September/October 1987, pp. 54—57. Rothschild, M. ”Political Advertising: A Neglected Policy Issue in Marketing." In Newman, 8. & Sheth, J. (eds.). Political Marketing: Readings and Annotated Bibliography. Chicago: American Marketing Association, 1985, pp. 142-155. Robbins, J. "Geodemographics: The New Magic." Campaigns and Elections, Spring 1980, pp. 25-34. Rust, R.. et al. "Efficient and Ineffient Media for Political Campaign Advertising." In Newman, B. & Sheth, J. (eds.). Political Marketang: Readings and Annotated Bibliography. Chicago: American Marketing Association, 1985, pp. 184-188. Sabato, L. The Ripe of Political Consultants. New York: Basic Books, 1981. Schneider, E. "Geodemographics Refined: An Alternative to Standardized Clustering." Campaigns and Elections. Winter 1983, pp. 62-73. 91 Schoenbach, K. and D. Weaver. "Finding the Unexpected: Cognitive Bonding in a Political Campaign." In Kraus, S. and R. Perloff (eds.). Mass Media and Political Thoaght: an information processing approach. Beverly Hills, London, New Delhi: Sage Publications. 1985. pp. 157-176. Shama, A. "Political Marketing: A Study of Voter Decision- Making Process and Candidate Marketing Strategy." In Newman, B. & Sheth, J. (eds.). Political Marketing: Readings and Annotated Bibliography. Chicago: American Marketing Association, 1985. Pp. 10-14. Swanson, Charles E. "The Frequency Structure of Television and Magazines." Journal of Advertisingygaaearch, Vol. 7, June 1967, pp. 8-14. Wells, William. "The Rise and Fall of Television Program Types." Journal of Advertising Research, Vol. 9, September 1969, pp. 21-27. 92